Magma recharge at Manam volcano, Papua New Guinea, identified through thermal and SO2 satellite remote sensing of open vent emissions

Magma recharge at Manam volcano, Papua New Guinea, identified through thermal and SO2 satellite remote sensing of open vent emissions | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Magma recharge at Manam volcano, Papua New Guinea, identified through thermal and SO 2 satellite remote sensing of open vent emissions Adam Cotterill, Emma Nicholson, Christopher Kilburn, Catherine Hayer This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3903120/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Oct, 2024 Read the published version in Bulletin of Volcanology → Version 1 posted 3 You are reading this latest preprint version Abstract Manam is one of the most frequently active volcanoes in Papua New Guinea and is a top contributor to global volcanic volatile emissions due to its persistent open vent degassing. Here, we present a multi-year time series (2018-2021) of thermal and SO 2 emissions for Manam from satellite remote sensing, which we interpret in the context of open vent feedbacks between magma supply, reservoir pressure, and outgassing. We classify the time series into four phases based on the varying SO 2 flux and observe a transient, yet substantial, increase in time-averaged SO 2 flux from background levels of ~0.6 kt day -1 to ~4.72 kt day -1 between March and July 2019. We also identify a transition from temporally-coupled to decoupled gas and thermal emissions during this period which we explain in the context of a magma recharge event that supplied new, volatile-rich magma to the shallow plumbing system beneath Manam. We infer that the arrival of this recharge magma triggered the series of eruptions between August 2018 and March 2019. These explosive events collectively removed 0.18 km 3 of degassed residual magma and signalled the onset of a renewed period of unrest that ultimately culminated in a major eruption on 28 June 2019. We quantify the magnitude of “excess” degassing at Manam after the removal of the inferred residual magma. SO 2 emissions reveal that ~0.18 km 3 of magma was supplied but only ~0.08km 3 was erupted between April 2019 and December 2021. We highlight how multi-parameter remote sensing observations over months to years enables interpretation of open vent processes that may be missed by short duration campaign measurements. magma recharge open vent remote sensing thermal anomalies SO2 degassing Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1. Introduction Open vent volcanism is sustained by the ascent and degassing of magma at shallow depths (Kazahaya et al. 1994 ; Harris et al. 1999 ; Shinohara 2008 ; Johnson et al. 2010 ; Palma et al. 2011 ), with variable contributions from both conduit convection and deep-derived segregated fluids that transfer both heat and volatiles to shallow reservoirs (Caricchi et al., 2018 ; Edmonds et al., 2022). Open vent systems exhibit a spectrum of eruptive styles: silica-rich magmas can form lava domes that may trigger vulcanian eruptions if collapse occurs (Stefan 1879 ; Robin et al. 1991 ; Wooster and Kaneko 1998 ; Calder et al. 1999 ; Young et al. 2003 ; James and Varley 2012 ; Mueller et al. 2013 ; Girina 2013 ; Flower and Carn 2015 ; Shevchenko et al. 2020 ), while more mafic magmas support open conduit conditions allowing increased mobility of both melt and volatiles and, in rare cases, maintain lava lakes over years to decades (Moussallam et al. 2016 ; Patrick et al. 2019 ; Gray et al. 2019 ; Lev et al. 2019 ). A common characteristic of open vent volcanoes is persistent degassing (Rose et al. 2013 ; Vergniolle and Métrich 2021 ), where the volume of magma required to supply the observed volatile flux exceeds that erupted; this is referred to as the “excess degassing phenomenon” (Kazahaya et al. 1994 ; Shinohara 2008 ). The fate of this unerupted degassed magma is often explained by conduit convection (Kazahaya et al. 1994 ; Beckett et al. 2014 ; Coppola et al. 2022 ) or by intrusion and endogenous crustal growth (Coppola et al., 2019 ). In both scenarios it remains unclear where this degassed magma accumulates in the crust. Typically, SO 2 flux is used to determine the magma supply rate (Allard et al. 1994 ; Andres and Kasgnoc 1998 ; Shinohara 2008 ) and the magma output is often approximated using thermal emissions (Wooster and Kaneko 1998 ; Laiolo et al. 2018 ; Coppola et al. 2019 , 2022 ). The balance between magma input and output, and changes in this budget through time, has been shown to be highly indicative of pressure and fluid dynamic perturbations within the shallow magma storage region, which can disturb open conduit processes and may ultimately lead to eruptions. Manam (-4.078, 145.038) is a frequently erupting mafic open vent stratovolcano located ~ 19 km off the northeast coast of mainland Papua New Guinea (Fig. 1 ) (Palfreyman and Cooke 1976 , 2021a). Manam is situated within the Western Bismarck Arc where the arc-continent subduction has ceased and melting of the remnant hanging slab is the source of volcanism (Abbott et al. 1994 ; Abbott 1995 ; Woodhead et al. 2010 ; Holm and Richards 2013 ) and, historically, it has erupted basaltic to basaltic andesite magma compositions. The volcano has two active craters that exhibit different activity. Main Crater hosts degassing from a broad fumarole field and is the vent from where most lava effusions originate. South Crater emits a near-constant dense gas plume and explosive eruptions tend to be generated from this crater. The top surface of the magma column was observed at shallow depths within South Crater on 22 May 2019 (Liu et al. 2020 ) but the temporal persistence of this state is currently unknown aside from occasional reports of incandescence (2021a). Manam is a top emitter of volcanic volatiles in a global context and is therefore an important case study to explore temporal variability in emissions and their relationship to subsurface processes. Satellite measurements of SO 2 using NASA’s Ozone Monitoring Instrument (OMI) between 2005–2015 indicate an average SO 2 flux of 1480 ± 750 tonnes day − 1 placing Manam as the 11th strongest emission source globally. (Carn et al. 2017 ). Similarly, Manam’s CO 2 flux has been estimated at 2760 ± 1570 tonnes day − 1 making it the 10th highest emitter of volcanic CO 2 (Aiuppa et al. 2019 ). Direct sampling of Manam’s plume using a Unoccupied Aerial Vehicle (UAV) mounted MultiGAS (Shinohara 2005 ; Aiuppa et al. 2007 ) in May 2019 revealed an elevated average SO 2 flux of 5150 ± [336/733] and CO 2 flux of 3760 ± [313/595] (Liu et al. 2020 ). These campaign-style measurements ranked Manam, albeit transiently, at 2nd and 5-7th respectively for volcanic SO 2 and CO 2 production globally. Manam generated 29 major and 139 minor eruptions between 2000 and 2021 (Palfreyman and Cooke 1976 , 2021a). Over the same interval, frequent clusters of thermal anomaly detections were identified from satellite multispectral imagery. Following large-scale evacuations associated with the major eruptions in 2004–2005 (Johnson 2013 ; Connell and Lutkehaus 2016 ), approximately 4000 residents had returned to the island by August 2021 and experience ongoing impacts to agriculture, settlement infrastructure, and water supplies due to persistent volcanic activity (J Sukua, Pers Com., 2021). A new phase of eruptive activity began in August 2018 after 11 months of quiescence. A series of 23 eruptions then took place until March 2019, including 5 major eruptions (i.e. eruption columns > 10 km) on 25 August 2018, 8 December 2018, 7 January 2019, 11 January 2019, and 23 January 2019. Lava effusions occurred between 27 September and 1 October 2018 and on 8 January 2019, which reached within 500 and 400 m of the coastline respectively. The largest eruption in recent years occurred on 28 June 2019, generating an eruption column that rose to 15.2 km asl, pyroclastic density currents, and a lava flow reaching within 700 m of the coastline. During this eruption, 3775 residents evacuated temporarily and 455 homes and agricultural gardens are reported to have been destroyed or damaged (2021a). Ground-based volcano monitoring is challenging due to the island setting, dense vegetation, steep topography and tropical climate, and therefore key gaps remain in observational capability despite the high level of volcanic risk. Improving understanding of the relationship between satellite remote sensing observations and subsurface volcanic processes are therefore critical to augmenting future volcano monitoring at this, and other, open vent volcanoes. Satellite remote sensing provides regular measurements and near-global coverage of volcanic SO 2 (Theys et al. 2019 ) and thermal emissions (Wright et al. 2004 ; Coppola et al. 2016 ) enabling monitoring of remote or inaccessible volcanoes, such as Manam. Here, we present the first multi-year time series from 2018 to 2021 of (a) SO 2 emissions derived from the European Space Agency (ESA) Sentinel-5P Tropospheric Monitoring Instrument (TROPOMI), (b) thermal anomaly detections from the National Aeronautic and Space Administration (NASA) Moderate Resolution Imaging Spectrometer (MODIS) instrument processed by the MODVOLC algorithm, and (c) surface temperature measurements of Main and South Crater using ESA’s Sentinel-2 MultiSpectral Instrument (MSI). We use this multi-parameter time series to interpret magmatic processes influencing transitions in open vent behaviour at Manam, quantify the magnitude of excess degassing and consequently the volume of unerupted degassed magma; and evaluate the time-varying contribution of Manam to global volcanic volatile emission inventories. 2. Methods 2.1 Sulphur Dioxide (SO 2 ) Emissions We quantify SO 2 emissions using the TROPOMI spectrometer on board ESA’s Sentinel-5P polar orbiting platform. TROPOMI has a spectral resolution of 0.25 to 0.54 nm and a spatial resolution of 3.5 × 7 km at launch (Veefkind et al., 2012) and updated to 3.5 × 5.5 km on 6 August 2019. TROPOMI observes every point on the Earth’s surface at least once per day. The TROPOMI Differential Optical Absorption Spectroscopy (DOAS) retrieval algorithm calculates SO 2 Vertical Column Densities (VCDs) for each pixel within its field of view (Theys et al., 2017); VCDs are then converted to column mass (Queißer et al., 2019). The total SO 2 mass loading for a given scene is calculated by summing the column mass of SO 2 contained within each pixel above 3 times the random noise. The TROPOMI instrumental response to SO 2 is height-dependent and therefore the assumed plume altitude is the main source of uncertainty in SO 2 retrievals. Gas plume or eruption column heights, where reported by RVO or Darwin Volcanic Ash Advisory Centre (VAAC), were used to represent the SO 2 plume altitude. Visual observations from May 2019 suggest that the buoyant gas plume generally rises between a few hundred metres to ~ 1 km above the summit (~ 1800 m asl) before dispersing laterally (Liu et al. 2020 ). Therefore, for days without reported plume height an altitude of 3 km was used. It is important to identify other SO 2 sources that might contaminate a scene resulting in overestimation of SO 2 mass flux from Manam. Identified contamination sources include two nearby volcanoes: Kadovar (-3.6069, 144.5878), which has been outgassing regularly in recent years (Plank et al. 2019 , 2020 , 2021b) and Ulawun (-5.0514, 151.3310), which has had five confirmed eruptions during the study period (Johnson 2013 ; Wood et al. 2019 , 2021c; McKee et al. 2021 ). TROPOMI scenes contaminated by SO 2 from external sources are identified from (a) activity reports for nearby volcanoes and (b) visual inspection of true colour Sentinel-2 images and TROPOMI VCD scenes (Fig. 2 ). Where possible, the measurement extent is delimited to include only the plume from Manam. If the external sources cannot be clearly separated then the contaminated images are omitted from the time series. Converting scene SO 2 mass into a flux requires knowledge of the residence time of SO 2 in the atmosphere. If the SO 2 lifetime exceeds 24 hours, then some proportion of the SO 2 mass in a TROPOMI scene will be residual from the previous day. Uncertainties in the lifetime of SO 2 in the atmosphere make converting total scene mass to SO 2 fluxes non-trivial, especially under different atmospheric conditions (McCormick Kilbride et al. 2019 ). Here, the method proposed by Fioletov et al., ( 2015 ) is used where under steady-state emissions the flux ( \(\varvec{\Phi }{\varvec{S}\varvec{O}}_{2}\) ) and SO 2 mass are related by Eq. 1 : $$\varvec{\Phi }{\varvec{S}\varvec{O}}_{2}=\frac{{\varvec{M}\varvec{S}\varvec{O}}_{2}}{\varvec{\tau }}$$ 1 where \({\varvec{M}\varvec{S}\varvec{O}}_{2}\) is SO 2 (tonnes) and \(\varvec{\tau }\) is the residence time of SO 2 in the atmosphere. Three estimates for residence time were used in an attempt to capture the uncertainty related to this variable: 1, 2 and 3 days based on residence times used in similar studies (Beirle et al. 2014 ; Laiolo et al. 2018 ; McCormick Kilbride et al. 2019 ; Liu et al. 2020 ). The maximum atmospheric residence time for SO 2 plumes from Papua New Guinea volcanoes has previously been estimated to be ~ 18 h with typical ages being < 12 h (McCormick et al. 2012 ). Therefore, the fluxes based on a 1-day residence time are used in the discussion. Lastly, we calculate time-averaged SO 2 fluxes by fitting a first-order polynomial to the cumulative \(\varvec{\Phi }{\varvec{S}\varvec{O}}_{2}\) emissions (Fig. 3 ). 2.2 Thermal Anomalies The MODVOLC algorithm measures the radiant heat flux, or volcanic radiative power (VRP), emitted by Manam (Wright et al. 2004 ; Wright 2016 ). MODVOLC uses Level 1B products from the MODIS multispectral instrument onboard the NASA Aqua and Terra satellites, providing a 1 km 2 pixel resolution for the infrared bands. These two satellites ensure coverage of most of Earth’s surface every 1–2 days. VRP is the total heat radiated across the area of the anomaly at the time of acquisition and is expressed in W or J s − 1 according to Eq. 2 (Coppola et al., 2013 ; Wooster et al., 2003 ; Wright et al., 2015 ). $$\varvec{V}\varvec{R}\varvec{P} \left({\varvec{\varphi }}_{\varvec{e}}\right)=1.89 \times {\varvec{A}}_{\varvec{P}\varvec{I}\varvec{X}} \times {\varvec{D}}_{\varvec{P}\varvec{I}\varvec{X}}$$ 2 where 1.89 is a best fit regression coefficient calculated using the MIR (Middle Infrared) method (which relates the VRP estimated by the simple power law used by MODVOLC to the expected value under the Planck function; Wooster et al., 2003 ), \({\varvec{A}}_{\varvec{P}\varvec{I}\varvec{X}}\) is the area of the pixel, and \({\varvec{D}}_{\varvec{P}\varvec{I}\varvec{X}}\) is the above-background MIR radiance of the pixel. When a hotspot is detected in more than one pixel, the total VRP is the sum of the VRP across all hotspot pixels. Thermal anomaly intensity is classified following Coppola et al ( 2016 ; also see “MIROVA,” 2022): <1 MW = Very Low, ≥ 1 MW = Low, ≥ 10 MW = Moderate, ≥ 100 MW = High, ≥1GW = Very High and ≥ 10 GW = Extreme. Manam’s two craters are within 1 km of each other and so the measured VRP likely includes thermal contributions from both due to MODIS’s pixel resolution. To determine the source of each anomaly, we use (a) thermal infrared (TIR) imagery from NASA’s Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) (90 m resolution) and (b) short-wave Infrared (SWIR) (20 m resolution) and True Colour Imagery (TCI) (10 m resolution) from ESA’s Sentinel-2 MSI (Fig. 4 ). 2.3 Surface Temperatures Surface temperature within each of Manam’s summit craters are derived from the Sentinel-2 MSI Level 1C product (Top of Atmosphere (TOA) Reflectance). First, TOA reflectance measured from the MSI band 11 (central wavelength of 1610 nm) is converted to radiance using Eq. 3 : $$\varvec{L}\varvec{\lambda }=\frac{{\varvec{Q}}_{\varvec{c}\varvec{a}\varvec{l}}{\varvec{E}}_{\varvec{e}\varvec{\lambda }}\left(\varvec{c}\varvec{o}\varvec{s}\varvec{\theta }\right)}{\varvec{\pi }\left(\frac{1}{\varvec{U}}\right)}/{10}^{4}$$ 3 where \({\varvec{Q}}_{\varvec{c}\varvec{a}\varvec{l}}\) = digital number (DN); \({\varvec{E}}_{\varvec{e}\varvec{\lambda }}\) = solar irradiance (W m 2 ); \(\varvec{\theta }\) = incidence angle (°), and U = quantification value (converts value to TOA). \({\varvec{E}}_{\varvec{e}\varvec{\lambda }}\) , \(\varvec{\theta }\) and U are drawn from the MSI image metadata. Pixel Integrated Temperatures (PIT) are then calculated for each band using Eq. 4 (adapted for use with MSI imagery from Francis and Rothery, 1987 ; Rothery, 1988; Harris, 2013 ), which is derived from the Planck function (Planck, 1901 ). $$\varvec{T}=\frac{{\varvec{C}}_{2}}{\varvec{\lambda }\varvec{l}\varvec{n}\left(\left[\varvec{\epsilon }{\varvec{\tau }\varvec{C}}_{1}{\varvec{\lambda }}^{-5}/{10}^{6}\varvec{\pi }{\varvec{L}}_{\varvec{\lambda }}\right]+1\right)}$$ 4 where \({\varvec{L}}_{\varvec{\lambda }}\) = radiance (Wm − 2 sr − 1 µm − 1 ); \({\varvec{C}}_{1}\) = 3.742 × 10 –16 (W m 2 ); \({\varvec{C}}_{2}\) = 0.0144 (mK); \(\varvec{\lambda }\) = wavelength (m); T = blackbody temperature (K); ε = emissivity of the radiating surface; and τ = atmospheric transmissivity. \({\varvec{C}}_{1}\) and \({\varvec{C}}_{2}\) are simplified constants representing hc 2 and hc/k , where h is Planck's constant (6.266 × 10 –34 J s), c is the speed of light (2.998 × 10 8 m s −1 ) and k is Boltzmann's constant (2987 µm K). Emissivity for the basaltic andesite lava erupted at Manam (Palfreyman and Cooke 1976 ; McKee 1981 ) is estimated as 0.852 for band 11, based on analogous lavas measured for emissivity in the John Hopkins ECOSTRES Spectral Library (Meerdink et al. 2019 ). Transmissivity for Manam was estimated using MODTRAN (MODerate resolution atmospheric TRANsmission) to be 0.892 for band 11 (Berk et al. 2014 ) (Table S1 ). 3. Results 3.1 SO 2 Emissions Daily SO 2 mass loadings from Manam between 6 May 2018 and 31 December 2021 have a mean of 1.1 kt day − 1 and a median of 0.47 kt day − 1 (Fig. 5 e). The time series is dominated by several short-duration, high-magnitude emissions associated with explosive eruptions with 75% of daily emissions below the mean. Using a 7-day moving-average, we define four degassing phases using a 1 kt day − 1 threshold to distinguish Phases 1 and 4 (below threshold) from Phases 2 and 3 (above threshold), which refer to the background and elevated degassing phases respectively. The elevated degassing phases are identified by the moving-average exceeding 1 kt day − 1 and subsequently not dropping below this threshold for more than 8 days. The transition between Phase 2 and 3 is demarcated by a gradient change of cumulative SO 2 emissions, which occurred from 21 July 2019 (Fig. 5 e). Phase 1: 6 May 2018–21 March 2019 SO 2 emissions for Phase 1 totalled 198.8 ± 8.3 kt over 319 days, with a time-averaged SO 2 flux of 0.62 kt day − 1 (Fig. 4 ). After an 11-month period of eruptive quiescence, explosive eruptive activity returned on 10 August 2018, after which 13 minor explosive eruptions, 5 major explosive eruptions and 6 effusive eruptions occurred during Phase 1. The daily emitted SO 2 mass exceeded 3 kt on ten occasions, all associated with explosive behaviour. The largest Phase 1 emission of 22.1 kt SO 2 is linked to explosive and effusive activity on 25 August 2018. Phase 2: 22 March 2019–20 July 2019 Phase 2 began when the 7-day moving average exceeded 1 kt day − 1 and did not drop below this threshold for more than 8 consecutive days. Manam emitted 539.9 ± 10.6 kt of SO 2 over 120 days, with a time-averaged flux of 4.72 kt day − 1 , during Phase 2 (Fig. 4 ). This period contained relatively few eruptive events with 5 minor eruptions and 1 major eruption on 28 June 2019. Despite the lower eruption frequency, SO 2 emissions prior to the major eruption were elevated substantially above the Phase 1 time-averaged flux. A total of 325.8 ± 7.6 kt SO 2 was released prior to the 28 June 2019 eruption, with most of these emissions not linked with documented eruptions. The 28 June 2019 eruption alone emitted 58.3 ± 0.5 kt of SO 2 and 82.1 ± 0.4 kt released over the following four days. Phase 3: 21 July 2019–16 October 2019 Phase 3 SO 2 emissions were reduced relative to Phase 2 with a total emitted mass of 127.9 ± 4.1 kt over 87 days and a time-averaged flux of 1.5 kt day − 1 (Fig. 4 ). However, emissions remained elevated above the background flux observed during Phase 1. Phase 3 emissions were mostly independent of eruptive activity, with only 3 minor eruptions occurring during this time. Phase 4: 17 October 2019–31 December 2021 Phase 4 began when the 7-day moving average fell below 1 kt day − 1 for 10 consecutive days and remained consistently below this threshold for the remainder of the time series. Phase 4 SO 2 emissions were 565.1 ± 24.9 kt over 806 days, with a time-averaged flux of 0.68 kt day − 1 (Fig. 4 ). This flux is comparable to that observed in Phase 1 and therefore we interpret these two phases as representative of the stable background SO 2 emission rate. The small peaks in SO 2 emissions during Phase 4 are associated with the 27 minor and 2 major explosive eruptions during this period. 3.2 Thermal Anomalies Thermal emissions have been detected at Manam sporadically since MODIS coverage began in 2002. During this study period, thermal emissions are characterised by periods of frequent elevated VRP, separated by intervals of several months to years with no detectable thermal output (Fig. 5 e). The time series is grouped into three discrete clusters of elevated VRP, where each cluster includes at least 5 thermal anomalies separated by intervals of no longer than 60 days. Cluster 1: 7 August 2018–18 July 2019 Cluster 1 began 3 days prior to the onset of explosive eruptive activity in August 2018 and is linked to the eruptions during late 2018 and early 2019. Thermal anomalies became more sporadic in May 2019 but subsequently increased in frequency coinciding with a series of eruptions in June 2019 that culminated in the 28 June 2019 major eruption. Cluster 1 contains 44 detected anomalies, 18 of which are classified as high intensity and are associated with explosive eruptions (Fig. 5 D). Two anomalies are classified as very high intensity and are both coincident with extensive lava flows that almost reached the coastline on 30 September 2018 (2211 MW) and between 8 and 11 January 2019 (1371 MW). Cluster 2: 29 July 2020–8 September 2020 Cluster 2 includes 5 low to moderate intensity thermal anomalies ranging between 8 and 50 MW. These detections coincide with a period of minor explosive eruptions reported between July and September 2020 (Fig. 5 d). Cluster 3: 19 June 2021–21 December 2021 Cluster 3 occurred during a period of unrest that began in June 2021 continuing throughout the remainder of 2021. The cluster began with a series of low to moderate intensity thermal anomalies following a minor explosive eruption on 23 June 2021. Increased thermal emissions in August 2021 were reflected in 11 anomalies, which included three high intensity detections. Four moderate to high intensity anomalies were detected following reports of Strombolian activity on 18 October 2021 and prior to the 20 October 2021 major eruption. Several low to moderate anomalies were detected in late November to December 2021 in the weeks prior to another major eruption on 22 December 2021 (Fig. 5 d). 3.3 Surface Temperatures Through the period of observation, Main and South Crater were completely obscured by cloud (meteorological or volcanic) in approximately 75% of the 277 Sentinel-2/MSI images available (Fig. 5 b). Surface temperatures were therefore calculated for Main and South Craters from 68 and 71 images, respectively. The daily maximum PIT for Main Crater ranged from 340 ± 83°C to 509 ± 124°C and 335 ± 79°C to 510 ± 124°C for South Crater. The mean inter-crater maximum PIT divergence was 31°C and a maximum of 158°C where the temperature of both craters could be measured together with South Crater having a higher maximum PIT 40 of 60 days, just 18 of these exceeding mean inter-crater divergence. There was just one occurrence where Main Crater maximum PIT exceeded South Crater by more than the mean divergence, a 57°C difference on 27 September 2018 during the emplacement of a lava flow from Main Crater. The maximum South Crater PIT during the study period was measured on 20 May 2019 but Main Crater was cloud-covered at the time, preventing direct comparison. However, the following simultaneous measurement of both craters—on 30 May 2019—has the second highest inter-crater temperature divergence of 122°C. A bright hotspot at South Crater in MSI thermal imagery was observed on this date and follows UAS in situ observations of shallow magma within South Crater on 22 May 2019 (Liu et al. 2020 ). Cloud-cover prevented MSI measurements throughout June 2019 and therefore the 28 June 2019 major eruption is not captured in this time series. However, following this eruption, a further two maximum pixel-integrated temperatures of 500 and 501°C, with means of 455°C and 473°C, were measured at South Crater on 14 July and 19 July 2019 respectively. This suggests a period of extended high temperatures at South Crater through May to July 2019, during which a major eruption occurred. The mean PIT on 13 August 2019 (403°C) and 18 August 2019 (408°C) decline more substantially than the maximum temperatures, 497°C and 475°C respectively, compared to the two retrievals in July (Fig. 5 a). As the difference between maximum and mean PIT is indicative of the spatial distribution of the hot radiating surface, we interpret this result to indicate that although hot material remained in South Crater it likely occupied a smaller portion of the crater area. 4. Discussion 4.1 Coupling between SO 2 and Thermal Emissions We present a combined time series of SO 2 and thermal emissions (Fig. 5 ), which are key parameters for observing changes in open vent activity where an established connection exists between a shallow reservoir and the surface (Wright et al. 2004 ; Sparks et al. 2012 ; Pyle et al. 2013 ; Blackett 2013 ; Aiuppa 2015 ). Thermal anomalies associated with volcanic edifices may indicate that magma is at or near the surface (Coppola et al., 2012; Dehn and Harris, 2015; Harris, 2013 ) and SO 2 emissions provide an insight into conduit permeability (Edmonds et al. 2003 ), magma convection (Shinohara 2008 ), and the volume of degassing magma present at shallow depths (Allard et al. 1994 ; Aiuppa et al. 2017 ). Thermal and SO 2 emissions have been shown to be coupled at during background level activity at open vent systems such as Stromboli (Italy) (Laiolo et al., 2022 , 2018 ), Bagana (Papua New Guinea) (McCormick et al. 2012 ; McCormick Kilbride et al. 2019 ), Batu Tara (Indonesia) (Laiolo et al., 2018 ), Tinakula (Solomon Islands) (Laiolo et al., 2018 ), and Mt. Etna (Italy) (Coppola et al., 2019 ; D’Aleo et al., 2019). Contrastingly, periods of uncoupled behaviour between these two parameters can signal transient disturbances to the magmatic system. At Stromboli, elevated SO2 emissions generally lag behind peaks in thermal emissions associated with paroxysms and lava flows (Laiolo et al. 2022 ). Periods of high SO2 flux but low radiant heat flux during quiescence phases are typically attributed to unerupted magmatic intrusions (e.g. Mt Etna 2005-06, Coppola et al., 2019 ). Conversely, periods of below-average SO2 flux but high radiant flux have been explained by extrusion of previously degassed magma (e.g., Tinakula 2006–2012; Laiolo et al., 2018 ). Here, we evaluate the degree to which the SO 2 and thermal emission timeseries at Manam are coupled for each of the four degassing phases (Fig. 5 ). The correlation between thermal and SO 2 emissions is calculated using weekly total emissions as it allows the mostly continuous SO 2 flux to be compared to the intermittent thermal emissions. The correlation between total weekly SO2 emissions and total weekly volcanic radiative power is weak across all phases with Phase 1 having the strongest correlation coefficient (r) of 0.27 (Fig. S2). These weak correlations are likely due to the fact that peak VRPs are associated with lava effusions whereas peak SO 2 emissions typically coincide with explosive eruptions as well as the fact that SO 2 emissions have a range of 3 orders of magnitude (0.92–96 kt) compared to 4 for thermal emissions (5–2792 MW). While the correlation between the magnitude of the two parameters is weak the temporal relationship of the SO 2 and thermal emissions throughout the time series can be used to interpret processes governing the observed activity and this temporal aspect is explored throughout the remainder of this work. Throughout Phase 1, peaks in SO 2 emission occurred coincident with periods of heightened radiant flux and were typically aligned with an observed eruption (Fig. 5 ). On five occasions, thermal anomalies begin to be detected days to weeks ahead of eruptions with coincident elevated SO 2 emissions (e.g. August 2018, October 2018, December 2018, January 2019 and March 2019, Fig. 5 , Fig. S1 ). This was not the case for the 8 January 2019 and 23 January 2019 eruptions, where SO 2 emissions peaked without thermal anomaly detections, likely due the presence of cloud and the ash-rich plume obscuring thermal detections. Throughout Phase 2, SO 2 and thermal emissions are not well correlated temporally, except around the 28 June 2019 eruption (Fig. 5 ). The elevated outgassing characterising this phase (4.72 ± 0.09 kt day − 1 ) occurred alongside four thermal anomalies in March-May, one of which was coincident with an above-average SO 2 emission. Magma was observed in situ deep within South Crater on 22 May 2019 (Liu et al. 2020 ) (Fig. 6 ) but no anomaly was detected, likely due to cloud cover obscuring the summit at the time of the satellite overpass (10:30 AM and 1:30 PM). MSI SWIR hotspots were observed on 20 May 2019 and 30 May 2019, both with high surface temperatures (Fig. 5 a), which suggest similar conditions were likely present on 22 May 2019. Persistent cloud cover during Phase 2 obscured MSI observations (Fig. 5 b) and, if MODIS acquisitions were similarly affected, under-reporting in the frequency of thermal anomaly detections is likely during this period. Nevertheless, the frequency of thermal anomalies increased in June 2019 alongside an increase in eruptive activity compared to the prior two months, including the 28 June 2019 major eruption. Following an 11-day period of subdued SO 2 emissions between 3 and 13 June 2019, during which < 7 kt was emitted over 9 days, elevated gas emissions are resumed alongside the escalating frequency of thermal anomalies. Elevated SO 2 emissions continued in Phase 3 (1.5 ± 0.05 kt day − 1 ) but were much reduced compared to Phase 2. Only one thermal anomaly was detected coinciding with an SO 2 emission of 1.9 kt on the day of a minor eruption 29 September 2019 (Fig. 5 ). Overall, Phase 3 represents an extended period where gas and thermal emissions appear to be decoupled, though both parameters appear to be declining in intensity. SO 2 emissions in Phase 4 (~ 0.68 ± 0.03 kt day − 1 ) returned to a comparable level to Phase 1 (~ 0.62 ± 0.03 kt day − 1 ). Phase 1 exhibited temporal coupling between thermal and SO 2 emissions, but this is not the case in Phase 4. Although instances where thermal anomalies coincide with above-average SO 2 emissions occur (e.g. 20 October 2021), most above-background emissions occur when no thermal anomalies are detected. During thermal Clusters 2 and 3, the magnitude of SO 2 emissions are unchanged relative to outside these periods of heightened thermal emissions (Fig. 5 ). Consequently, we propose that both coupled and decoupled regimes in SO 2 and thermal emissions are present within the study period, and that a transition from coupled to decoupled behaviour occurred following the last eruption of Phase 1 on 29 March 2019. Phase 1 operates under the coupled regime, where peaks in the two parameters are well correlated in time but not in absolute magnitude. In contrast, Phases 2–4 show little to no correlation between either the timing or magnitude of thermal and gas emissions and therefore represent a decoupled regime. 4.2 Is persistent outgassing balanced by magma flux? Mass balance calculations at mafic open vent volcanoes, globally, suggest that the amount of magma required to sustain observed gas fluxes is generally far greater than that erupted (Edmonds et al., 2022; Kazahaya et al., 1994 ; Laiolo et al., 2022 ; McCormick et al., 2012 ). Constraining the mass balance of magma in terms of total inputs and outputs, and variations over time, is key to relating observed gas emissions to the magmatic and eruptive processes operating at open vent volcanoes. SO 2 emissions and radiant flux can be used to infer the amounts of magma supplied to the shallow magmatic system (input) and erupted (output) respectively (Harris et al. 1999 ; Coppola et al. 2019 ). Here we use TROPOMI-derived SO 2 masses, estimated erupted volumes based on plume heights reported in Volcanic Ash Advisory Bulletin (Darwin VAAC), and lava flow inundation areas estimated from Sentinel-2 satellite multispectral imagery using ArcGIS to calculate the mass balance. The volume of magma required to generate the observed SO 2 emissions is calculated using Eq. 5 : $$V= \frac{f}{c \rho \gamma {\Delta }S} \times {10}^{-9}$$ 5 where \(V\) = magma volume (km 3 ), \(f\) = measured SO 2 flux (kg d −1 ), \(c\) = S to SO 2 conversion constant ( \(c\) = 2), \(\rho\) = magma density (2640 kg m −3 ), \(\gamma\) = vesicularity (expressed as melt fraction i.e. 1 = 0% porosity, 0.7 = 30% porosity), and \({\Delta }S\) = degassed sulphur (ppm \(\times\) 10 −6 ). Density is calculated using the method of Bottinga and Weill ( 1970 ) using bulk rock compositions are from Palfreyman and Cooke ( 1976 ) and McKee ( 1981 ). The values of vesicularity and both initial and degassed melt sulphur contents are unconstrained for recent eruptive products; therefore, the vesicularity term is varied between 0 and 30%, and the total degassed sulphur is approximated as 0.2 ± 0.02 wt%, based on the range of undegassed sulphur contents in melt inclusions from arc magmas (Lloyd et al. 2014 ; Taracsák et al. 2022 ). Manam emitted 1432 ± 48 kt of SO 2 between 6 May 2018 and 31 December 2021 implying degassing of 0.12 to 0.22 km 3 of magma (Eq. 5 ) (TABLE 1). The magma volumes required to yield the cumulative SO 2 emissions for each degassing phase are summarised in (TABLE 1). Effusively Erupted Magma Volume Six lava flows were emplaced between May 2018 and December 2021, identified in satellite imagery. The area inundated by each flow was measured using ArcGIS from MSI, ASTER, and NASA’s Landsat 7 & 8 multispectral imagery (Fig. S3). The volume of each flow was approximated based on the average thickness of the September-October 2018 flow, which was estimated as ~ 3.5 m based on a digital elevation model (DEM) (Fig. S4) created from an Unoccupied Aerial System (DJI Mavic 2 Pro) video of Manam’s North East Valley. A porosity of 18.8% was measured (see supplementary materials) in a lava sample from the distal portion of the 28 June 2019 lava flow and used to convert bulk volume to dense rock equivalent (DRE) (Table S2). Explosively Erupted Magma Volume The volume of magma erupted during explosive events can be estimated using the relationship between eruption plume heights (H, km) and erupted volumes (V, km 3 DRE) presented by Mastin et al. ( 2009 ). This relation is empirical, based on a catalogue of 34 moderate to large explosive eruptions spanning mafic to silicic magma compositions (Eq. 6): \(H=\) 25.9 + 6.64log 10 (V) (6) An uncertainty of approximately one order of magnitude is associated with using Eq. 6 to calculate erupted volume from plume height (L. Mastin, Pers Com.). Here, we use Eq. 6 to calculate erupted volumes at Manam based on explosive eruption column heights reported in RVO bulletins and Darwin VAAC reports (Fig. S5). Magma Flux Balance Over the duration of the study period, more magma was erupted (~ 0.25 km 3 ) than the maximum estimate of magma entering the system (~ 0.22 km 3 ). The eruptions during Phase 1 were responsible for 70% of the erupted magma over the entire time series (FI. 6a). In contrast, the supply of magma was relatively steady as indicated by the consistent cumulative magma input gradient in Phases 1, 3, and 4 (Fig. 6 a). The daily magma net balance shows a steady low magma supply indicated by Manam’s persistent degassing alongside intermittent high magnitude magma outputs (Fig. 6 b). VRP as a Proxy for System Pressure The height of magma columns and lava lakes can vary substantially over days to months, and this is particularly well observed at open vent systems. At open-vent systems like Manam, where the magmatic system is open to the atmosphere, variations in magma column height have been interpreted to reflect changing pressure within the shallow magma reservoirs (Patrick et al., 2015; Moussallam et al., 2016 ; Lev et al., 2019 , Calvari et al., 2011 , Johnson et al., 2018). High system pressure is reflected in an elevated magma column which may rise high enough to be detected as a thermal anomaly and in some cases visible within the crater. Conversely, lower pressure in the magmatic system results in the magma column being too deep to be detected as a thermal anomaly. While eruptions are associated with high pressure conditions, lava effusions typically produce high VRP due to their large radiating surface (Blackett 2017 ) and anomalies linked to explosive eruptions have high VRP due to the radiated heat of large volumes of magma being ejected. Therefore, only non-eruption related thermal anomalies are used here to indicate relative pressure. 4.3 Interpretation – A magma recharge event captured from space? Here we interpret the observed SO 2 and thermal emissions, the coupling of these parameters, the magma flux balance, and periods of high pressure within the context of a magma recharge event to examine the processes and feedbacks at work throughout the study period. The 10 August 2018 eruption ended 11 months of quiescence and passing degassing following the end of 2017 eruptive period on 10 September 2017 (2021a). The absence of thermal anomalies during this interruptive period (Fig. 4 ) suggests a low magma column level and therefore low system pressure. We infer that magma residual from the 2017 eruptive period continued to degas over this period, sustaining the observed SO 2 emissions and driving sluggish convection in the shallow plumbing system (Kazahaya et al. 1994 ; Allard 1997 ; Witter et al. 2004 ; Beckett et al. 2014 ). However, continued degassing of this residual magma without replenishment would drive volatile depletion, cooling, and crystallisation, initially limiting the mobility of melts and subsequently the permeable migration of exsolved fluids (Edmonds et al. 2022a). Formation of a crystal-rich, semi-permeable plug may be partially responsible for the absence of thermal anomalies in the interruptive period (Stix et al. 1997 ; Diller et al. 2006 ; Hall et al. 2015 ; Gaunt et al. 2020 ). In this context, we interpret that the time-averaged Phase 1 SO 2 emissions of ~ 0.62 kt day − 1 represent the supply of volatiles derived from second boiling of residual magma, migrating slowly through the semi-permeable plug (Fig. 7 a). Thermal anomalies were first detected on 7 August 2018 (Fig. 5 d) indicating increasing pressure which was followed by the series of explosive and effusive eruptions in August-September 2018. This increased pressure was likely caused by a volatile-rich magma recharge (Andronico and Corsaro, 2011; Grapenthin et al., 2022; Patrick et al., 2019 a, 2015; Viccaro et al., 2015), that would have began arriving at the shallow storage region several months prior to the onset of eruptive activity based on estimates from similar systems (Cannata et al. 2018 ; Aiuppa et al. 2021 ; Petrone et al. 2022 ). The presence of the semi-permeable plug would have initially inhibited pressure release (Diller et al. 2006 ; Battaglia et al. 2019 ) but continued increasing pressure would eventually exceed plug’s strength (Woitischek et al. 2020 ) causing fracturing and complete failure, resulting in the 10 August 2018 eruption. The 13 August 2018 lava effusion (2021a) via the reopened conduit would have further decreased pressure in the magmatic system. During Phase 1, we calculate that 0.18 km 3 of magma was erupted, compared to 0.03 km 3 required to sustain the observed gas emissions (Fig. 6 , Table 1); this result implies that the erupted magma was extensively degassed prior to eruption. We interpret that the degassed residual magma continued to be removed by the August and September 2018 eruptions (Fig. 7 b), including the 25 August 2018 major explosive eruption and the intermittent effusive activity from 9 September to 1 October 2018 (Fig. 5 ). The eruptive activity in January 2019 removed an estimated 0.11 km 3 (42% of total erupted material during this study) and therefore likely expelled most of the remaining residual magma. The elevated SO 2 emissions of Phase 2 commenced following the 29 March 2019 eruption that removed the final remnants of 2017 residual magma and re-established an open conduit state. Removing residual magma present within the conduit would have reduced the lithostatic load in the upper magmatic system, thereby promoting continued ascent of volatile-rich recharge magma (Calvari et al. 2011 ) and a positive feedback between magma ascent, decompression, and volatile exsolution. During Phase 2 (March – July 2019), we estimate that 0.08 km 3 of magma entered the shallow plumbing system compared to 0.04 km 3 erupted indicating Manam was in state of open vent excess degassing (Rose et al. 2013 ; Edmonds et al. 2022a; Vergniolle and Métrich 2022 ). In-situ observations of magma within South Crater on 22 May 2019 and two high surface temperature retrievals on 20 and 30 May 2019 (Fig. 5 a) suggest at least transient periods of high pressure raising the magma column. However, the decoupled thermal and SO 2 emissions during early Phase 2 (Fig. 7 c), together with the lack of detected anomalies and infrequent eruptions (Fig. 5 d), suggests that the magma level and consequently reservoir pressure, was highly variable. Unlike Phase 1, increased SO 2 emissions are not explicitly linked to eruptions or to thermal anomaly detections; therefore we suggest that surface outgassing involved an enhanced contribution from open system fluxing of volatiles independent of magma ascent, transferred from degassing of recharge magma within the shallow reservoir (Edmonds et al. 2022a). The later stage of Phase 2 (June – July 2019) (Fig. 7 d) displayed increased thermal emissions and eruptive activity compared to earlier in Phase 2. Two surface temperature measurements of > 450°C in South Crater (Fig. 5 a) alongside increased frequency of thermal anomaly detections and recorded eruptions suggest that the magma level remained high in the conduit caused by high reservoir pressure. A marked reduction in daily SO 2 emissions around early June suggests that permeability within the conduit was briefly obstructed (Fig. 5 ). However, the subsequent return to elevated outgassing levels following the 7 June 2019 eruption suggests that explosive activity may have reopened degassing pathways. We propose that the intense degassing and resulting dehydration-driven crystallisation during Phase 2 may have promoted the development of another plug (Applegarth et al., 2013; Couch et al., 2003 ; Gaunt et al., 2020 ; Lipman et al., 1985 ), reducing the permeability of the magma and consequently promoting gas accumulation beneath the plug (Stix et al. 1997 ; Sparks 1997 ; Burgisser et al. 2011 ). Three minor explosive eruptions on 7, 8, and 18 June 2019 likely evidence increasing strain on the plug but it appears that each failed to fully re-open the conduit. We attribute the 28 June 2019 major eruption to the eventual catastrophic failure of this plug, releasing accumulated volatiles and triggering rapid downward-propagating decompression. Explosive activity during this event generated a 15.2 km eruption column, released 58.3 kt of SO 2 , and erupted ~ 0.018 km 3 of magma (7% of total erupted material during this study period). SO 2 emissions during Phase 3 remained above background levels but reduced from ~ 4.72 kt day − 1 to ~ 1.54 kt day − 1 , which may reflect a gradual depletion of volatiles in the shallow magmatic system. SO 2 and thermal emissions remained decoupled in Phase 3. To explain the low frequency of both thermal anomalies and reported eruptions, we infer that the removal of a substantial volume of magma and volatiles would have reduced the pressure in the shallow magmatic system considerably (Anderson et al. 2015 ; Patrick et al. 2020 ; Barrière et al. 2022 ) (Fig. 7 e).The fact that surface SO 2 emissions were sustained at elevated fluxes throughout this period (Fig. 5 e) signals that conduit convection and volatile fluxing remained active despite the reduction in system pressure. During this period, SO 2 emissions imply that 0.02 km 3 of magma was supplied compared to just 0.002 km 3 being erupted (Table 1), indicating that Manam maintained a state of open vent excess degassing. Average daily SO 2 emissions in Phase 4 (~ 0.68 kt day − 1 ) returned to similar levels to Phase 1 (~ 0.62 kt day − 1 ), which can be considered a return to background degassing levels. Reservoir pressure was likely low during November 2019 to May 2021 since there were just 5 thermal anomaly detections (Cluster 2) associated with a series of 6 minor eruptions in July-September 2020. The return to background SO 2 emissions suggests that the initially volatile-rich recharge magma had become relatively depleted. The dominant process responsible for volatile transport would likely have reverted from volatile fluxing to conduit convection, where gas emissions were once again tied to magma transport within the conduit (Fig. 7 f) (Beckett et al. 2014 ; Edmonds et al. 2022a).. This is supported by not only the decrease in SO 2 emissions compared to the previous two phases (Fig. 3 ) but also the reduction in excess degassing with 0.09 km 3 of magma supplied compared to 0.04 km 3 erupted, approximately 75% of which was erupted between August-December 2021. Together, the sparsity of thermal anomalies prior to the onset of Cluster 3 in August 2021 alongside the persistent background SO 2 emissions indicates that emissions were decoupled during Phase 4 (Fig. 5 ). The Cluster 3 thermal emissions suggest that Manam was periodically experiencing high pressure throughout the second half of 2021, raising the magma level close to the surface. These anomalies were associated temporally with the eruptive activity reported throughout this period, yet most were not linked to substantial SO 2 emission peaks as had been the case during Phase 1 (Fig. 5 ). As such, we interpret that the transient periods of eruptive activity during this background level degassing phase are unlikely to reflect a further recharge of volatile-rich magma reaching the shallow plumbing system. Instead, the lack of SO 2 emissions indicates the involvement of comparatively volatile-depleted magma, where the increased pressure to drive explosive events is more likely the result of reduced conduit permeability through cooling and dehydration-driven crystallisation of the residual unerupted magma (Applegarth et al., 2013; Lipman et al., 1985 ). A repeating cycle of partial closing and re-opening of the conduit continued throughout the remainder of 2021, with more substantial and protracted reductions in permeability likely preceding the major eruptions on 20 October 2021 and 22 December 2021 (Hall et al. 2015 ; Battaglia et al. 2019 ). 4.4 Long-term SO 2 Emission Variability The time series of SO 2 emissions at Manam presented in this chapter demonstrates the variability of volcanic volatile emissions over months to years, even at persistently degassing volcanoes. The annual mean daily SO 2 emissions (based on a 1 day residence time) are presented alongside the annual mean daily SO 2 emissions (2005–2015) first reported by Carn et al. ( 2017 ) in Fig. 8 . Carn et al. ( 2017 ) identified a declining trend in annual SO 2 emissions and the inclusion of the data from this study shows that this declining trend continues despite the elevated emissions during 2019 (Fig. 8 ). The 2019 annual daily mean emissions of 2.2 kt day − 1 exceeds substantially the 2015–2021 mean (1.4 kt day − 1 ) and represents a striking departure from the long-term trend, to which emissions in 2020 and 2021 return (Fig. 8 ). This observation demonstrates how open vent volcanoes, such as Manam, can exhibit wide fluctuations in emissions, which are superimposed on decadal trends. If placed within the global SO 2 inventory compiled by Carn et al. ( 2017 ), Manam’s SO 2 emissions would be ranked 31st in Phase 1, 3rd in Phase 3, 10th in Phase 3, and 25th in Phase 4 (Table S3). This variability has also been recognised at other open vent systems, including Bagana (Papua New Guinea; (McCormick Kilbride et al. 2023 ) and consequently highlights the inherent limitations and uncertainties associated with compiling global volcanic volatile inventories, especially where short duration or campaign measurements are relied on. Additionally, this analysis emphasises the need for a multi-parametric approach to interpreting changes in degassing behaviour and eruptive activity at open vent volcanic systems, as the processes responsible for modulating degassing are varied and interpretations potentially ambiguous when derived from emission rates alone. 5. Conclusion We have used a multi-parameter remote sensing approach to investigate the subsurface processes responsible for the period of elevated volcanic activity observed at Manam between August 2018 to December 2021. Using satellite-based measurements of thermal and sulphur dioxide (SO 2 ) emissions, combined with in situ observations of volcanic activity, we quantify the relative inputs and outputs of magma and gas to the shallow conduit and surface—and consequently the varying extent of excess degassing through time. From these timeseries, four distinct phases of volcanic activity are identified between 2018 and 2021. To explain these phases in the context of volcanological processes, we propose that eruptive activity at Manam during the period of observation was driven by the injection and eruption of a volatile-rich recharge magma. In this conceptual model, initial eruptions in August 2018—triggered by positive reservoir pressure changes after a year long period of repose—removed previously degassed residual magma to re-open the conduit and promote efficient fluxing of segregated volatiles through the shallow magmatic system, accounting for the very high SO 2 fluxes (4.72 kt day − 1 ) observed in March-June 2019. A period of lower emissions, both thermal and SO 2 , immediately prior to the major eruption on 28 June 2019 points to reduced permeability and ultimately failure of a conduit plug as a likely trigger mechanism. We suggest plug formation may have been promoted by the extended period of enhanced degassing and resulting dehydration-driven crystallisation in the shallow conduit. The multi-parameter timeseries has allowed the estimation of Manam’s magma budget and resolved the previously unknown fate of the magma supplying Manam’s high SO 2 flux in 2019 (Liu et al. 2020 ). Overall, we calculate that the magma output exceeds the magma input if we consider the entire timeseries examined in this study. However, since this output component is dominated by eruptions from August 2018 to March 2019—which we infer to involve residual degassed magma from 2017—this suggests that the magma budget is balanced over long timescales and that degassed magma is eventually erupted at Manam rather than intruded. In the context of long-term SO 2 degassing trends, the period of enhanced degassing in 2019 is superimposed on a long-term declining trend in emissions at Manam and is therefore not representative of the time-averaged degassing behaviour of this volcano. This substantial temporal variability is not unique to Manam, and has been recognised at other strong open vent emitters (e.g., Bagana, Papua New Guinea; McCormick Kilbride et al. 2023 ). These observations highlight an important limitation to acknowledge when extrapolating short-term or campaign measurements at open vent volcanoes to long-term emissions contributions within global volcanic volatile inventories. Declarations Acknowledgments: The authors wish to thank the Ima Itikarai and Kila Mulina (Rabaul Volcanological Observatory) for providing a sample of the 28 June 2019 lava flow. We acknowledge Keiran Wood (University of Manchester) who piloted the Unoccupied Aerial System which captured footage Manam which was used to create the orthomosiac model of the October 2018 lava flow. Author Contribution: Adam Cotterill, Emma Nicholson & Christopher Kilburn contributed to this studies conception and design. Material preparation, data collection and analysis were performed by Adam Cotterill and Emma Nicholson. Sentinel-5P TROPOMI SO 2 mass retrieval process was provided by Catherine Hayer. The first draft of the manuscript was written by Adam Cotterill and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. This work was supported by the UK Natural Environment Research Council (Grant Number NE/S007229/1). Conflict of interest: The authors declare no competing interests. References Abbott LD (1995) Neogene tectonic reconstruction of the Adelbert-Finisterre-New Britain collision, northern Papua New Guinea. J Southeast Asian Earth Sci 11:33–51. https://doi.org/10.1016/0743-9547(94)00032-A Abbott LD, Silver EA, Galewsky J (1994) Structural evolution of a modern arc-continent collision in Papua New Guinea. Tectonics 13:1007–1034. https://doi.org/10.1029/94TC01623 Aiuppa A (2015) Volcanic-gas monitoring. In: Volcanism and Global Environmental Change. Cambridge University Press, pp 81–96 Aiuppa A, Bitetto M, Delle Donne D, et al (2021) Volcanic CO2 tracks the incubation period of basaltic paroxysms. Sci Adv 7:eabh0191. https://doi.org/10.1126/sciadv.abh0191 Aiuppa A, Bitetto M, Francofonte V, et al (2017) A CO2-gas precursor to the March 2015 Villarrica volcano eruption. Geochem Geophys Geosystems 18:2120–2132. https://doi.org/10.1002/2017GC006892 Aiuppa A, Fischer TP, Plank T, Bani P (2019) CO2 flux emissions from the Earth’s most actively degassing volcanoes, 2005–2015. Sci Rep 9:5442. https://doi.org/10.1038/s41598-019-41901-y Aiuppa A, Moretti R, Federico C, et al (2007) Forecasting Etna eruptions by real-time observation of volcanic gas composition. Geology 35:1115–1118. https://doi.org/10.1130/G24149A.1 Allard P (1997) Endogenous magma degassing and storage at Mount Etna. Geophys Res Lett 24:2219–2222. https://doi.org/10.1029/97GL02101 Allard P, Carbonnelle J, Métrich N, et al (1994) Sulphur output and magma degassing budget of Stromboli volcano. Nature 368:326–330. https://doi.org/10.1038/368326a0 Anderson KR, Poland MP, Johnson JH, Miklius A (2015) Episodic Deflation–Inflation Events at Kīlauea Volcano and Implications for the Shallow Magma System. In: Hawaiian Volcanoes. American Geophysical Union (AGU), pp 229–250 Andres RJ, Kasgnoc AD (1998) A time-averaged inventory of subaerial volcanic sulfur emissions. J Geophys Res Atmospheres 103:25251–25261. https://doi.org/10.1029/98JD02091 Applegarth LJ, Tuffen H, James MR, Pinkerton H (2013a) Degassing-driven crystallisation in basalts. Earth-Sci Rev 116:1–16. https://doi.org/10.1016/j.earscirev.2012.10.007 Applegarth LJane, Tuffen H, James MR, et al (2013b) Direct observations of degassing-induced crystallization in basalts. Geology 41:243–246. https://doi.org/10.1130/G33641.1 Barrière J, d’Oreye N, Smets B, et al (2022) Intra-Crater Eruption Dynamics at Nyiragongo (D.R. Congo), 2002–2021. J Geophys Res Solid Earth 127:e2021JB023858. https://doi.org/10.1029/2021JB023858 Battaglia J, Hidalgo S, Bernard B, et al (2019) Autopsy of an eruptive phase of Tungurahua volcano (Ecuador) through coupling of seismo-acoustic and SO2 recordings with ash characteristics. Earth Planet Sci Lett 511:223–232. https://doi.org/10.1016/j.epsl.2019.01.042 Beckett FM, Burton M, Mader HM, et al (2014) Conduit convection driving persistent degassing at basaltic volcanoes. J Volcanol Geotherm Res 283:19–35. https://doi.org/10.1016/j.jvolgeores.2014.06.006 Beirle S, Hörmann C, Penning de Vries M, et al (2014) Estimating the volcanic emission rate and atmospheric lifetime of SO 2 from space: a case study for Kīlauea volcano, Hawai`i. Atmospheric Chem Phys 14:8309–8322. https://doi.org/10.5194/acp-14-8309-2014 Berk A, Conforti P, Kennett R, et al (2014) MODTRAN® 6: A major upgrade of the MODTRAN® radiative transfer code. In: 2014 6th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS). pp 1–4 Blackett M (2013) Review of the utility of infrared remote sensing for detecting and monitoring volcanic activity with the case study of shortwave infrared data for Lascar Volcano from 2001–2005. Geol Soc Lond Spec Publ 380:107–135. https://doi.org/10.1144/SP380.10 Blackett M (2017) An Overview of Infrared Remote Sensing of Volcanic Activity. J Imaging 3:13. https://doi.org/10.3390/jimaging3020013 Bottinga Y, Weill DF (1970) Densities of liquid silicate systems calculated from partial molar volumes of oxide components. Am J Sci 269:169–182 Burgisser A, Arbaret L, Druitt TH, Giachetti T (2011) Pre-explosive conduit conditions of the 1997 Vulcanian explosions at Soufrière Hills Volcano, Montserrat: II. Overpressure and depth distributions. J Volcanol Geotherm Res 199:193–205. https://doi.org/10.1016/j.jvolgeores.2010.11.014 Calder ES, Cole PD, Dade WB, et al (1999) Mobility of pyroclastic flows and surges at the Soufriere Hills Volcano, Montserrat. Geophys Res Lett 26:537–540. https://doi.org/10.1029/1999GL900051 Calvari S, Spampinato L, Bonaccorso A, et al (2011) Lava effusion — A slow fuse for paroxysms at Stromboli volcano? Earth Planet Sci Lett 301:317–323. https://doi.org/10.1016/j.epsl.2010.11.015 Cannata A, Di Grazia G, Giuffrida M, et al (2018) Space-Time Evolution of Magma Storage and Transfer at Mt. Etna Volcano (Italy): The 2015–2016 Reawakening of Voragine Crater. Geochem Geophys Geosystems 19:471–495. https://doi.org/10.1002/2017GC007296 Caricchi L, Sheldrake TE, Blundy J (2018) Modulation of magmatic processes by CO2 flushing. Earth Planet Sci Lett 491:160–171. https://doi.org/10.1016/j.epsl.2018.03.042 Carn SA, Fioletov VE, McLinden CA, et al (2017) A decade of global volcanic SO2 emissions measured from space. Sci Rep 7:44095. https://doi.org/10.1038/srep44095 Connell J, Lutkehaus N (2016) Another Manam? The forced migration of the population of Manam Island, Papua New Guinea, due to volcanic eruptions 2004–2005. International Organisation for Migration Coppola D, Laiolo M, Cigolini C, et al (2016) Enhanced volcanic hot-spot detection using MODIS IR data: results from the MIROVA system. Geol Soc Lond Spec Publ 426:181–205. https://doi.org/10.1144/SP426.5 Coppola D, Laiolo M, Massimetti F, Cigolini C (2019) Monitoring endogenous growth of open-vent volcanoes by balancing thermal and SO2 emissions data derived from space. Sci Rep 9:9394. https://doi.org/10.1038/s41598-019-45753-4 Coppola D, Laiolo M, Piscopo D, Cigolini C (2013) Rheological control on the radiant density of active lava flows and domes. J Volcanol Geotherm Res 249:39–48. https://doi.org/10.1016/j.jvolgeores.2012.09.005 Coppola D, Valade S, Masias P, et al (2022) Shallow magma convection evidenced by excess degassing and thermal radiation during the dome-forming Sabancaya eruption (2012–2020). Bull Volcanol 84:16. https://doi.org/10.1007/s00445-022-01523-1 Couch S, Sparks RSJ, Carroll MR (2003) The Kinetics of Degassing-Induced Crystallization at Soufrière Hills Volcano, Montserrat. J Petrol 44:1477–1502. https://doi.org/10.1093/petrology/44.8.1477 Diller K, Clarke AB, Voight B, Neri A (2006) Mechanisms of conduit plug formation: Implications for vulcanian explosions. Geophys Res Lett 33:. https://doi.org/10.1029/2006GL027391 Edmonds, Liu EJ, Cashman KV (2022a) Open-vent volcanoes fuelled by depth-integrated magma degassing. Bull Volcanol 84:28. https://doi.org/10.1007/s00445-021-01522-8 Edmonds M, Mason E, Hogg O (2022b) Volcanic Outgassing of Volatile Trace Metals. Annu Rev Earth Planet Sci 50:79–98. https://doi.org/10.1146/annurev-earth-070921-062047 Edmonds M, Oppenheimer C, Pyle DM, et al (2003) SO2 emissions from Soufrière Hills Volcano and their relationship to conduit permeability, hydrothermal interaction and degassing regime. J Volcanol Geotherm Res 124:23–43. https://doi.org/10.1016/S0377-0273(03)00041-6 Fioletov VE, McLinden CA, Krotkov N, Li C (2015) Lifetimes and emissions of SO2 from point sources estimated from OMI. Geophys Res Lett 42:1969–1976. https://doi.org/10.1002/2015GL063148 Flower VJB, Carn SA (2015) Characterising volcanic cycles at Soufriere Hills Volcano, Montserrat: Time series analysis of multi-parameter satellite data. J Volcanol Geotherm Res 304:82–93. https://doi.org/10.1016/j.jvolgeores.2015.07.035 Francis PW, Rothery DA (1987) Using the Landsat Thematic Mapper to detect and monitor active volcanoes: An example from Lascar volcano, northern Chile. Geology 15:614–617. https://doi.org/10.1130/0091-7613(1987)152.0.CO;2 Gaunt HE, Burgisser A, Mothes PA, et al (2020) Triggering of the powerful 14 July 2013 Vulcanian explosion at Tungurahua Volcano, Ecuador. J Volcanol Geotherm Res 392:106762. https://doi.org/10.1016/j.jvolgeores.2019.106762 Girina OA (2013) Chronology of Bezymianny Volcano activity, 1956–2010. J Volcanol Geotherm Res 263:22–41. https://doi.org/10.1016/j.jvolgeores.2013.05.002 Gray DM, Burton-Johnson A, Fretwell PT (2019) Evidence for a lava lake on Mt. Michael volcano, Saunders Island (South Sandwich Islands) from Landsat, Sentinel-2 and ASTER satellite imagery. J Volcanol Geotherm Res 379:60–71. https://doi.org/10.1016/j.jvolgeores.2019.05.002 Hall ML, Steele AL, Bernard B, et al (2015) Sequential plug formation, disintegration by Vulcanian explosions, and the generation of granular Pyroclastic Density Currents at Tungurahua volcano (2013–2014), Ecuador. J Volcanol Geotherm Res 306:90–103. https://doi.org/10.1016/j.jvolgeores.2015.09.009 Harris A (2013) Thermal Remote Sensing of Active Volcanoes: A User’s Manual. Cambridge University Press Harris AJL, Flynn LP, Rothery DA, et al (1999) Mass flux measurements at active lava lakes: Implications for magma recycling. J Geophys Res Solid Earth 104:7117–7136. https://doi.org/10.1029/98JB02731 Holm RJ, Richards SW (2013) A re-evaluation of arc–continent collision and along-arc variation in the Bismarck Sea region, Papua New Guinea. Aust J Earth Sci 60:605–619. https://doi.org/10.1080/08120099.2013.824505 James MR, Varley N (2012) Identification of structural controls in an active lava dome with high resolution DEMs: Volcán de Colima, Mexico. Geophys Res Lett 39:. https://doi.org/10.1029/2012GL054245 Johnson ER, Wallace PJ, Cashman KV, Delgado Granados H (2010) Degassing of volatiles (H2O, CO2, S, Cl) during ascent, crystallization, and eruption at mafic monogenetic volcanoes in central Mexico. J Volcanol Geotherm Res 197:225–238. https://doi.org/10.1016/j.jvolgeores.2010.02.017 Johnson RW (2013) Fire Mountains of the Islands: A History of Volcanic Eruptions and Disaster Management in Papua New Guinea and the Solomon Islands. ANU E Press Kazahaya K, Shinohara H, Saito G (1994) Excessive degassing of Izu-Oshima volcano: magma convection in a conduit. Bull Volcanol 56:207–216. https://doi.org/10.1007/BF00279605 Laiolo M, Delle Donne D, Coppola D, et al (2022) Shallow magma dynamics at open-vent volcanoes tracked by coupled thermal and SO2 observations. Earth Planet Sci Lett 594:117726. https://doi.org/10.1016/j.epsl.2022.117726 Laiolo M, Massimetti F, Cigolini C, et al (2018) Long-term eruptive trends from space-based thermal and SO2 emissions: a comparative analysis of Stromboli, Batu Tara and Tinakula volcanoes. Bull Volcanol 80:68. https://doi.org/10.1007/s00445-018-1242-0 Lev E, Ruprecht P, Oppenheimer C, et al (2019) A global synthesis of lava lake dynamics. J Volcanol Geotherm Res 381:16–31. https://doi.org/10.1016/j.jvolgeores.2019.04.010 Lipman PW, Banks NG, Rhodes JM (1985) Degassing-induced crystallization of basaltic magma and effects on lava rheology. Nature 317:604–607. https://doi.org/10.1038/317604a0 Liu EJ, Aiuppa A, Alan A, et al (2020) Aerial strategies advance volcanic gas measurements at inaccessible, strongly degassing volcanoes. Sci Adv 6:eabb9103. https://doi.org/10.1126/sciadv.abb9103 Lloyd AS, Ruprecht P, Hauri EH, et al (2014) NanoSIMS results from olivine-hosted melt embayments: Magma ascent rate during explosive basaltic eruptions. J Volcanol Geotherm Res 283:1–18. https://doi.org/10.1016/j.jvolgeores.2014.06.002 Mastin LG, Guffanti M, Servranckx R, et al (2009) A multidisciplinary effort to assign realistic source parameters to models of volcanic ash-cloud transport and dispersion during eruptions. J Volcanol Geotherm Res 186:10–21. https://doi.org/10.1016/j.jvolgeores.2009.01.008 McCormick BT, Edmonds M, Mather TA, Carn SA (2012) First synoptic analysis of volcanic degassing in Papua New Guinea. Geochem Geophys Geosystems 13:. https://doi.org/10.1029/2011GC003945 McCormick Kilbride BT, Mulina K, Wadge G, et al (2019) Multi-year Satellite Observations of Sulfur Dioxide Gas Emissions and Lava Extrusion at Bagana Volcano, Papua New Guinea. Front Earth Sci 7:. https://doi.org/10.3389/feart.2019.00009 McCormick Kilbride BT, Nicholson EJ, Wood KT, et al (2023) Temporal Variability in Gas Emissions at Bagana Volcano Revealed by Aerial, Ground, and Satellite Observations. Geochem Geophys Geosystems 24:e2022GC010786. https://doi.org/10.1029/2022GC010786 McKee CO (1981) Geomorphology, geology and petrology of Manam volcano. Cooke-Ravian Vol Volcanol Pap 23–38 McKee K, Smith CM, Reath K, et al (2021) Evaluating the state-of-the-art in remote volcanic eruption characterization Part II: Ulawun volcano, Papua New Guinea. J Volcanol Geotherm Res 420:107381. https://doi.org/10.1016/j.jvolgeores.2021.107381 Meerdink SK, Hook SJ, Roberts DA, Abbott EA (2019) The ECOSTRESS spectral library version 1.0. Remote Sens Environ 230:111196. https://doi.org/10.1016/j.rse.2019.05.015 Moussallam Y, Bani P, Curtis A, et al (2016) Sustaining persistent lava lakes: Observations from high-resolution gas measurements at Villarrica volcano, Chile. Earth Planet Sci Lett 454:237–247. https://doi.org/10.1016/j.epsl.2016.09.012 Mueller SB, Varley NR, Kueppers U, et al (2013) Quantification of magma ascent rate through rockfall monitoring at the growing/collapsing lava dome of Volcán de Colima, Mexico. Solid Earth 4:201–213. https://doi.org/10.5194/se-4-201-2013 Palfreyman W, Cooke R (1976) Eruptive history of Manam volcano, Papua New Guinea. Volcanism Australas 201–210 Palma JL, Blake S, Calder ES (2011) Constraints on the rates of degassing and convection in basaltic open-vent volcanoes. Geochem Geophys Geosystems 12:. https://doi.org/10.1029/2011GC003715 Patrick M, Swanson D, Orr T (2019) A review of controls on lava lake level: insights from Halema‘uma‘u Crater, Kīlauea Volcano. Bull Volcanol 81:13. https://doi.org/10.1007/s00445-019-1268-y Patrick MR, Houghton BF, Anderson KR, et al (2020) The cascading origin of the 2018 Kīlauea eruption and implications for future forecasting. Nat Commun 11:5646. https://doi.org/10.1038/s41467-020-19190-1 Petrone CM, Mollo S, Gertisser R, et al (2022) Magma recharge and mush rejuvenation drive paroxysmal activity at Stromboli volcano. Nat Commun 13:7717. https://doi.org/10.1038/s41467-022-35405-z Planck M (1901) On the Law of Distribution of Energy in the Normal Spectrum. Ann Phys 4:1–11 Plank S, Walter T, Martinis S, Cesca S (2020) Multi-sensor satellite imagery analysis of the growth and collapse of a littoral lava dome during the 2018/19 eruption of Kadovar Volcano, Papua New Guinea. 22:1423 Plank S, Walter TR, Martinis S, Cesca S (2019) Growth and collapse of a littoral lava dome during the 2018/19 eruption of Kadovar Volcano, Papua New Guinea, analyzed by multi-sensor satellite imagery. J Volcanol Geotherm Res 388:106704. https://doi.org/10.1016/j.jvolgeores.2019.106704 Pyle DM, Mather TA, Biggs J (2013) Remote sensing of volcanoes and volcanic processes: integrating observation and modelling – introduction. Geol Soc Lond Spec Publ 380:1–13. https://doi.org/10.1144/SP380.14 Robin C, Camus G, Gourgaud A (1991) Eruptive and magmatic cycles at Fuego de Colima volcano (Mexico). J Volcanol Geotherm Res 45:209–225. https://doi.org/10.1016/0377-0273(91)90060-D Rose WI, Palma J, Delgado Granados H, Varley N (2013) Open-vent volcanism and related hazards: Overview. Spec Pap Geol Soc Am 498:vii–xiii. https://doi.org/10.1130/2013.2498(00) Rothery DA, Francis PW, Wood CA (1988) Volcano monitoring using short wavelength infrared data from satellites. J Geophys Res Solid Earth 93:7993–8008. https://doi.org/10.1029/JB093iB07p07993 Shevchenko AV, Dvigalo VN, Walter TR, et al (2020) The rebirth and evolution of Bezymianny volcano, Kamchatka after the 1956 sector collapse. Commun Earth Environ 1:1–15. https://doi.org/10.1038/s43247-020-00014-5 Shinohara H (2008) Excess degassing from volcanoes and its role on eruptive and intrusive activity. Rev Geophys 46: Shinohara H (2005) A new technique to estimate volcanic gas composition: plume measurements with a portable multi-sensor system. J Volcanol Geotherm Res 143:319–333. https://doi.org/10.1016/j.jvolgeores.2004.12.004 Sparks RSJ (1997) Causes and consequences of pressurisation in lava dome eruptions. Earth Planet Sci Lett 150:177–189. https://doi.org/10.1016/S0012-821X(97)00109-X Sparks RSJ, Biggs J, Neuberg JW (2012) Monitoring volcanoes. Science 335:1310–1311 Stefan J (1879) Uber die Beziehung zwischen der Warmestrahlung und der Temperatur, Sitzungsberichte der mathematisch-naturwissenschaftlichen Classe der kaiserlichen. Akad Wiss 79:S-391 Stix J, Torres RC, Narváez M L, et al (1997) A model of vulcanian eruptions at Galeras volcano, Colombia. J Volcanol Geotherm Res 77:285–303. https://doi.org/10.1016/S0377-0273(96)00100-X Taracsák Z, Longpré M-A, Tartèse R, et al (2022) Highly Oxidising Conditions in Volatile-Rich El Hierro Magmas: Implications for Ocean Island Magmatism. J Petrol 63:egac011. https://doi.org/10.1093/petrology/egac011 Theys N, Hedelt P, De Smedt I, et al (2019) Global monitoring of volcanic SO2 degassing with unprecedented resolution from TROPOMI onboard Sentinel-5 Precursor. Sci Rep 9:2643. https://doi.org/10.1038/s41598-019-39279-y Vergniolle S, Métrich N (2021) Open-vent volcanoes: a preface to the special issue. Bull Volcanol 83:29, s00445-021-01454–3. https://doi.org/10.1007/s00445-021-01454-3 Vergniolle S, Métrich N (2022) An interpretative view of open-vent volcanoes. Bull Volcanol 84:83. https://doi.org/10.1007/s00445-022-01581-5 Witter JB, Kress VC, Delmelle P, Stix J (2004) Volatile degassing, petrology, and magma dynamics of the Villarrica Lava Lake, Southern Chile. J Volcanol Geotherm Res 134:303–337. https://doi.org/10.1016/j.jvolgeores.2004.03.002 Woitischek J, Edmonds M, Woods AW (2020) The control of magma crystallinity on the fluctuations in gas composition at open vent basaltic volcanoes. Sci Rep 10:14862. https://doi.org/10.1038/s41598-020-71667-7 Wood M, Foale S, Gabriel J (2019) Anticipating Ulawun Volcano in New Britain, Papua New Guinea. Anthropol Forum 1–12. https://doi.org/10.1080/00664677.2019.1647831 Woodhead J, Hergt J, Sandiford M, Johnson W (2010) The big crunch: Physical and chemical expressions of arc/continent collision in the Western Bismarck arc. J Volcanol Geotherm Res 190:11–24. https://doi.org/10.1016/j.jvolgeores.2009.03.003 Wooster M, Kaneko T (1998) Satellite thermal analyses of lava dome effusion rates at Unzen Volcano, Japan. J Geophys Res Solid Earth 103:20935–20947 Wooster MJ, Zhukov B, Oertel D (2003) Fire radiative energy for quantitative study of biomass burning: derivation from the BIRD experimental satellite and comparison to MODIS fire products. Remote Sens Environ 86:83–107. https://doi.org/10.1016/S0034-4257(03)00070-1 Wright R (2016) MODVOLC: 14 years of autonomous observations of effusive volcanism from space. Geol Soc Lond Spec Publ 426:23–53. https://doi.org/10.1144/SP426.12 Wright R, Blackett M, Hill‐Butler C (2015) Some observations regarding the thermal flux from Earth’s erupting volcanoes for the period of 2000 to 2014. Geophys Res Lett 42:282–289. https://doi.org/10.1002/2014GL061997 Wright R, Flynn LP, Garbeil H, et al (2004) MODVOLC: near-real-time thermal monitoring of global volcanism. J Volcanol Geotherm Res 135:29–49. https://doi.org/10.1016/j.jvolgeores.2003.12.008 Young S, Voight B, Duffell H (2003) Magma extrusion dynamics revealed by high-frequency gas monitoring at Soufriere Hills Volcano, Montserrat. Geol Soc Lond Spec Publ 213:219–230 (2021a) Global Volcanism Program | Manam. In: Smithson. Inst. Glob. Volcanism Program. https://volcano.si.edu/volcano.cfm?vn=251020. Accessed 9 Feb 2020 (2021b) Global Volcanism Program | Kadovar. https://volcano.si.edu/volcano.cfm?vn=251002. Accessed 8 May 2021 (2021c) Global Volcanism Program | Ulawun. In: Smithson. Inst. Glob. Volcanism Program. https://volcano.si.edu/volcano.cfm?vn=252120. Accessed 8 May 2021 (2022) MIROVA. https://www.mirovaweb.it/?action=volcanoDetails&volcano_id=251020#lrp. Accessed 27 May 2021 Tables Tables 1 and 2 are available in the Supplementary Files section. Supplementary Files Table1.pdf Table 1 Summary of SO 2 emissions by phase and the calculated magma volume required to supply observed SO 2 emissions Table2.pdf Table 2 Volume of erupted magma by phase. Explosive erupted volumes calculated from eruption column heights using Equation 6 (Mastin et al., 2009). Eruption column heights recorded by Darwin VAAC and the Rabaul Volcanological Observatory observer on Manam. Effusive eruptions are estimated from satellite observation of lava flows and a representative thickness calculated from an orthomosaic from the September-October 2018 lava flow against a older digital elevation model (see supplementary material) and using a measured porosity (18.8%) from the September-October 2018 lava flow. SupplementaryMaterials.docx Cite Share Download PDF Status: Published Journal Publication published 07 Oct, 2024 Read the published version in Bulletin of Volcanology → Version 1 posted Reviewers agreed at journal 04 Mar, 2024 Reviewers invited by journal 27 Feb, 2024 First submitted to journal 26 Jan, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3903120","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":275425001,"identity":"3708b375-8cf4-4945-9fe5-0becd8a90348","order_by":0,"name":"Adam Cotterill","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABC0lEQVRIie2RsWrDMBCGLwSiReD1Skv1ChIFT6Z5lYOAvHgoFEqhIfWULO4DlL5Eu3Q2GOzFD6Bu6VBPGTx1rmXXtIOSdsygbzoJPv3/IQCP5yihnzHfXgEXLAXY9ufZQQW5VUjCmcry8fIfCnRKBIYOK8Fm8fEMy2g1Zw/vNoVPHneqJbgUgJpcCtZNaKDUyHkle4WdJhdIsFAp6twZYyg0k7RAjhqGlKck7IpNCTBOXYYw8eegiGZQ4K22yv1eRZpkTJl9K4ZbpegUdzFV724Mlfokq/tiyFWmr5Fkpda8ca5/XsWvpl1GAduU07a9jeaCFS/dcCcCpqVzfcuvx3AsvP8jPR6Px/M3X7BPWISOPM+rAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0009-0001-1315-3087","institution":"University College London","correspondingAuthor":true,"prefix":"","firstName":"Adam","middleName":"","lastName":"Cotterill","suffix":""},{"id":275425002,"identity":"7fd1009d-09db-49ea-8d13-868b0a1d2da8","order_by":1,"name":"Emma Nicholson","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Emma","middleName":"","lastName":"Nicholson","suffix":""},{"id":275425003,"identity":"7da64832-24bf-4d86-916b-83810188280f","order_by":2,"name":"Christopher Kilburn","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Christopher","middleName":"","lastName":"Kilburn","suffix":""},{"id":275425004,"identity":"fa464e86-ebbd-4076-aa9b-19e84b9bb50b","order_by":3,"name":"Catherine Hayer","email":"","orcid":"","institution":"EUMETSAT","correspondingAuthor":false,"prefix":"","firstName":"Catherine","middleName":"","lastName":"Hayer","suffix":""}],"badges":[],"createdAt":"2024-01-27 13:43:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3903120/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3903120/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00445-024-01772-2","type":"published","date":"2024-10-07T15:58:05+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":51969540,"identity":"28d66e3f-c72e-4d95-9268-ada53747f30c","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":533717,"visible":true,"origin":"","legend":"\u003cp\u003eLeft – Map of Papua New Guinea with the locations of Manam (M), Kadovar (K) and Ulawun (U). Right - True colour image from 17th August 2020 Sentinel-2 overpass with key features of Manam Island used for Sentinel-2 imagery processing annotated.\u003c/p\u003e","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/2fab5e87fe6673c29522b2bf.jpg"},{"id":51969539,"identity":"70b3a620-b20f-4ca1-9873-29f212e638ee","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2518847,"visible":true,"origin":"","legend":"\u003cp\u003eTROPOMI SO2 Vertical Column Density (VCD) interpolated at 15 km altitude over Manam on 28th June 2019 showing the SO2 plume from the major eruption of Manam that day. This total emitted SO2 mass was 58.3 kt on 28th June 2019. The total emitted SO2 mass is calculated by summing all pixels within the view extent. N.B. this retrieval has a pixel resolution of 3.5 x 7 km as it was taken prior to the resolution improvement to 3.5 x 5.5 km on 6 August 2019.\u003c/p\u003e","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/3d96f34576901bd3328505ed.jpg"},{"id":51969535,"identity":"0d30388c-d752-48a1-9290-77032c65f4c2","added_by":"auto","created_at":"2024-03-04 18:41:44","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":297066,"visible":true,"origin":"","legend":"\u003cp\u003eSolid lines show cumulative SO2 under 1 day (blue), 2 day (red), and 3 day (green) residence time regimes. Each regime is divided by the 4 identified emission phases. A polynomial was fitted for each regime and phase (dashed lines) and the gradient of each line indicates the average daily flux.\u003c/p\u003e","description":"","filename":"Figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/75e365d0ff8c4c52356310a2.jpg"},{"id":51971041,"identity":"ac9b9abf-e312-4ee2-a451-d4467cee0e09","added_by":"auto","created_at":"2024-03-04 18:49:44","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1341274,"visible":true,"origin":"","legend":"\u003cp\u003eA) Timeseries of MODVOLC detected thermal anomalies from 2015 -2021 at Manam and observed activity. Horizontal coloured dashed lines correspond to Volcanic Radiative Power intensities. B) ASTER infrared imagery used to visually identify the location of anomalies. Black markers indicate an anomaly present in the key region represented by the row marker on the y-axis. Yellow highlighted markers represent an Aster anomaly on the same day as a MODVOLC detection.\u003c/p\u003e","description":"","filename":"Figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/2d038e85be6449f76c3e4c40.jpg"},{"id":51969536,"identity":"3fe5c903-04b5-47a4-93e8-268322968e1d","added_by":"auto","created_at":"2024-03-04 18:41:44","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1447625,"visible":true,"origin":"","legend":"\u003cp\u003eCombined timeseries of Maximum Pixel Integrated Temperature (A), MSI Cloud Cover (B), Mean Pixel Integrated Temperature (C), MODVOLC thermal anomaly detection (D), SO2 emissions (E) and activity reported by Rabaul Volcanological Observatory and Global Volcanism Progam (F).\u003c/p\u003e","description":"","filename":"Figure5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/364a254537b23f52759c417e.jpg"},{"id":51969543,"identity":"bc839e33-790c-473b-983f-dd4ad07e3e07","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":782404,"visible":true,"origin":"","legend":"\u003cp\u003eA) Cumulative magma output (i.e. effusively and explosively erupted magma) (red) and cumulative magma input (magma reaching the exsolution level for SO2 at depth) (blue). The minimum and maximum magma input values are based on varying the assumed sulphur content (0.2 ± 0.02 wt %) and vesicularity (0-30%) of the melt used in the petrological method to calculate magma volumes from SO2 flux (Devine et al., 1984). B) The daily net magma balance calculated as magma output volume subtracted from magma input volume.\u003c/p\u003e","description":"","filename":"Figure6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/6509e1e9dc9fbffeff4d20e5.jpg"},{"id":51969544,"identity":"80101f3a-fbd9-431c-bfa7-a119159811b7","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":814558,"visible":true,"origin":"","legend":"\u003cp\u003eConceptual model of Manam's shallow plumbing system and processes responsible for the observed activity, thermal anomalies and SO2 emissions. “Coupled” and “Decoupled” indicates the relationship between thermal and SO2 emissions during each phase; see main text for detail.\u003c/p\u003e","description":"","filename":"Figure7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/e3c5ad97a7fd1f493c70a0e8.jpg"},{"id":51969541,"identity":"f8a9a307-f337-43b6-a15c-60356ee3bd39","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":727070,"visible":true,"origin":"","legend":"\u003cp\u003eManam annual daily SO2 emissions (kt/d) measured using OMI (Ozone Monitoring Instrument) (green) (Carn et al. 2017), TROPOMI derived annual daily SO2 emissions based a 1 day residence time (blue) (this study), mean daily SO2 trend calculated by fitting a first order polynomial to the timeseries (orange dotted line) and mean annual emissions 2015-2012 – 1.36 kt/d (black dotted line).\u003c/p\u003e","description":"","filename":"Figure8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/85348237b656a39624292c96.jpg"},{"id":66597435,"identity":"dbfcd92f-205e-4053-a137-84301e220c20","added_by":"auto","created_at":"2024-10-14 16:10:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9163388,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/1c12b96d-da1b-4411-96b8-f5cb034f4223.pdf"},{"id":51969538,"identity":"fef5abe4-f2f3-472b-ad01-20115c6344fa","added_by":"auto","created_at":"2024-03-04 18:41:44","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":48948,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e Summary of SO\u003csub\u003e2\u003c/sub\u003e emissions by phase and the calculated magma volume required to supply observed SO\u003csub\u003e2\u003c/sub\u003e emissions\u003c/p\u003e","description":"","filename":"Table1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/1aa715ad2841c8b3640b0848.pdf"},{"id":51969546,"identity":"b3e7d137-0aa3-418e-a982-b4260b5a43ec","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":116519,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 2 \u003c/strong\u003eVolume of erupted magma by phase. Explosive erupted volumes calculated from eruption column heights using Equation 6 (Mastin et al., 2009). Eruption column heights recorded by Darwin VAAC and the Rabaul Volcanological Observatory observer on Manam. Effusive eruptions are estimated from satellite observation of lava flows and a representative thickness calculated from an orthomosaic from the September-October 2018 lava flow against a older digital elevation model (see supplementary material) and using a measured porosity (18.8%) from the September-October 2018 lava flow.\u003c/p\u003e","description":"","filename":"Table2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/e5116fc0b6cef6ea69e40757.pdf"},{"id":51969542,"identity":"3f622568-c676-41cd-8eca-4db9cc6eb533","added_by":"auto","created_at":"2024-03-04 18:41:45","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2654509,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-3903120/v1/85a9250319210ae85e0500bc.docx"}],"financialInterests":"","formattedTitle":"\u003cp\u003e\u003cstrong\u003eMagma recharge at Manam volcano, Papua New Guinea, identified through thermal and SO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e satellite remote sensing of open vent emissions\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eOpen vent volcanism is sustained by the ascent and degassing of magma at shallow depths (Kazahaya et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Harris et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Shinohara \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Johnson et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Palma et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), with variable contributions from both conduit convection and deep-derived segregated fluids that transfer both heat and volatiles to shallow reservoirs (Caricchi et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Edmonds et al., 2022). Open vent systems exhibit a spectrum of eruptive styles: silica-rich magmas can form lava domes that may trigger vulcanian eruptions if collapse occurs (Stefan \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e1879\u003c/span\u003e; Robin et al. \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Wooster and Kaneko \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Calder et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Young et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; James and Varley \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Mueller et al. \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Girina \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Flower and Carn \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Shevchenko et al. \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), while more mafic magmas support open conduit conditions allowing increased mobility of both melt and volatiles and, in rare cases, maintain lava lakes over years to decades (Moussallam et al. \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Patrick et al. \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Gray et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Lev et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eA common characteristic of open vent volcanoes is persistent degassing (Rose et al. \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Vergniolle and M\u0026eacute;trich \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), where the volume of magma required to supply the observed volatile flux exceeds that erupted; this is referred to as the \u0026ldquo;excess degassing phenomenon\u0026rdquo; (Kazahaya et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Shinohara \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). The fate of this unerupted degassed magma is often explained by conduit convection (Kazahaya et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Beckett et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Coppola et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) or by intrusion and endogenous crustal growth (Coppola et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In both scenarios it remains unclear where this degassed magma accumulates in the crust. Typically, SO\u003csub\u003e2\u003c/sub\u003e flux is used to determine the magma supply rate (Allard et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Andres and Kasgnoc \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Shinohara \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) and the magma output is often approximated using thermal emissions (Wooster and Kaneko \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Laiolo et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Coppola et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The balance between magma input and output, and changes in this budget through time, has been shown to be highly indicative of pressure and fluid dynamic perturbations within the shallow magma storage region, which can disturb open conduit processes and may ultimately lead to eruptions.\u003c/p\u003e \u003cp\u003eManam (-4.078, 145.038) is a frequently erupting mafic open vent stratovolcano located\u0026thinsp;~\u0026thinsp;19 km off the northeast coast of mainland Papua New Guinea (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) (Palfreyman and Cooke \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1976\u003c/span\u003e, 2021a). Manam is situated within the Western Bismarck Arc where the arc-continent subduction has ceased and melting of the remnant hanging slab is the source of volcanism (Abbott et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Abbott \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Woodhead et al. \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Holm and Richards \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) and, historically, it has erupted basaltic to basaltic andesite magma compositions. The volcano has two active craters that exhibit different activity. Main Crater hosts degassing from a broad fumarole field and is the vent from where most lava effusions originate. South Crater emits a near-constant dense gas plume and explosive eruptions tend to be generated from this crater. The top surface of the magma column was observed at shallow depths within South Crater on 22 May 2019 (Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) but the temporal persistence of this state is currently unknown aside from occasional reports of incandescence (2021a).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eManam is a top emitter of volcanic volatiles in a global context and is therefore an important case study to explore temporal variability in emissions and their relationship to subsurface processes. Satellite measurements of SO\u003csub\u003e2\u003c/sub\u003e using NASA\u0026rsquo;s Ozone Monitoring Instrument (OMI) between 2005\u0026ndash;2015 indicate an average SO\u003csub\u003e2\u003c/sub\u003e flux of 1480\u0026thinsp;\u0026plusmn;\u0026thinsp;750 tonnes day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e placing Manam as the 11th strongest emission source globally. (Carn et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Similarly, Manam\u0026rsquo;s CO\u003csub\u003e2\u003c/sub\u003e flux has been estimated at 2760\u0026thinsp;\u0026plusmn;\u0026thinsp;1570 tonnes day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e making it the 10th highest emitter of volcanic CO\u003csub\u003e2\u003c/sub\u003e (Aiuppa et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Direct sampling of Manam\u0026rsquo;s plume using a Unoccupied Aerial Vehicle (UAV) mounted MultiGAS (Shinohara \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Aiuppa et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) in May 2019 revealed an elevated average SO\u003csub\u003e2\u003c/sub\u003e flux of 5150 \u0026plusmn; [336/733] and CO\u003csub\u003e2\u003c/sub\u003e flux of 3760 \u0026plusmn; [313/595] (Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). These campaign-style measurements ranked Manam, albeit transiently, at 2nd and 5-7th respectively for volcanic SO\u003csub\u003e2\u003c/sub\u003e and CO\u003csub\u003e2\u003c/sub\u003e production globally.\u003c/p\u003e \u003cp\u003eManam generated 29 major and 139 minor eruptions between 2000 and 2021 (Palfreyman and Cooke \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1976\u003c/span\u003e, 2021a). Over the same interval, frequent clusters of thermal anomaly detections were identified from satellite multispectral imagery. Following large-scale evacuations associated with the major eruptions in 2004\u0026ndash;2005 (Johnson \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Connell and Lutkehaus \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), approximately 4000 residents had returned to the island by August 2021 and experience ongoing impacts to agriculture, settlement infrastructure, and water supplies due to persistent volcanic activity (J Sukua, Pers Com., 2021).\u003c/p\u003e \u003cp\u003eA new phase of eruptive activity began in August 2018 after 11 months of quiescence. A series of 23 eruptions then took place until March 2019, including 5 major eruptions (i.e. eruption columns\u0026thinsp;\u0026gt;\u0026thinsp;10 km) on 25 August 2018, 8 December 2018, 7 January 2019, 11 January 2019, and 23 January 2019. Lava effusions occurred between 27 September and 1 October 2018 and on 8 January 2019, which reached within 500 and 400 m of the coastline respectively. The largest eruption in recent years occurred on 28 June 2019, generating an eruption column that rose to 15.2 km asl, pyroclastic density currents, and a lava flow reaching within 700 m of the coastline. During this eruption, 3775 residents evacuated temporarily and 455 homes and agricultural gardens are reported to have been destroyed or damaged (2021a). Ground-based volcano monitoring is challenging due to the island setting, dense vegetation, steep topography and tropical climate, and therefore key gaps remain in observational capability despite the high level of volcanic risk. Improving understanding of the relationship between satellite remote sensing observations and subsurface volcanic processes are therefore critical to augmenting future volcano monitoring at this, and other, open vent volcanoes.\u003c/p\u003e \u003cp\u003eSatellite remote sensing provides regular measurements and near-global coverage of volcanic SO\u003csub\u003e2\u003c/sub\u003e (Theys et al. \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) and thermal emissions (Wright et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Coppola et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) enabling monitoring of remote or inaccessible volcanoes, such as Manam. Here, we present the first multi-year time series from 2018 to 2021 of (a) SO\u003csub\u003e2\u003c/sub\u003e emissions derived from the European Space Agency (ESA) Sentinel-5P Tropospheric Monitoring Instrument (TROPOMI), (b) thermal anomaly detections from the National Aeronautic and Space Administration (NASA) Moderate Resolution Imaging Spectrometer (MODIS) instrument processed by the MODVOLC algorithm, and (c) surface temperature measurements of Main and South Crater using ESA\u0026rsquo;s Sentinel-2 MultiSpectral Instrument (MSI). We use this multi-parameter time series to interpret magmatic processes influencing transitions in open vent behaviour at Manam, quantify the magnitude of excess degassing and consequently the volume of unerupted degassed magma; and evaluate the time-varying contribution of Manam to global volcanic volatile emission inventories.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Sulphur Dioxide (SO\u003csub\u003e2\u003c/sub\u003e) Emissions\u003c/h2\u003e \u003cp\u003eWe quantify SO\u003csub\u003e2\u003c/sub\u003e emissions using the TROPOMI spectrometer on board ESA\u0026rsquo;s Sentinel-5P polar orbiting platform. TROPOMI has a spectral resolution of 0.25 to 0.54 nm and a spatial resolution of 3.5 \u0026times; 7 km at launch (Veefkind et al., 2012) and updated to 3.5 \u0026times; 5.5 km on 6 August 2019. TROPOMI observes every point on the Earth\u0026rsquo;s surface at least once per day. The TROPOMI Differential Optical Absorption Spectroscopy (DOAS) retrieval algorithm calculates SO\u003csub\u003e2\u003c/sub\u003e Vertical Column Densities (VCDs) for each pixel within its field of view (Theys et al., 2017); VCDs are then converted to column mass (Quei\u0026szlig;er et al., 2019). The total SO\u003csub\u003e2\u003c/sub\u003e mass loading for a given scene is calculated by summing the column mass of SO\u003csub\u003e2\u003c/sub\u003e contained within each pixel above 3 times the random noise.\u003c/p\u003e \u003cp\u003eThe TROPOMI instrumental response to SO\u003csub\u003e2\u003c/sub\u003e is height-dependent and therefore the assumed plume altitude is the main source of uncertainty in SO\u003csub\u003e2\u003c/sub\u003e retrievals. Gas plume or eruption column heights, where reported by RVO or Darwin Volcanic Ash Advisory Centre (VAAC), were used to represent the SO\u003csub\u003e2\u003c/sub\u003e plume altitude. Visual observations from May 2019 suggest that the buoyant gas plume generally rises between a few hundred metres to ~\u0026thinsp;1 km above the summit (~\u0026thinsp;1800 m asl) before dispersing laterally (Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Therefore, for days without reported plume height an altitude of 3 km was used.\u003c/p\u003e \u003cp\u003eIt is important to identify other SO\u003csub\u003e2\u003c/sub\u003e sources that might contaminate a scene resulting in overestimation of SO\u003csub\u003e2\u003c/sub\u003e mass flux from Manam. Identified contamination sources include two nearby volcanoes: Kadovar (-3.6069, 144.5878), which has been outgassing regularly in recent years (Plank et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, 2021b) and Ulawun (-5.0514, 151.3310), which has had five confirmed eruptions during the study period (Johnson \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Wood et al. \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, 2021c; McKee et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). TROPOMI scenes contaminated by SO\u003csub\u003e2\u003c/sub\u003e from external sources are identified from (a) activity reports for nearby volcanoes and (b) visual inspection of true colour Sentinel-2 images and TROPOMI VCD scenes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Where possible, the measurement extent is delimited to include only the plume from Manam. If the external sources cannot be clearly separated then the contaminated images are omitted from the time series.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eConverting scene SO\u003csub\u003e2\u003c/sub\u003e mass into a flux requires knowledge of the residence time of SO\u003csub\u003e2\u003c/sub\u003e in the atmosphere. If the SO\u003csub\u003e2\u003c/sub\u003e lifetime exceeds 24 hours, then some proportion of the SO\u003csub\u003e2\u003c/sub\u003e mass in a TROPOMI scene will be residual from the previous day. Uncertainties in the lifetime of SO\u003csub\u003e2\u003c/sub\u003e in the atmosphere make converting total scene mass to SO\u003csub\u003e2\u003c/sub\u003e fluxes non-trivial, especially under different atmospheric conditions (McCormick Kilbride et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Here, the method proposed by Fioletov et al., (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) is used where under steady-state emissions the flux (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varvec{\\Phi }{\\varvec{S}\\varvec{O}}_{2}\\)\u003c/span\u003e\u003c/span\u003e) and SO\u003csub\u003e2\u003c/sub\u003e mass are related by Eq.\u0026nbsp;\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\varvec{\\Phi }{\\varvec{S}\\varvec{O}}_{2}=\\frac{{\\varvec{M}\\varvec{S}\\varvec{O}}_{2}}{\\varvec{\\tau }}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{M}\\varvec{S}\\varvec{O}}_{2}\\)\u003c/span\u003e\u003c/span\u003e is SO\u003csub\u003e2\u003c/sub\u003e (tonnes) and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varvec{\\tau }\\)\u003c/span\u003e\u003c/span\u003e is the residence time of SO\u003csub\u003e2\u003c/sub\u003e in the atmosphere. Three estimates for residence time were used in an attempt to capture the uncertainty related to this variable: 1, 2 and 3 days based on residence times used in similar studies (Beirle et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Laiolo et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; McCormick Kilbride et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The maximum atmospheric residence time for SO\u003csub\u003e2\u003c/sub\u003e plumes from Papua New Guinea volcanoes has previously been estimated to be ~\u0026thinsp;18 h with typical ages being \u0026lt;\u0026thinsp;12 h (McCormick et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Therefore, the fluxes based on a 1-day residence time are used in the discussion. Lastly, we calculate time-averaged SO\u003csub\u003e2\u003c/sub\u003e fluxes by fitting a first-order polynomial to the cumulative \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varvec{\\Phi }{\\varvec{S}\\varvec{O}}_{2}\\)\u003c/span\u003e\u003c/span\u003e emissions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Thermal Anomalies\u003c/h2\u003e \u003cp\u003eThe MODVOLC algorithm measures the radiant heat flux, or volcanic radiative power (VRP), emitted by Manam (Wright et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Wright \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). MODVOLC uses Level 1B products from the MODIS multispectral instrument onboard the NASA Aqua and Terra satellites, providing a 1 km\u003csup\u003e2\u003c/sup\u003e pixel resolution for the infrared bands. These two satellites ensure coverage of most of Earth\u0026rsquo;s surface every 1\u0026ndash;2 days. VRP is the total heat radiated across the area of the anomaly at the time of acquisition and is expressed in W or J s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e according to Eq.\u0026nbsp;\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (Coppola et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Wooster et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Wright et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\varvec{V}\\varvec{R}\\varvec{P} \\left({\\varvec{\\varphi }}_{\\varvec{e}}\\right)=1.89 \\times {\\varvec{A}}_{\\varvec{P}\\varvec{I}\\varvec{X}} \\times {\\varvec{D}}_{\\varvec{P}\\varvec{I}\\varvec{X}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere 1.89 is a best fit regression coefficient calculated using the MIR (Middle Infrared) method (which relates the VRP estimated by the simple power law used by MODVOLC to the expected value under the Planck function; Wooster et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{A}}_{\\varvec{P}\\varvec{I}\\varvec{X}}\\)\u003c/span\u003e\u003c/span\u003e is the area of the pixel, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{D}}_{\\varvec{P}\\varvec{I}\\varvec{X}}\\)\u003c/span\u003e\u003c/span\u003e is the above-background MIR radiance of the pixel. When a hotspot is detected in more than one pixel, the total VRP is the sum of the VRP across all hotspot pixels. Thermal anomaly intensity is classified following Coppola et al (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; also see \u0026ldquo;MIROVA,\u0026rdquo; 2022): \u0026lt;1 MW\u0026thinsp;=\u0026thinsp;Very Low, \u0026ge;\u0026thinsp;1 MW\u0026thinsp;=\u0026thinsp;Low, \u0026ge;\u0026thinsp;10 MW\u0026thinsp;=\u0026thinsp;Moderate, \u0026ge;\u0026thinsp;100 MW\u0026thinsp;=\u0026thinsp;High, \u0026ge;1GW\u0026thinsp;=\u0026thinsp;Very High and \u0026ge;\u0026thinsp;10 GW\u0026thinsp;=\u0026thinsp;Extreme.\u003c/p\u003e \u003cp\u003eManam\u0026rsquo;s two craters are within 1 km of each other and so the measured VRP likely includes thermal contributions from both due to MODIS\u0026rsquo;s pixel resolution. To determine the source of each anomaly, we use (a) thermal infrared (TIR) imagery from NASA\u0026rsquo;s Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) (90 m resolution) and (b) short-wave Infrared (SWIR) (20 m resolution) and True Colour Imagery (TCI) (10 m resolution) from ESA\u0026rsquo;s Sentinel-2 MSI (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Surface Temperatures\u003c/h2\u003e \u003cp\u003eSurface temperature within each of Manam\u0026rsquo;s summit craters are derived from the Sentinel-2 MSI Level 1C product (Top of Atmosphere (TOA) Reflectance). First, TOA reflectance measured from the MSI band 11 (central wavelength of 1610 nm) is converted to radiance using Eq.\u0026nbsp;\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e:\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\varvec{L}\\varvec{\\lambda }=\\frac{{\\varvec{Q}}_{\\varvec{c}\\varvec{a}\\varvec{l}}{\\varvec{E}}_{\\varvec{e}\\varvec{\\lambda }}\\left(\\varvec{c}\\varvec{o}\\varvec{s}\\varvec{\\theta }\\right)}{\\varvec{\\pi }\\left(\\frac{1}{\\varvec{U}}\\right)}/{10}^{4}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{Q}}_{\\varvec{c}\\varvec{a}\\varvec{l}}\\)\u003c/span\u003e\u003c/span\u003e = digital number (DN); \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{E}}_{\\varvec{e}\\varvec{\\lambda }}\\)\u003c/span\u003e\u003c/span\u003e = solar irradiance (W m\u003csup\u003e2\u003c/sup\u003e); \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varvec{\\theta }\\)\u003c/span\u003e\u003c/span\u003e = incidence angle (\u0026deg;), and \u003cb\u003eU\u003c/b\u003e = quantification value (converts value to TOA). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{E}}_{\\varvec{e}\\varvec{\\lambda }}\\)\u003c/span\u003e\u003c/span\u003e,\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varvec{\\theta }\\)\u003c/span\u003e\u003c/span\u003e and \u003cb\u003eU\u003c/b\u003e are drawn from the MSI image metadata.\u003c/p\u003e \u003cp\u003ePixel Integrated Temperatures (PIT) are then calculated for each band using Eq.\u0026nbsp;\u003cspan refid=\"Equ4\" class=\"InternalRef\"\u003e4\u003c/span\u003e (adapted for use with MSI imagery from Francis and Rothery, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e1987\u003c/span\u003e; Rothery, 1988; Harris, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), which is derived from the Planck function (Planck, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e1901\u003c/span\u003e).\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\varvec{T}=\\frac{{\\varvec{C}}_{2}}{\\varvec{\\lambda }\\varvec{l}\\varvec{n}\\left(\\left[\\varvec{\\epsilon }{\\varvec{\\tau }\\varvec{C}}_{1}{\\varvec{\\lambda }}^{-5}/{10}^{6}\\varvec{\\pi }{\\varvec{L}}_{\\varvec{\\lambda }}\\right]+1\\right)}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{L}}_{\\varvec{\\lambda }}\\)\u003c/span\u003e\u003c/span\u003e = radiance (Wm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e sr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e \u0026micro;m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e); \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{C}}_{1}\\)\u003c/span\u003e\u003c/span\u003e = 3.742 \u0026times; 10\u003csup\u003e\u0026ndash;16\u003c/sup\u003e (W m\u003csup\u003e2\u003c/sup\u003e); \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{C}}_{2}\\)\u003c/span\u003e\u003c/span\u003e = 0.0144 (mK); \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varvec{\\lambda }\\)\u003c/span\u003e\u003c/span\u003e = wavelength (m); \u003cb\u003eT\u003c/b\u003e = blackbody temperature (K); \u003cb\u003eε\u003c/b\u003e\u0026thinsp;=\u0026thinsp;emissivity of the radiating surface; and \u003cb\u003eτ\u003c/b\u003e\u0026thinsp;=\u0026thinsp;atmospheric transmissivity. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{C}}_{1}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{C}}_{2}\\)\u003c/span\u003e\u003c/span\u003e are simplified constants representing \u003cb\u003ehc\u003c/b\u003e\u003csup\u003e\u003cb\u003e2\u003c/b\u003e\u003c/sup\u003e and \u003cb\u003ehc/k\u003c/b\u003e, where \u003cb\u003eh\u003c/b\u003e is Planck's constant (6.266 \u0026times; 10\u003csup\u003e\u0026ndash;34\u003c/sup\u003e J s), \u003cb\u003ec\u003c/b\u003e is the speed of light (2.998 \u0026times; 10\u003csup\u003e8\u003c/sup\u003e m s\u003csup\u003e\u0026minus;1\u003c/sup\u003e) and \u003cb\u003ek\u003c/b\u003e is Boltzmann's constant (2987 \u0026micro;m K).\u003c/p\u003e \u003cp\u003eEmissivity for the basaltic andesite lava erupted at Manam (Palfreyman and Cooke \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1976\u003c/span\u003e; McKee \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e1981\u003c/span\u003e) is estimated as 0.852 for band 11, based on analogous lavas measured for emissivity in the John Hopkins ECOSTRES Spectral Library (Meerdink et al. \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Transmissivity for Manam was estimated using MODTRAN (MODerate resolution atmospheric TRANsmission) to be 0.892 for band 11 (Berk et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1 SO\u003csub\u003e2\u003c/sub\u003e Emissions\u003c/h2\u003e \u003cp\u003eDaily SO\u003csub\u003e2\u003c/sub\u003e mass loadings from Manam between 6 May 2018 and 31 December 2021 have a mean of 1.1 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and a median of 0.47 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee). The time series is dominated by several short-duration, high-magnitude emissions associated with explosive eruptions with 75% of daily emissions below the mean. Using a 7-day moving-average, we define four degassing phases using a 1 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e threshold to distinguish Phases 1 and 4 (below threshold) from Phases 2 and 3 (above threshold), which refer to the background and elevated degassing phases respectively. The elevated degassing phases are identified by the moving-average exceeding 1 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and subsequently not dropping below this threshold for more than 8 days. The transition between Phase 2 and 3 is demarcated by a gradient change of cumulative SO\u003csub\u003e2\u003c/sub\u003e emissions, which occurred from 21 July 2019 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003ePhase 1: 6 May 2018\u0026ndash;21 March 2019\u003c/em\u003e \u003c/p\u003e \u003cp\u003eSO\u003csub\u003e2\u003c/sub\u003e emissions for Phase 1 totalled 198.8\u0026thinsp;\u0026plusmn;\u0026thinsp;8.3 kt over 319 days, with a time-averaged SO\u003csub\u003e2\u003c/sub\u003e flux of 0.62 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). After an 11-month period of eruptive quiescence, explosive eruptive activity returned on 10 August 2018, after which 13 minor explosive eruptions, 5 major explosive eruptions and 6 effusive eruptions occurred during Phase 1. The daily emitted SO\u003csub\u003e2\u003c/sub\u003e mass exceeded 3 kt on ten occasions, all associated with explosive behaviour. The largest Phase 1 emission of 22.1 kt SO\u003csub\u003e2\u003c/sub\u003e is linked to explosive and effusive activity on 25 August 2018.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePhase 2: 22 March 2019\u0026ndash;20 July 2019\u003c/em\u003e \u003c/p\u003e \u003cp\u003ePhase 2 began when the 7-day moving average exceeded 1 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and did not drop below this threshold for more than 8 consecutive days. Manam emitted 539.9\u0026thinsp;\u0026plusmn;\u0026thinsp;10.6 kt of SO\u003csub\u003e2\u003c/sub\u003e over 120 days, with a time-averaged flux of 4.72 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, during Phase 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This period contained relatively few eruptive events with 5 minor eruptions and 1 major eruption on 28 June 2019. Despite the lower eruption frequency, SO\u003csub\u003e2\u003c/sub\u003e emissions prior to the major eruption were elevated substantially above the Phase 1 time-averaged flux. A total of 325.8\u0026thinsp;\u0026plusmn;\u0026thinsp;7.6 kt SO\u003csub\u003e2\u003c/sub\u003e was released prior to the 28 June 2019 eruption, with most of these emissions not linked with documented eruptions. The 28 June 2019 eruption alone emitted 58.3\u0026thinsp;\u0026plusmn;\u0026thinsp;0.5 kt of SO\u003csub\u003e2\u003c/sub\u003e and 82.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.4 kt released over the following four days.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePhase 3: 21 July 2019\u0026ndash;16 October 2019\u003c/em\u003e \u003c/p\u003e \u003cp\u003ePhase 3 SO\u003csub\u003e2\u003c/sub\u003e emissions were reduced relative to Phase 2 with a total emitted mass of 127.9\u0026thinsp;\u0026plusmn;\u0026thinsp;4.1 kt over 87 days and a time-averaged flux of 1.5 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). However, emissions remained elevated above the background flux observed during Phase 1. Phase 3 emissions were mostly independent of eruptive activity, with only 3 minor eruptions occurring during this time.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePhase 4: 17 October 2019\u0026ndash;31 December 2021\u003c/em\u003e \u003c/p\u003e \u003cp\u003ePhase 4 began when the 7-day moving average fell below 1 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for 10 consecutive days and remained consistently below this threshold for the remainder of the time series. Phase 4 SO\u003csub\u003e2\u003c/sub\u003e emissions were 565.1\u0026thinsp;\u0026plusmn;\u0026thinsp;24.9 kt over 806 days, with a time-averaged flux of 0.68 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This flux is comparable to that observed in Phase 1 and therefore we interpret these two phases as representative of the stable background SO\u003csub\u003e2\u003c/sub\u003e emission rate. The small peaks in SO\u003csub\u003e2\u003c/sub\u003e emissions during Phase 4 are associated with the 27 minor and 2 major explosive eruptions during this period.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Thermal Anomalies\u003c/h2\u003e \u003cp\u003eThermal emissions have been detected at Manam sporadically since MODIS coverage began in 2002. During this study period, thermal emissions are characterised by periods of frequent elevated VRP, separated by intervals of several months to years with no detectable thermal output (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee). The time series is grouped into three discrete clusters of elevated VRP, where each cluster includes at least 5 thermal anomalies separated by intervals of no longer than 60 days.\u003c/p\u003e \u003cp\u003e \u003cem\u003eCluster 1: 7 August 2018\u0026ndash;18 July 2019\u003c/em\u003e \u003c/p\u003e \u003cp\u003eCluster 1 began 3 days prior to the onset of explosive eruptive activity in August 2018 and is linked to the eruptions during late 2018 and early 2019. Thermal anomalies became more sporadic in May 2019 but subsequently increased in frequency coinciding with a series of eruptions in June 2019 that culminated in the 28 June 2019 major eruption. Cluster 1 contains 44 detected anomalies, 18 of which are classified as high intensity and are associated with explosive eruptions (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD). Two anomalies are classified as very high intensity and are both coincident with extensive lava flows that almost reached the coastline on 30 September 2018 (2211 MW) and between 8 and 11 January 2019 (1371 MW).\u003c/p\u003e \u003cp\u003e \u003cem\u003eCluster 2: 29 July 2020\u0026ndash;8 September 2020\u003c/em\u003e \u003c/p\u003e \u003cp\u003eCluster 2 includes 5 low to moderate intensity thermal anomalies ranging between 8 and 50 MW. These detections coincide with a period of minor explosive eruptions reported between July and September 2020 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed).\u003c/p\u003e \u003cp\u003e \u003cem\u003eCluster 3: 19 June 2021\u0026ndash;21 December 2021\u003c/em\u003e \u003c/p\u003e \u003cp\u003eCluster 3 occurred during a period of unrest that began in June 2021 continuing throughout the remainder of 2021. The cluster began with a series of low to moderate intensity thermal anomalies following a minor explosive eruption on 23 June 2021. Increased thermal emissions in August 2021 were reflected in 11 anomalies, which included three high intensity detections. Four moderate to high intensity anomalies were detected following reports of Strombolian activity on 18 October 2021 and prior to the 20 October 2021 major eruption. Several low to moderate anomalies were detected in late November to December 2021 in the weeks prior to another major eruption on 22 December 2021 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Surface Temperatures\u003c/h2\u003e \u003cp\u003eThrough the period of observation, Main and South Crater were completely obscured by cloud (meteorological or volcanic) in approximately 75% of the 277 Sentinel-2/MSI images available (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Surface temperatures were therefore calculated for Main and South Craters from 68 and 71 images, respectively. The daily maximum PIT for Main Crater ranged from 340\u0026thinsp;\u0026plusmn;\u0026thinsp;83\u0026deg;C to 509\u0026thinsp;\u0026plusmn;\u0026thinsp;124\u0026deg;C and 335\u0026thinsp;\u0026plusmn;\u0026thinsp;79\u0026deg;C to 510\u0026thinsp;\u0026plusmn;\u0026thinsp;124\u0026deg;C for South Crater. The mean inter-crater maximum PIT divergence was 31\u0026deg;C and a maximum of 158\u0026deg;C where the temperature of both craters could be measured together with South Crater having a higher maximum PIT 40 of 60 days, just 18 of these exceeding mean inter-crater divergence. There was just one occurrence where Main Crater maximum PIT exceeded South Crater by more than the mean divergence, a 57\u0026deg;C difference on 27 September 2018 during the emplacement of a lava flow from Main Crater.\u003c/p\u003e \u003cp\u003eThe maximum South Crater PIT during the study period was measured on 20 May 2019 but Main Crater was cloud-covered at the time, preventing direct comparison. However, the following simultaneous measurement of both craters\u0026mdash;on 30 May 2019\u0026mdash;has the second highest inter-crater temperature divergence of 122\u0026deg;C. A bright hotspot at South Crater in MSI thermal imagery was observed on this date and follows UAS in situ observations of shallow magma within South Crater on 22 May 2019 (Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Cloud-cover prevented MSI measurements throughout June 2019 and therefore the 28 June 2019 major eruption is not captured in this time series. However, following this eruption, a further two maximum pixel-integrated temperatures of 500 and 501\u0026deg;C, with means of 455\u0026deg;C and 473\u0026deg;C, were measured at South Crater on 14 July and 19 July 2019 respectively. This suggests a period of extended high temperatures at South Crater through May to July 2019, during which a major eruption occurred. The mean PIT on 13 August 2019 (403\u0026deg;C) and 18 August 2019 (408\u0026deg;C) decline more substantially than the maximum temperatures, 497\u0026deg;C and 475\u0026deg;C respectively, compared to the two retrievals in July (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). As the difference between maximum and mean PIT is indicative of the spatial distribution of the hot radiating surface, we interpret this result to indicate that although hot material remained in South Crater it likely occupied a smaller portion of the crater area.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Coupling between SO\u003csub\u003e2\u003c/sub\u003e and Thermal Emissions\u003c/h2\u003e \u003cp\u003eWe present a combined time series of SO\u003csub\u003e2\u003c/sub\u003e and thermal emissions (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), which are key parameters for observing changes in open vent activity where an established connection exists between a shallow reservoir and the surface (Wright et al. \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Sparks et al. \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Pyle et al. \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Blackett \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Aiuppa \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Thermal anomalies associated with volcanic edifices may indicate that magma is at or near the surface (Coppola et al., 2012; Dehn and Harris, 2015; Harris, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) and SO\u003csub\u003e2\u003c/sub\u003e emissions provide an insight into conduit permeability (Edmonds et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), magma convection (Shinohara \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), and the volume of degassing magma present at shallow depths (Allard et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Aiuppa et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThermal and SO\u003csub\u003e2\u003c/sub\u003e emissions have been shown to be coupled at during background level activity at open vent systems such as Stromboli (Italy) (Laiolo et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), Bagana (Papua New Guinea) (McCormick et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; McCormick Kilbride et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), Batu Tara (Indonesia) (Laiolo et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), Tinakula (Solomon Islands) (Laiolo et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and Mt. Etna (Italy) (Coppola et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; D\u0026rsquo;Aleo et al., 2019). Contrastingly, periods of uncoupled behaviour between these two parameters can signal transient disturbances to the magmatic system. At Stromboli, elevated SO2 emissions generally lag behind peaks in thermal emissions associated with paroxysms and lava flows (Laiolo et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Periods of high SO2 flux but low radiant heat flux during quiescence phases are typically attributed to unerupted magmatic intrusions (e.g. Mt Etna 2005-06, Coppola et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Conversely, periods of below-average SO2 flux but high radiant flux have been explained by extrusion of previously degassed magma (e.g., Tinakula 2006\u0026ndash;2012; Laiolo et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHere, we evaluate the degree to which the SO\u003csub\u003e2\u003c/sub\u003e and thermal emission timeseries at Manam are coupled for each of the four degassing phases (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The correlation between thermal and SO\u003csub\u003e2\u003c/sub\u003e emissions is calculated using weekly total emissions as it allows the mostly continuous SO\u003csub\u003e2\u003c/sub\u003e flux to be compared to the intermittent thermal emissions. The correlation between total weekly SO2 emissions and total weekly volcanic radiative power is weak across all phases with Phase 1 having the strongest correlation coefficient (r) of 0.27 (Fig. S2). These weak correlations are likely due to the fact that peak VRPs are associated with lava effusions whereas peak SO\u003csub\u003e2\u003c/sub\u003e emissions typically coincide with explosive eruptions as well as the fact that SO\u003csub\u003e2\u003c/sub\u003e emissions have a range of 3 orders of magnitude (0.92\u0026ndash;96 kt) compared to 4 for thermal emissions (5\u0026ndash;2792 MW). While the correlation between the magnitude of the two parameters is weak the temporal relationship of the SO\u003csub\u003e2\u003c/sub\u003e and thermal emissions throughout the time series can be used to interpret processes governing the observed activity and this temporal aspect is explored throughout the remainder of this work.\u003c/p\u003e \u003cp\u003eThroughout Phase 1, peaks in SO\u003csub\u003e2\u003c/sub\u003e emission occurred coincident with periods of heightened radiant flux and were typically aligned with an observed eruption (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). On five occasions, thermal anomalies begin to be detected days to weeks ahead of eruptions with coincident elevated SO\u003csub\u003e2\u003c/sub\u003e emissions (e.g. August 2018, October 2018, December 2018, January 2019 and March 2019, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). This was not the case for the 8 January 2019 and 23 January 2019 eruptions, where SO\u003csub\u003e2\u003c/sub\u003e emissions peaked without thermal anomaly detections, likely due the presence of cloud and the ash-rich plume obscuring thermal detections.\u003c/p\u003e \u003cp\u003eThroughout Phase 2, SO\u003csub\u003e2\u003c/sub\u003e and thermal emissions are not well correlated temporally, except around the 28 June 2019 eruption (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The elevated outgassing characterising this phase (4.72\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) occurred alongside four thermal anomalies in March-May, one of which was coincident with an above-average SO\u003csub\u003e2\u003c/sub\u003e emission. Magma was observed in situ deep within South Crater on 22 May 2019 (Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e) but no anomaly was detected, likely due to cloud cover obscuring the summit at the time of the satellite overpass (10:30 AM and 1:30 PM). MSI SWIR hotspots were observed on 20 May 2019 and 30 May 2019, both with high surface temperatures (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea), which suggest similar conditions were likely present on 22 May 2019. Persistent cloud cover during Phase 2 obscured MSI observations (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb) and, if MODIS acquisitions were similarly affected, under-reporting in the frequency of thermal anomaly detections is likely during this period. Nevertheless, the frequency of thermal anomalies increased in June 2019 alongside an increase in eruptive activity compared to the prior two months, including the 28 June 2019 major eruption. Following an 11-day period of subdued SO\u003csub\u003e2\u003c/sub\u003e emissions between 3 and 13 June 2019, during which\u0026thinsp;\u0026lt;\u0026thinsp;7 kt was emitted over 9 days, elevated gas emissions are resumed alongside the escalating frequency of thermal anomalies.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eElevated SO\u003csub\u003e2\u003c/sub\u003e emissions continued in Phase 3 (1.5\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) but were much reduced compared to Phase 2. Only one thermal anomaly was detected coinciding with an SO\u003csub\u003e2\u003c/sub\u003e emission of 1.9 kt on the day of a minor eruption 29 September 2019 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Overall, Phase 3 represents an extended period where gas and thermal emissions appear to be decoupled, though both parameters appear to be declining in intensity.\u003c/p\u003e \u003cp\u003eSO\u003csub\u003e2\u003c/sub\u003e emissions in Phase 4 (~\u0026thinsp;0.68\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) returned to a comparable level to Phase 1 (~\u0026thinsp;0.62\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). Phase 1 exhibited temporal coupling between thermal and SO\u003csub\u003e2\u003c/sub\u003e emissions, but this is not the case in Phase 4. Although instances where thermal anomalies coincide with above-average SO\u003csub\u003e2\u003c/sub\u003e emissions occur (e.g. 20 October 2021), most above-background emissions occur when no thermal anomalies are detected. During thermal Clusters 2 and 3, the magnitude of SO\u003csub\u003e2\u003c/sub\u003e emissions are unchanged relative to outside these periods of heightened thermal emissions (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eConsequently, we propose that both coupled and decoupled regimes in SO\u003csub\u003e2\u003c/sub\u003e and thermal emissions are present within the study period, and that a transition from coupled to decoupled behaviour occurred following the last eruption of Phase 1 on 29 March 2019. Phase 1 operates under the coupled regime, where peaks in the two parameters are well correlated in time but not in absolute magnitude. In contrast, Phases 2\u0026ndash;4 show little to no correlation between either the timing or magnitude of thermal and gas emissions and therefore represent a decoupled regime.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Is persistent outgassing balanced by magma flux?\u003c/h2\u003e \u003cp\u003eMass balance calculations at mafic open vent volcanoes, globally, suggest that the amount of magma required to sustain observed gas fluxes is generally far greater than that erupted (Edmonds et al., 2022; Kazahaya et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Laiolo et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; McCormick et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Constraining the mass balance of magma in terms of total inputs and outputs, and variations over time, is key to relating observed gas emissions to the magmatic and eruptive processes operating at open vent volcanoes. SO\u003csub\u003e2\u003c/sub\u003e emissions and radiant flux can be used to infer the amounts of magma supplied to the shallow magmatic system (input) and erupted (output) respectively (Harris et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Coppola et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Here we use TROPOMI-derived SO\u003csub\u003e2\u003c/sub\u003e masses, estimated erupted volumes based on plume heights reported in Volcanic Ash Advisory Bulletin (Darwin VAAC), and lava flow inundation areas estimated from Sentinel-2 satellite multispectral imagery using ArcGIS to calculate the mass balance.\u003c/p\u003e \u003cp\u003eThe volume of magma required to generate the observed SO\u003csub\u003e2\u003c/sub\u003e emissions is calculated using Eq.\u0026nbsp;\u003cspan refid=\"Equ5\" class=\"InternalRef\"\u003e5\u003c/span\u003e:\u003cdiv id=\"Equ5\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ5\" name=\"EquationSource\"\u003e\n$$V= \\frac{f}{c \\rho \\gamma {\\Delta }S} \\times {10}^{-9}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e5\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(V\\)\u003c/span\u003e\u003c/span\u003e = magma volume (km\u003csup\u003e3\u003c/sup\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(f\\)\u003c/span\u003e\u003c/span\u003e = measured SO\u003csub\u003e2\u003c/sub\u003e flux (kg d\u003csup\u003e\u0026minus;1\u003c/sup\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(c\\)\u003c/span\u003e\u003c/span\u003e = S to SO\u003csub\u003e2\u003c/sub\u003e conversion constant (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(c\\)\u003c/span\u003e\u003c/span\u003e = 2), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\rho\\)\u003c/span\u003e\u003c/span\u003e = magma density (2640 kg m\u003csup\u003e\u0026minus;3\u003c/sup\u003e), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\gamma\\)\u003c/span\u003e\u003c/span\u003e = vesicularity (expressed as melt fraction i.e. 1\u0026thinsp;=\u0026thinsp;0% porosity, 0.7\u0026thinsp;=\u0026thinsp;30% porosity), and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\Delta }S\\)\u003c/span\u003e\u003c/span\u003e = degassed sulphur (ppm \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\times\\)\u003c/span\u003e\u003c/span\u003e 10\u003csup\u003e\u0026minus;6\u003c/sup\u003e). Density is calculated using the method of Bottinga and Weill (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1970\u003c/span\u003e) using bulk rock compositions are from Palfreyman and Cooke (\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1976\u003c/span\u003e) and McKee (\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e1981\u003c/span\u003e). The values of vesicularity and both initial and degassed melt sulphur contents are unconstrained for recent eruptive products; therefore, the vesicularity term is varied between 0 and 30%, and the total degassed sulphur is approximated as 0.2 \u0026plusmn; 0.02 wt%, based on the range of undegassed sulphur contents in melt inclusions from arc magmas (Lloyd et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Taracs\u0026aacute;k et al. \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eManam emitted 1432\u0026thinsp;\u0026plusmn;\u0026thinsp;48 kt of SO\u003csub\u003e2\u003c/sub\u003e between 6 May 2018 and 31 December 2021 implying degassing of 0.12 to 0.22 km\u003csup\u003e3\u003c/sup\u003e of magma (Eq.\u0026nbsp;\u003cspan refid=\"Equ5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) (TABLE 1). The magma volumes required to yield the cumulative SO\u003csub\u003e2\u003c/sub\u003e emissions for each degassing phase are summarised in (TABLE 1).\u003c/p\u003e \u003cp\u003e \u003cem\u003eEffusively Erupted Magma Volume\u003c/em\u003e \u003c/p\u003e \u003cp\u003eSix lava flows were emplaced between May 2018 and December 2021, identified in satellite imagery. The area inundated by each flow was measured using ArcGIS from MSI, ASTER, and NASA\u0026rsquo;s Landsat 7 \u0026amp; 8 multispectral imagery (Fig. S3). The volume of each flow was approximated based on the average thickness of the September-October 2018 flow, which was estimated as ~\u0026thinsp;3.5 m based on a digital elevation model (DEM) (Fig. S4) created from an Unoccupied Aerial System (DJI Mavic 2 Pro) video of Manam\u0026rsquo;s North East Valley. A porosity of 18.8% was measured (see supplementary materials) in a lava sample from the distal portion of the 28 June 2019 lava flow and used to convert bulk volume to dense rock equivalent (DRE) (Table S2).\u003c/p\u003e \u003cp\u003e \u003cem\u003eExplosively Erupted Magma Volume\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe volume of magma erupted during explosive events can be estimated using the relationship between eruption plume heights (H, km) and erupted volumes (V, km\u003csup\u003e3\u003c/sup\u003e DRE) presented by Mastin et al. (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). This relation is empirical, based on a catalogue of 34 moderate to large explosive eruptions spanning mafic to silicic magma compositions (Eq.\u0026nbsp;6):\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(H=\\)\u003c/span\u003e \u003c/span\u003e25.9\u0026thinsp;+\u0026thinsp;6.64log\u003csub\u003e10\u003c/sub\u003e (V) (6)\u003c/p\u003e \u003cp\u003eAn uncertainty of approximately one order of magnitude is associated with using Eq.\u0026nbsp;6 to calculate erupted volume from plume height (L. Mastin, Pers Com.). Here, we use Eq.\u0026nbsp;6 to calculate erupted volumes at Manam based on explosive eruption column heights reported in RVO bulletins and Darwin VAAC reports (Fig. S5).\u003c/p\u003e \u003cp\u003e \u003cem\u003eMagma Flux Balance\u003c/em\u003e \u003c/p\u003e \u003cp\u003eOver the duration of the study period, more magma was erupted (~\u0026thinsp;0.25 km\u003csup\u003e3\u003c/sup\u003e) than the maximum estimate of magma entering the system (~\u0026thinsp;0.22 km\u003csup\u003e3\u003c/sup\u003e). The eruptions during Phase 1 were responsible for 70% of the erupted magma over the entire time series (FI. 6a). In contrast, the supply of magma was relatively steady as indicated by the consistent cumulative magma input gradient in Phases 1, 3, and 4 (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). The daily magma net balance shows a steady low magma supply indicated by Manam\u0026rsquo;s persistent degassing alongside intermittent high magnitude magma outputs (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003e \u003cem\u003eVRP as a Proxy for System Pressure\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe height of magma columns and lava lakes can vary substantially over days to months, and this is particularly well observed at open vent systems. At open-vent systems like Manam, where the magmatic system is open to the atmosphere, variations in magma column height have been interpreted to reflect changing pressure within the shallow magma reservoirs (Patrick et al., 2015; Moussallam et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Lev et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Calvari et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, Johnson et al., 2018). High system pressure is reflected in an elevated magma column which may rise high enough to be detected as a thermal anomaly and in some cases visible within the crater. Conversely, lower pressure in the magmatic system results in the magma column being too deep to be detected as a thermal anomaly. While eruptions are associated with high pressure conditions, lava effusions typically produce high VRP due to their large radiating surface (Blackett \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) and anomalies linked to explosive eruptions have high VRP due to the radiated heat of large volumes of magma being ejected. Therefore, only non-eruption related thermal anomalies are used here to indicate relative pressure.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Interpretation \u0026ndash; A magma recharge event captured from space?\u003c/h2\u003e \u003cp\u003eHere we interpret the observed SO\u003csub\u003e2\u003c/sub\u003e and thermal emissions, the coupling of these parameters, the magma flux balance, and periods of high pressure within the context of a magma recharge event to examine the processes and feedbacks at work throughout the study period.\u003c/p\u003e \u003cp\u003eThe 10 August 2018 eruption ended 11 months of quiescence and passing degassing following the end of 2017 eruptive period on 10 September 2017 (2021a). The absence of thermal anomalies during this interruptive period (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) suggests a low magma column level and therefore low system pressure. We infer that magma residual from the 2017 eruptive period continued to degas over this period, sustaining the observed SO\u003csub\u003e2\u003c/sub\u003e emissions and driving sluggish convection in the shallow plumbing system (Kazahaya et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Allard \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Witter et al. \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Beckett et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). However, continued degassing of this residual magma without replenishment would drive volatile depletion, cooling, and crystallisation, initially limiting the mobility of melts and subsequently the permeable migration of exsolved fluids (Edmonds et al. 2022a). Formation of a crystal-rich, semi-permeable plug may be partially responsible for the absence of thermal anomalies in the interruptive period (Stix et al. \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Diller et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Hall et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Gaunt et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn this context, we interpret that the time-averaged Phase 1 SO\u003csub\u003e2\u003c/sub\u003e emissions of ~\u0026thinsp;0.62 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e represent the supply of volatiles derived from second boiling of residual magma, migrating slowly through the semi-permeable plug (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). Thermal anomalies were first detected on 7 August 2018 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed) indicating increasing pressure which was followed by the series of explosive and effusive eruptions in August-September 2018. This increased pressure was likely caused by a volatile-rich magma recharge (Andronico and Corsaro, 2011; Grapenthin et al., 2022; Patrick et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2019\u003c/span\u003ea, 2015; Viccaro et al., 2015), that would have began arriving at the shallow storage region several months prior to the onset of eruptive activity based on estimates from similar systems (Cannata et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Aiuppa et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Petrone et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The presence of the semi-permeable plug would have initially inhibited pressure release (Diller et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Battaglia et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) but continued increasing pressure would eventually exceed plug\u0026rsquo;s strength (Woitischek et al. \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) causing fracturing and complete failure, resulting in the 10 August 2018 eruption. The 13 August 2018 lava effusion (2021a) via the reopened conduit would have further decreased pressure in the magmatic system.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDuring Phase 1, we calculate that 0.18 km\u003csup\u003e3\u003c/sup\u003e of magma was erupted, compared to 0.03 km\u003csup\u003e3\u003c/sup\u003e required to sustain the observed gas emissions (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Table\u0026nbsp;1); this result implies that the erupted magma was extensively degassed prior to eruption. We interpret that the degassed residual magma continued to be removed by the August and September 2018 eruptions (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb), including the 25 August 2018 major explosive eruption and the intermittent effusive activity from 9 September to 1 October 2018 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The eruptive activity in January 2019 removed an estimated 0.11 km\u003csup\u003e3\u003c/sup\u003e (42% of total erupted material during this study) and therefore likely expelled most of the remaining residual magma. The elevated SO\u003csub\u003e2\u003c/sub\u003e emissions of Phase 2 commenced following the 29 March 2019 eruption that removed the final remnants of 2017 residual magma and re-established an open conduit state. Removing residual magma present within the conduit would have reduced the lithostatic load in the upper magmatic system, thereby promoting continued ascent of volatile-rich recharge magma (Calvari et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) and a positive feedback between magma ascent, decompression, and volatile exsolution.\u003c/p\u003e \u003cp\u003eDuring Phase 2 (March \u0026ndash; July 2019), we estimate that 0.08 km\u003csup\u003e3\u003c/sup\u003e of magma entered the shallow plumbing system compared to 0.04 km\u003csup\u003e3\u003c/sup\u003e erupted indicating Manam was in state of open vent excess degassing (Rose et al. \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Edmonds et al. 2022a; Vergniolle and M\u0026eacute;trich \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). In-situ observations of magma within South Crater on 22 May 2019 and two high surface temperature retrievals on 20 and 30 May 2019 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea) suggest at least transient periods of high pressure raising the magma column. However, the decoupled thermal and SO\u003csub\u003e2\u003c/sub\u003e emissions during early Phase 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec), together with the lack of detected anomalies and infrequent eruptions (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed), suggests that the magma level and consequently reservoir pressure, was highly variable. Unlike Phase 1, increased SO\u003csub\u003e2\u003c/sub\u003e emissions are not explicitly linked to eruptions or to thermal anomaly detections; therefore we suggest that surface outgassing involved an enhanced contribution from open system fluxing of volatiles independent of magma ascent, transferred from degassing of recharge magma within the shallow reservoir (Edmonds et al. 2022a).\u003c/p\u003e \u003cp\u003eThe later stage of Phase 2 (June \u0026ndash; July 2019) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ed) displayed increased thermal emissions and eruptive activity compared to earlier in Phase 2. Two surface temperature measurements of \u0026gt;\u0026thinsp;450\u0026deg;C in South Crater (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea) alongside increased frequency of thermal anomaly detections and recorded eruptions suggest that the magma level remained high in the conduit caused by high reservoir pressure. A marked reduction in daily SO\u003csub\u003e2\u003c/sub\u003e emissions around early June suggests that permeability within the conduit was briefly obstructed (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). However, the subsequent return to elevated outgassing levels following the 7 June 2019 eruption suggests that explosive activity may have reopened degassing pathways. We propose that the intense degassing and resulting dehydration-driven crystallisation during Phase 2 may have promoted the development of another plug (Applegarth et al., 2013; Couch et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Gaunt et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Lipman et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e1985\u003c/span\u003e), reducing the permeability of the magma and consequently promoting gas accumulation beneath the plug (Stix et al. \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Sparks \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Burgisser et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Three minor explosive eruptions on 7, 8, and 18 June 2019 likely evidence increasing strain on the plug but it appears that each failed to fully re-open the conduit. We attribute the 28 June 2019 major eruption to the eventual catastrophic failure of this plug, releasing accumulated volatiles and triggering rapid downward-propagating decompression. Explosive activity during this event generated a 15.2 km eruption column, released 58.3 kt of SO\u003csub\u003e2\u003c/sub\u003e, and erupted\u0026thinsp;~\u0026thinsp;0.018 km\u003csup\u003e3\u003c/sup\u003e of magma (7% of total erupted material during this study period).\u003c/p\u003e \u003cp\u003eSO\u003csub\u003e2\u003c/sub\u003e emissions during Phase 3 remained above background levels but reduced from ~\u0026thinsp;4.72 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to ~\u0026thinsp;1.54 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, which may reflect a gradual depletion of volatiles in the shallow magmatic system. SO\u003csub\u003e2\u003c/sub\u003e and thermal emissions remained decoupled in Phase 3. To explain the low frequency of both thermal anomalies and reported eruptions, we infer that the removal of a substantial volume of magma and volatiles would have reduced the pressure in the shallow magmatic system considerably (Anderson et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Patrick et al. \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Barri\u0026egrave;re et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ee).The fact that surface SO\u003csub\u003e2\u003c/sub\u003e emissions were sustained at elevated fluxes throughout this period (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee) signals that conduit convection and volatile fluxing remained active despite the reduction in system pressure. During this period, SO\u003csub\u003e2\u003c/sub\u003e emissions imply that 0.02 km\u003csup\u003e3\u003c/sup\u003e of magma was supplied compared to just 0.002 km\u003csup\u003e3\u003c/sup\u003e being erupted (Table\u0026nbsp;1), indicating that Manam maintained a state of open vent excess degassing.\u003c/p\u003e \u003cp\u003eAverage daily SO\u003csub\u003e2\u003c/sub\u003e emissions in Phase 4 (~\u0026thinsp;0.68 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) returned to similar levels to Phase 1 (~\u0026thinsp;0.62 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), which can be considered a return to background degassing levels. Reservoir pressure was likely low during November 2019 to May 2021 since there were just 5 thermal anomaly detections (Cluster 2) associated with a series of 6 minor eruptions in July-September 2020. The return to background SO\u003csub\u003e2\u003c/sub\u003e emissions suggests that the initially volatile-rich recharge magma had become relatively depleted. The dominant process responsible for volatile transport would likely have reverted from volatile fluxing to conduit convection, where gas emissions were once again tied to magma transport within the conduit (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ef) (Beckett et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Edmonds et al. 2022a).. This is supported by not only the decrease in SO\u003csub\u003e2\u003c/sub\u003e emissions compared to the previous two phases (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) but also the reduction in excess degassing with 0.09 km\u003csup\u003e3\u003c/sup\u003e of magma supplied compared to 0.04 km\u003csup\u003e3\u003c/sup\u003e erupted, approximately 75% of which was erupted between August-December 2021.\u003c/p\u003e \u003cp\u003eTogether, the sparsity of thermal anomalies prior to the onset of Cluster 3 in August 2021 alongside the persistent background SO\u003csub\u003e2\u003c/sub\u003e emissions indicates that emissions were decoupled during Phase 4 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The Cluster 3 thermal emissions suggest that Manam was periodically experiencing high pressure throughout the second half of 2021, raising the magma level close to the surface. These anomalies were associated temporally with the eruptive activity reported throughout this period, yet most were not linked to substantial SO\u003csub\u003e2\u003c/sub\u003e emission peaks as had been the case during Phase 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). As such, we interpret that the transient periods of eruptive activity during this background level degassing phase are unlikely to reflect a further recharge of volatile-rich magma reaching the shallow plumbing system. Instead, the lack of SO\u003csub\u003e2\u003c/sub\u003e emissions indicates the involvement of comparatively volatile-depleted magma, where the increased pressure to drive explosive events is more likely the result of reduced conduit permeability through cooling and dehydration-driven crystallisation of the residual unerupted magma (Applegarth et al., 2013; Lipman et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e1985\u003c/span\u003e). A repeating cycle of partial closing and re-opening of the conduit continued throughout the remainder of 2021, with more substantial and protracted reductions in permeability likely preceding the major eruptions on 20 October 2021 and 22 December 2021 (Hall et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Battaglia et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Long-term SO\u003csub\u003e2\u003c/sub\u003e Emission Variability\u003c/h2\u003e \u003cp\u003eThe time series of SO\u003csub\u003e2\u003c/sub\u003e emissions at Manam presented in this chapter demonstrates the variability of volcanic volatile emissions over months to years, even at persistently degassing volcanoes. The annual mean daily SO\u003csub\u003e2\u003c/sub\u003e emissions (based on a 1 day residence time) are presented alongside the annual mean daily SO\u003csub\u003e2\u003c/sub\u003e emissions (2005\u0026ndash;2015) first reported by Carn et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e. Carn et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) identified a declining trend in annual SO\u003csub\u003e2\u003c/sub\u003e emissions and the inclusion of the data from this study shows that this declining trend continues despite the elevated emissions during 2019 (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The 2019 annual daily mean emissions of 2.2 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e exceeds substantially the 2015\u0026ndash;2021 mean (1.4 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and represents a striking departure from the long-term trend, to which emissions in 2020 and 2021 return (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). This observation demonstrates how open vent volcanoes, such as Manam, can exhibit wide fluctuations in emissions, which are superimposed on decadal trends. If placed within the global SO\u003csub\u003e2\u003c/sub\u003e inventory compiled by Carn et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), Manam\u0026rsquo;s SO\u003csub\u003e2\u003c/sub\u003e emissions would be ranked 31st in Phase 1, 3rd in Phase 3, 10th in Phase 3, and 25th in Phase 4 (Table S3). This variability has also been recognised at other open vent systems, including Bagana (Papua New Guinea; (McCormick Kilbride et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and consequently highlights the inherent limitations and uncertainties associated with compiling global volcanic volatile inventories, especially where short duration or campaign measurements are relied on. Additionally, this analysis emphasises the need for a multi-parametric approach to interpreting changes in degassing behaviour and eruptive activity at open vent volcanic systems, as the processes responsible for modulating degassing are varied and interpretations potentially ambiguous when derived from emission rates alone.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eWe have used a multi-parameter remote sensing approach to investigate the subsurface processes responsible for the period of elevated volcanic activity observed at Manam between August 2018 to December 2021. Using satellite-based measurements of thermal and sulphur dioxide (SO\u003csub\u003e2\u003c/sub\u003e) emissions, combined with in situ observations of volcanic activity, we quantify the relative inputs and outputs of magma and gas to the shallow conduit and surface\u0026mdash;and consequently the varying extent of excess degassing through time.\u003c/p\u003e \u003cp\u003eFrom these timeseries, four distinct phases of volcanic activity are identified between 2018 and 2021. To explain these phases in the context of volcanological processes, we propose that eruptive activity at Manam during the period of observation was driven by the injection and eruption of a volatile-rich recharge magma. In this conceptual model, initial eruptions in August 2018\u0026mdash;triggered by positive reservoir pressure changes after a year long period of repose\u0026mdash;removed previously degassed residual magma to re-open the conduit and promote efficient fluxing of segregated volatiles through the shallow magmatic system, accounting for the very high SO\u003csub\u003e2\u003c/sub\u003e fluxes (4.72 kt day\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) observed in March-June 2019. A period of lower emissions, both thermal and SO\u003csub\u003e2\u003c/sub\u003e, immediately prior to the major eruption on 28 June 2019 points to reduced permeability and ultimately failure of a conduit plug as a likely trigger mechanism. We suggest plug formation may have been promoted by the extended period of enhanced degassing and resulting dehydration-driven crystallisation in the shallow conduit.\u003c/p\u003e \u003cp\u003eThe multi-parameter timeseries has allowed the estimation of Manam\u0026rsquo;s magma budget and resolved the previously unknown fate of the magma supplying Manam\u0026rsquo;s high SO\u003csub\u003e2\u003c/sub\u003e flux in 2019 (Liu et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Overall, we calculate that the magma output exceeds the magma input if we consider the entire timeseries examined in this study. However, since this output component is dominated by eruptions from August 2018 to March 2019\u0026mdash;which we infer to involve residual degassed magma from 2017\u0026mdash;this suggests that the magma budget is balanced over long timescales and that degassed magma is eventually erupted at Manam rather than intruded.\u003c/p\u003e \u003cp\u003eIn the context of long-term SO\u003csub\u003e2\u003c/sub\u003e degassing trends, the period of enhanced degassing in 2019 is superimposed on a long-term declining trend in emissions at Manam and is therefore not representative of the time-averaged degassing behaviour of this volcano. This substantial temporal variability is not unique to Manam, and has been recognised at other strong open vent emitters (e.g., Bagana, Papua New Guinea; McCormick Kilbride et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These observations highlight an important limitation to acknowledge when extrapolating short-term or campaign measurements at open vent volcanoes to long-term emissions contributions within global volcanic volatile inventories.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments:\u0026nbsp;\u003c/strong\u003eThe authors wish to thank the Ima Itikarai and Kila Mulina (Rabaul Volcanological Observatory) for providing a sample of the 28 June 2019 lava flow. We acknowledge Keiran Wood (University of Manchester) who piloted the Unoccupied Aerial System which captured footage Manam which was used to create the orthomosiac model of the October 2018 lava flow.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contribution:\u0026nbsp;\u003c/strong\u003eAdam Cotterill, Emma Nicholson \u0026amp; Christopher Kilburn contributed to this studies conception and design. Material preparation, data collection and analysis were performed by Adam Cotterill and Emma Nicholson. Sentinel-5P TROPOMI SO\u003csub\u003e2\u003c/sub\u003e mass retrieval process was provided by Catherine Hayer. The first draft of the manuscript was written by Adam Cotterill and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. This work was supported by the UK Natural Environment Research Council (Grant Number NE/S007229/1).\u003c/p\u003e\n\u003cp\u003eConflict of interest: The authors declare no competing interests.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAbbott LD (1995) Neogene tectonic reconstruction of the Adelbert-Finisterre-New Britain collision, northern Papua New Guinea. J Southeast Asian Earth Sci 11:33\u0026ndash;51. https://doi.org/10.1016/0743-9547(94)00032-A\u003c/li\u003e\n \u003cli\u003eAbbott LD, Silver EA, Galewsky J (1994) Structural evolution of a modern arc-continent collision in Papua New Guinea. Tectonics 13:1007\u0026ndash;1034. https://doi.org/10.1029/94TC01623\u003c/li\u003e\n \u003cli\u003eAiuppa A (2015) Volcanic-gas monitoring. In: Volcanism and Global Environmental Change. Cambridge University Press, pp 81\u0026ndash;96\u003c/li\u003e\n \u003cli\u003eAiuppa A, Bitetto M, Delle Donne D, et al (2021) Volcanic CO2 tracks the incubation period of basaltic paroxysms. Sci Adv 7:eabh0191. https://doi.org/10.1126/sciadv.abh0191\u003c/li\u003e\n \u003cli\u003eAiuppa A, Bitetto M, Francofonte V, et al (2017) A CO2-gas precursor to the March 2015 Villarrica volcano eruption. Geochem Geophys Geosystems 18:2120\u0026ndash;2132. https://doi.org/10.1002/2017GC006892\u003c/li\u003e\n \u003cli\u003eAiuppa A, Fischer TP, Plank T, Bani P (2019) CO2 flux emissions from the Earth\u0026rsquo;s most actively degassing volcanoes, 2005\u0026ndash;2015. Sci Rep 9:5442. https://doi.org/10.1038/s41598-019-41901-y\u003c/li\u003e\n \u003cli\u003eAiuppa A, Moretti R, Federico C, et al (2007) Forecasting Etna eruptions by real-time observation of volcanic gas composition. Geology 35:1115\u0026ndash;1118. https://doi.org/10.1130/G24149A.1\u003c/li\u003e\n \u003cli\u003eAllard P (1997) Endogenous magma degassing and storage at Mount Etna. Geophys Res Lett 24:2219\u0026ndash;2222. https://doi.org/10.1029/97GL02101\u003c/li\u003e\n \u003cli\u003eAllard P, Carbonnelle J, M\u0026eacute;trich N, et al (1994) Sulphur output and magma degassing budget of Stromboli volcano. Nature 368:326\u0026ndash;330. https://doi.org/10.1038/368326a0\u003c/li\u003e\n \u003cli\u003eAnderson KR, Poland MP, Johnson JH, Miklius A (2015) Episodic Deflation\u0026ndash;Inflation Events at Kīlauea Volcano and Implications for the Shallow Magma System. In: Hawaiian Volcanoes. American Geophysical Union (AGU), pp 229\u0026ndash;250\u003c/li\u003e\n \u003cli\u003eAndres RJ, Kasgnoc AD (1998) A time-averaged inventory of subaerial volcanic sulfur emissions. J Geophys Res Atmospheres 103:25251\u0026ndash;25261. https://doi.org/10.1029/98JD02091\u003c/li\u003e\n \u003cli\u003eApplegarth LJ, Tuffen H, James MR, Pinkerton H (2013a) Degassing-driven crystallisation in basalts. Earth-Sci Rev 116:1\u0026ndash;16. https://doi.org/10.1016/j.earscirev.2012.10.007\u003c/li\u003e\n \u003cli\u003eApplegarth LJane, Tuffen H, James MR, et al (2013b) Direct observations of degassing-induced crystallization in basalts. Geology 41:243\u0026ndash;246. https://doi.org/10.1130/G33641.1\u003c/li\u003e\n \u003cli\u003eBarri\u0026egrave;re J, d\u0026rsquo;Oreye N, Smets B, et al (2022) Intra-Crater Eruption Dynamics at Nyiragongo (D.R. Congo), 2002\u0026ndash;2021. J Geophys Res Solid Earth 127:e2021JB023858. https://doi.org/10.1029/2021JB023858\u003c/li\u003e\n \u003cli\u003eBattaglia J, Hidalgo S, Bernard B, et al (2019) Autopsy of an eruptive phase of Tungurahua volcano (Ecuador) through coupling of seismo-acoustic and SO2 recordings with ash characteristics. Earth Planet Sci Lett 511:223\u0026ndash;232. https://doi.org/10.1016/j.epsl.2019.01.042\u003c/li\u003e\n \u003cli\u003eBeckett FM, Burton M, Mader HM, et al (2014) Conduit convection driving persistent degassing at basaltic volcanoes. J Volcanol Geotherm Res 283:19\u0026ndash;35. https://doi.org/10.1016/j.jvolgeores.2014.06.006\u003c/li\u003e\n \u003cli\u003eBeirle S, H\u0026ouml;rmann C, Penning de Vries M, et al (2014) Estimating the volcanic emission rate and atmospheric lifetime of SO\u003csub\u003e2\u003c/sub\u003e from space: a case study for Kīlauea volcano, Hawai`i. Atmospheric Chem Phys 14:8309\u0026ndash;8322. https://doi.org/10.5194/acp-14-8309-2014\u003c/li\u003e\n \u003cli\u003eBerk A, Conforti P, Kennett R, et al (2014) MODTRAN\u0026reg; 6: A major upgrade of the MODTRAN\u0026reg; radiative transfer code. In: 2014 6th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS). pp 1\u0026ndash;4\u003c/li\u003e\n \u003cli\u003eBlackett M (2013) Review of the utility of infrared remote sensing for detecting and monitoring volcanic activity with the case study of shortwave infrared data for Lascar Volcano from 2001\u0026ndash;2005. Geol Soc Lond Spec Publ 380:107\u0026ndash;135. https://doi.org/10.1144/SP380.10\u003c/li\u003e\n \u003cli\u003eBlackett M (2017) An Overview of Infrared Remote Sensing of Volcanic Activity. J Imaging 3:13. https://doi.org/10.3390/jimaging3020013\u003c/li\u003e\n \u003cli\u003eBottinga Y, Weill DF (1970) Densities of liquid silicate systems calculated from partial molar volumes of oxide components. Am J Sci 269:169\u0026ndash;182\u003c/li\u003e\n \u003cli\u003eBurgisser A, Arbaret L, Druitt TH, Giachetti T (2011) Pre-explosive conduit conditions of the 1997 Vulcanian explosions at Soufri\u0026egrave;re Hills Volcano, Montserrat: II. Overpressure and depth distributions. J Volcanol Geotherm Res 199:193\u0026ndash;205. https://doi.org/10.1016/j.jvolgeores.2010.11.014\u003c/li\u003e\n \u003cli\u003eCalder ES, Cole PD, Dade WB, et al (1999) Mobility of pyroclastic flows and surges at the Soufriere Hills Volcano, Montserrat. Geophys Res Lett 26:537\u0026ndash;540. https://doi.org/10.1029/1999GL900051\u003c/li\u003e\n \u003cli\u003eCalvari S, Spampinato L, Bonaccorso A, et al (2011) Lava effusion \u0026mdash; A slow fuse for paroxysms at Stromboli volcano? Earth Planet Sci Lett 301:317\u0026ndash;323. https://doi.org/10.1016/j.epsl.2010.11.015\u003c/li\u003e\n \u003cli\u003eCannata A, Di Grazia G, Giuffrida M, et al (2018) Space-Time Evolution of Magma Storage and Transfer at Mt. Etna Volcano (Italy): The 2015\u0026ndash;2016 Reawakening of Voragine Crater. Geochem Geophys Geosystems 19:471\u0026ndash;495. https://doi.org/10.1002/2017GC007296\u003c/li\u003e\n \u003cli\u003eCaricchi L, Sheldrake TE, Blundy J (2018) Modulation of magmatic processes by CO2 flushing. Earth Planet Sci Lett 491:160\u0026ndash;171. https://doi.org/10.1016/j.epsl.2018.03.042\u003c/li\u003e\n \u003cli\u003eCarn SA, Fioletov VE, McLinden CA, et al (2017) A decade of global volcanic SO2 emissions measured from space. Sci Rep 7:44095. https://doi.org/10.1038/srep44095\u003c/li\u003e\n \u003cli\u003eConnell J, Lutkehaus N (2016) Another Manam? The forced migration of the population of Manam Island, Papua New Guinea, due to volcanic eruptions 2004\u0026ndash;2005. International Organisation for Migration\u003c/li\u003e\n \u003cli\u003eCoppola D, Laiolo M, Cigolini C, et al (2016) Enhanced volcanic hot-spot detection using MODIS IR data: results from the MIROVA system. Geol Soc Lond Spec Publ 426:181\u0026ndash;205. https://doi.org/10.1144/SP426.5\u003c/li\u003e\n \u003cli\u003eCoppola D, Laiolo M, Massimetti F, Cigolini C (2019) Monitoring endogenous growth of open-vent volcanoes by balancing thermal and SO2 emissions data derived from space. Sci Rep 9:9394. https://doi.org/10.1038/s41598-019-45753-4\u003c/li\u003e\n \u003cli\u003eCoppola D, Laiolo M, Piscopo D, Cigolini C (2013) Rheological control on the radiant density of active lava flows and domes. J Volcanol Geotherm Res 249:39\u0026ndash;48. https://doi.org/10.1016/j.jvolgeores.2012.09.005\u003c/li\u003e\n \u003cli\u003eCoppola D, Valade S, Masias P, et al (2022) Shallow magma convection evidenced by excess degassing and thermal radiation during the dome-forming Sabancaya eruption (2012\u0026ndash;2020). Bull Volcanol 84:16. https://doi.org/10.1007/s00445-022-01523-1\u003c/li\u003e\n \u003cli\u003eCouch S, Sparks RSJ, Carroll MR (2003) The Kinetics of Degassing-Induced Crystallization at Soufri\u0026egrave;re Hills Volcano, Montserrat. J Petrol 44:1477\u0026ndash;1502. https://doi.org/10.1093/petrology/44.8.1477\u003c/li\u003e\n \u003cli\u003eDiller K, Clarke AB, Voight B, Neri A (2006) Mechanisms of conduit plug formation: Implications for vulcanian explosions. Geophys Res Lett 33:. https://doi.org/10.1029/2006GL027391\u003c/li\u003e\n \u003cli\u003eEdmonds, Liu EJ, Cashman KV (2022a) Open-vent volcanoes fuelled by depth-integrated magma degassing. Bull Volcanol 84:28. https://doi.org/10.1007/s00445-021-01522-8\u003c/li\u003e\n \u003cli\u003eEdmonds M, Mason E, Hogg O (2022b) Volcanic Outgassing of Volatile Trace Metals. Annu Rev Earth Planet Sci 50:79\u0026ndash;98. https://doi.org/10.1146/annurev-earth-070921-062047\u003c/li\u003e\n \u003cli\u003eEdmonds M, Oppenheimer C, Pyle DM, et al (2003) SO2 emissions from Soufri\u0026egrave;re Hills Volcano and their relationship to conduit permeability, hydrothermal interaction and degassing regime. J Volcanol Geotherm Res 124:23\u0026ndash;43. https://doi.org/10.1016/S0377-0273(03)00041-6\u003c/li\u003e\n \u003cli\u003eFioletov VE, McLinden CA, Krotkov N, Li C (2015) Lifetimes and emissions of SO2 from point sources estimated from OMI. Geophys Res Lett 42:1969\u0026ndash;1976. https://doi.org/10.1002/2015GL063148\u003c/li\u003e\n \u003cli\u003eFlower VJB, Carn SA (2015) Characterising volcanic cycles at Soufriere Hills Volcano, Montserrat: Time series analysis of multi-parameter satellite data. J Volcanol Geotherm Res 304:82\u0026ndash;93. https://doi.org/10.1016/j.jvolgeores.2015.07.035\u003c/li\u003e\n \u003cli\u003eFrancis PW, Rothery DA (1987) Using the Landsat Thematic Mapper to detect and monitor active volcanoes: An example from Lascar volcano, northern Chile. Geology 15:614\u0026ndash;617. https://doi.org/10.1130/0091-7613(1987)15\u0026lt;614:UTLTMT\u0026gt;2.0.CO;2\u003c/li\u003e\n \u003cli\u003eGaunt HE, Burgisser A, Mothes PA, et al (2020) Triggering of the powerful 14 July 2013 Vulcanian explosion at Tungurahua Volcano, Ecuador. J Volcanol Geotherm Res 392:106762. https://doi.org/10.1016/j.jvolgeores.2019.106762\u003c/li\u003e\n \u003cli\u003eGirina OA (2013) Chronology of Bezymianny Volcano activity, 1956\u0026ndash;2010. J Volcanol Geotherm Res 263:22\u0026ndash;41. https://doi.org/10.1016/j.jvolgeores.2013.05.002\u003c/li\u003e\n \u003cli\u003eGray DM, Burton-Johnson A, Fretwell PT (2019) Evidence for a lava lake on Mt. Michael volcano, Saunders Island (South Sandwich Islands) from Landsat, Sentinel-2 and ASTER satellite imagery. J Volcanol Geotherm Res 379:60\u0026ndash;71. https://doi.org/10.1016/j.jvolgeores.2019.05.002\u003c/li\u003e\n \u003cli\u003eHall ML, Steele AL, Bernard B, et al (2015) Sequential plug formation, disintegration by Vulcanian explosions, and the generation of granular Pyroclastic Density Currents at Tungurahua volcano (2013\u0026ndash;2014), Ecuador. J Volcanol Geotherm Res 306:90\u0026ndash;103. https://doi.org/10.1016/j.jvolgeores.2015.09.009\u003c/li\u003e\n \u003cli\u003eHarris A (2013) Thermal Remote Sensing of Active Volcanoes: A User\u0026rsquo;s Manual. Cambridge University Press\u003c/li\u003e\n \u003cli\u003eHarris AJL, Flynn LP, Rothery DA, et al (1999) Mass flux measurements at active lava lakes: Implications for magma recycling. J Geophys Res Solid Earth 104:7117\u0026ndash;7136. https://doi.org/10.1029/98JB02731\u003c/li\u003e\n \u003cli\u003eHolm RJ, Richards SW (2013) A re-evaluation of arc\u0026ndash;continent collision and along-arc variation in the Bismarck Sea region, Papua New Guinea. Aust J Earth Sci 60:605\u0026ndash;619. https://doi.org/10.1080/08120099.2013.824505\u003c/li\u003e\n \u003cli\u003eJames MR, Varley N (2012) Identification of structural controls in an active lava dome with high resolution DEMs: Volc\u0026aacute;n de Colima, Mexico. Geophys Res Lett 39:. https://doi.org/10.1029/2012GL054245\u003c/li\u003e\n \u003cli\u003eJohnson ER, Wallace PJ, Cashman KV, Delgado Granados H (2010) Degassing of volatiles (H2O, CO2, S, Cl) during ascent, crystallization, and eruption at mafic monogenetic volcanoes in central Mexico. J Volcanol Geotherm Res 197:225\u0026ndash;238. https://doi.org/10.1016/j.jvolgeores.2010.02.017\u003c/li\u003e\n \u003cli\u003eJohnson RW (2013) Fire Mountains of the Islands: A History of Volcanic Eruptions and Disaster Management in Papua New Guinea and the Solomon Islands. ANU E Press\u003c/li\u003e\n \u003cli\u003eKazahaya K, Shinohara H, Saito G (1994) Excessive degassing of Izu-Oshima volcano: magma convection in a conduit. Bull Volcanol 56:207\u0026ndash;216. https://doi.org/10.1007/BF00279605\u003c/li\u003e\n \u003cli\u003eLaiolo M, Delle Donne D, Coppola D, et al (2022) Shallow magma dynamics at open-vent volcanoes tracked by coupled thermal and SO2 observations. Earth Planet Sci Lett 594:117726. https://doi.org/10.1016/j.epsl.2022.117726\u003c/li\u003e\n \u003cli\u003eLaiolo M, Massimetti F, Cigolini C, et al (2018) Long-term eruptive trends from space-based thermal and SO2 emissions: a comparative analysis of Stromboli, Batu Tara and Tinakula volcanoes. Bull Volcanol 80:68. https://doi.org/10.1007/s00445-018-1242-0\u003c/li\u003e\n \u003cli\u003eLev E, Ruprecht P, Oppenheimer C, et al (2019) A global synthesis of lava lake dynamics. J Volcanol Geotherm Res 381:16\u0026ndash;31. https://doi.org/10.1016/j.jvolgeores.2019.04.010\u003c/li\u003e\n \u003cli\u003eLipman PW, Banks NG, Rhodes JM (1985) Degassing-induced crystallization of basaltic magma and effects on lava rheology. Nature 317:604\u0026ndash;607. https://doi.org/10.1038/317604a0\u003c/li\u003e\n \u003cli\u003eLiu EJ, Aiuppa A, Alan A, et al (2020) Aerial strategies advance volcanic gas measurements at inaccessible, strongly degassing volcanoes. Sci Adv 6:eabb9103. https://doi.org/10.1126/sciadv.abb9103\u003c/li\u003e\n \u003cli\u003eLloyd AS, Ruprecht P, Hauri EH, et al (2014) NanoSIMS results from olivine-hosted melt embayments: Magma ascent rate during explosive basaltic eruptions. J Volcanol Geotherm Res 283:1\u0026ndash;18. https://doi.org/10.1016/j.jvolgeores.2014.06.002\u003c/li\u003e\n \u003cli\u003eMastin LG, Guffanti M, Servranckx R, et al (2009) A multidisciplinary effort to assign realistic source parameters to models of volcanic ash-cloud transport and dispersion during eruptions. J Volcanol Geotherm Res 186:10\u0026ndash;21. https://doi.org/10.1016/j.jvolgeores.2009.01.008\u003c/li\u003e\n \u003cli\u003eMcCormick BT, Edmonds M, Mather TA, Carn SA (2012) First synoptic analysis of volcanic degassing in Papua New Guinea. Geochem Geophys Geosystems 13:. https://doi.org/10.1029/2011GC003945\u003c/li\u003e\n \u003cli\u003eMcCormick Kilbride BT, Mulina K, Wadge G, et al (2019) Multi-year Satellite Observations of Sulfur Dioxide Gas Emissions and Lava Extrusion at Bagana Volcano, Papua New Guinea. Front Earth Sci 7:. https://doi.org/10.3389/feart.2019.00009\u003c/li\u003e\n \u003cli\u003eMcCormick Kilbride BT, Nicholson EJ, Wood KT, et al (2023) Temporal Variability in Gas Emissions at Bagana Volcano Revealed by Aerial, Ground, and Satellite Observations. Geochem Geophys Geosystems 24:e2022GC010786. https://doi.org/10.1029/2022GC010786\u003c/li\u003e\n \u003cli\u003eMcKee CO (1981) Geomorphology, geology and petrology of Manam volcano. Cooke-Ravian Vol Volcanol Pap 23\u0026ndash;38\u003c/li\u003e\n \u003cli\u003eMcKee K, Smith CM, Reath K, et al (2021) Evaluating the state-of-the-art in remote volcanic eruption characterization Part II: Ulawun volcano, Papua New Guinea. J Volcanol Geotherm Res 420:107381. https://doi.org/10.1016/j.jvolgeores.2021.107381\u003c/li\u003e\n \u003cli\u003eMeerdink SK, Hook SJ, Roberts DA, Abbott EA (2019) The ECOSTRESS spectral library version 1.0. Remote Sens Environ 230:111196. https://doi.org/10.1016/j.rse.2019.05.015\u003c/li\u003e\n \u003cli\u003eMoussallam Y, Bani P, Curtis A, et al (2016) Sustaining persistent lava lakes: Observations from high-resolution gas measurements at Villarrica volcano, Chile. Earth Planet Sci Lett 454:237\u0026ndash;247. https://doi.org/10.1016/j.epsl.2016.09.012\u003c/li\u003e\n \u003cli\u003eMueller SB, Varley NR, Kueppers U, et al (2013) Quantification of magma ascent rate through rockfall monitoring at the growing/collapsing lava dome of Volc\u0026aacute;n de Colima, Mexico. Solid Earth 4:201\u0026ndash;213. https://doi.org/10.5194/se-4-201-2013\u003c/li\u003e\n \u003cli\u003ePalfreyman W, Cooke R (1976) Eruptive history of Manam volcano, Papua New Guinea. Volcanism Australas 201\u0026ndash;210\u003c/li\u003e\n \u003cli\u003ePalma JL, Blake S, Calder ES (2011) Constraints on the rates of degassing and convection in basaltic open-vent volcanoes. Geochem Geophys Geosystems 12:. https://doi.org/10.1029/2011GC003715\u003c/li\u003e\n \u003cli\u003ePatrick M, Swanson D, Orr T (2019) A review of controls on lava lake level: insights from Halema\u0026lsquo;uma\u0026lsquo;u Crater, Kīlauea Volcano. Bull Volcanol 81:13. https://doi.org/10.1007/s00445-019-1268-y\u003c/li\u003e\n \u003cli\u003ePatrick MR, Houghton BF, Anderson KR, et al (2020) The cascading origin of the 2018 Kīlauea eruption and implications for future forecasting. Nat Commun 11:5646. https://doi.org/10.1038/s41467-020-19190-1\u003c/li\u003e\n \u003cli\u003ePetrone CM, Mollo S, Gertisser R, et al (2022) Magma recharge and mush rejuvenation drive paroxysmal activity at Stromboli volcano. Nat Commun 13:7717. https://doi.org/10.1038/s41467-022-35405-z\u003c/li\u003e\n \u003cli\u003ePlanck M (1901) On the Law of Distribution of Energy in the Normal Spectrum. Ann Phys 4:1\u0026ndash;11\u003c/li\u003e\n \u003cli\u003ePlank S, Walter T, Martinis S, Cesca S (2020) Multi-sensor satellite imagery analysis of the growth and collapse of a littoral lava dome during the 2018/19 eruption of Kadovar Volcano, Papua New Guinea. 22:1423\u003c/li\u003e\n \u003cli\u003ePlank S, Walter TR, Martinis S, Cesca S (2019) Growth and collapse of a littoral lava dome during the 2018/19 eruption of Kadovar Volcano, Papua New Guinea, analyzed by multi-sensor satellite imagery. J Volcanol Geotherm Res 388:106704. https://doi.org/10.1016/j.jvolgeores.2019.106704\u003c/li\u003e\n \u003cli\u003ePyle DM, Mather TA, Biggs J (2013) Remote sensing of volcanoes and volcanic processes: integrating observation and modelling \u0026ndash; introduction. Geol Soc Lond Spec Publ 380:1\u0026ndash;13. https://doi.org/10.1144/SP380.14\u003c/li\u003e\n \u003cli\u003eRobin C, Camus G, Gourgaud A (1991) Eruptive and magmatic cycles at Fuego de Colima volcano (Mexico). J Volcanol Geotherm Res 45:209\u0026ndash;225. https://doi.org/10.1016/0377-0273(91)90060-D\u003c/li\u003e\n \u003cli\u003eRose WI, Palma J, Delgado Granados H, Varley N (2013) Open-vent volcanism and related hazards: Overview. Spec Pap Geol Soc Am 498:vii\u0026ndash;xiii. https://doi.org/10.1130/2013.2498(00)\u003c/li\u003e\n \u003cli\u003eRothery DA, Francis PW, Wood CA (1988) Volcano monitoring using short wavelength infrared data from satellites. J Geophys Res Solid Earth 93:7993\u0026ndash;8008. https://doi.org/10.1029/JB093iB07p07993\u003c/li\u003e\n \u003cli\u003eShevchenko AV, Dvigalo VN, Walter TR, et al (2020) The rebirth and evolution of Bezymianny volcano, Kamchatka after the 1956 sector collapse. Commun Earth Environ 1:1\u0026ndash;15. https://doi.org/10.1038/s43247-020-00014-5\u003c/li\u003e\n \u003cli\u003eShinohara H (2008) Excess degassing from volcanoes and its role on eruptive and intrusive activity. Rev Geophys 46:\u003c/li\u003e\n \u003cli\u003eShinohara H (2005) A new technique to estimate volcanic gas composition: plume measurements with a portable multi-sensor system. J Volcanol Geotherm Res 143:319\u0026ndash;333. https://doi.org/10.1016/j.jvolgeores.2004.12.004\u003c/li\u003e\n \u003cli\u003eSparks RSJ (1997) Causes and consequences of pressurisation in lava dome eruptions. Earth Planet Sci Lett 150:177\u0026ndash;189. https://doi.org/10.1016/S0012-821X(97)00109-X\u003c/li\u003e\n \u003cli\u003eSparks RSJ, Biggs J, Neuberg JW (2012) Monitoring volcanoes. Science 335:1310\u0026ndash;1311\u003c/li\u003e\n \u003cli\u003eStefan J (1879) Uber die Beziehung zwischen der Warmestrahlung und der Temperatur, Sitzungsberichte der mathematisch-naturwissenschaftlichen Classe der kaiserlichen. Akad Wiss 79:S-391\u003c/li\u003e\n \u003cli\u003eStix J, Torres RC, Narv\u0026aacute;ez M L, et al (1997) A model of vulcanian eruptions at Galeras volcano, Colombia. J Volcanol Geotherm Res 77:285\u0026ndash;303. https://doi.org/10.1016/S0377-0273(96)00100-X\u003c/li\u003e\n \u003cli\u003eTaracs\u0026aacute;k Z, Longpr\u0026eacute; M-A, Tart\u0026egrave;se R, et al (2022) Highly Oxidising Conditions in Volatile-Rich El Hierro Magmas: Implications for Ocean Island Magmatism. J Petrol 63:egac011. https://doi.org/10.1093/petrology/egac011\u003c/li\u003e\n \u003cli\u003eTheys N, Hedelt P, De Smedt I, et al (2019) Global monitoring of volcanic SO2 degassing with unprecedented resolution from TROPOMI onboard Sentinel-5 Precursor. Sci Rep 9:2643. https://doi.org/10.1038/s41598-019-39279-y\u003c/li\u003e\n \u003cli\u003eVergniolle S, M\u0026eacute;trich N (2021) Open-vent volcanoes: a preface to the special issue. Bull Volcanol 83:29, s00445-021-01454\u0026ndash;3. https://doi.org/10.1007/s00445-021-01454-3\u003c/li\u003e\n \u003cli\u003eVergniolle S, M\u0026eacute;trich N (2022) An interpretative view of open-vent volcanoes. Bull Volcanol 84:83. https://doi.org/10.1007/s00445-022-01581-5\u003c/li\u003e\n \u003cli\u003eWitter JB, Kress VC, Delmelle P, Stix J (2004) Volatile degassing, petrology, and magma dynamics of the Villarrica Lava Lake, Southern Chile. J Volcanol Geotherm Res 134:303\u0026ndash;337. https://doi.org/10.1016/j.jvolgeores.2004.03.002\u003c/li\u003e\n \u003cli\u003eWoitischek J, Edmonds M, Woods AW (2020) The control of magma crystallinity on the fluctuations in gas composition at open vent basaltic volcanoes. Sci Rep 10:14862. https://doi.org/10.1038/s41598-020-71667-7\u003c/li\u003e\n \u003cli\u003eWood M, Foale S, Gabriel J (2019) Anticipating Ulawun Volcano in New Britain, Papua New Guinea. Anthropol Forum 1\u0026ndash;12. https://doi.org/10.1080/00664677.2019.1647831\u003c/li\u003e\n \u003cli\u003eWoodhead J, Hergt J, Sandiford M, Johnson W (2010) The big crunch: Physical and chemical expressions of arc/continent collision in the Western Bismarck arc. J Volcanol Geotherm Res 190:11\u0026ndash;24. https://doi.org/10.1016/j.jvolgeores.2009.03.003\u003c/li\u003e\n \u003cli\u003eWooster M, Kaneko T (1998) Satellite thermal analyses of lava dome effusion rates at Unzen Volcano, Japan. J Geophys Res Solid Earth 103:20935\u0026ndash;20947\u003c/li\u003e\n \u003cli\u003eWooster MJ, Zhukov B, Oertel D (2003) Fire radiative energy for quantitative study of biomass burning: derivation from the BIRD experimental satellite and comparison to MODIS fire products. Remote Sens Environ 86:83\u0026ndash;107. https://doi.org/10.1016/S0034-4257(03)00070-1\u003c/li\u003e\n \u003cli\u003eWright R (2016) MODVOLC: 14 years of autonomous observations of effusive volcanism from space. Geol Soc Lond Spec Publ 426:23\u0026ndash;53. https://doi.org/10.1144/SP426.12\u003c/li\u003e\n \u003cli\u003eWright R, Blackett M, Hill‐Butler C (2015) Some observations regarding the thermal flux from Earth\u0026rsquo;s erupting volcanoes for the period of 2000 to 2014. Geophys Res Lett 42:282\u0026ndash;289. https://doi.org/10.1002/2014GL061997\u003c/li\u003e\n \u003cli\u003eWright R, Flynn LP, Garbeil H, et al (2004) MODVOLC: near-real-time thermal monitoring of global volcanism. J Volcanol Geotherm Res 135:29\u0026ndash;49. https://doi.org/10.1016/j.jvolgeores.2003.12.008\u003c/li\u003e\n \u003cli\u003eYoung S, Voight B, Duffell H (2003) Magma extrusion dynamics revealed by high-frequency gas monitoring at Soufriere Hills Volcano, Montserrat. Geol Soc Lond Spec Publ 213:219\u0026ndash;230\u003c/li\u003e\n \u003cli\u003e(2021a) Global Volcanism Program | Manam. In: Smithson. Inst. Glob. Volcanism Program. https://volcano.si.edu/volcano.cfm?vn=251020. Accessed 9 Feb 2020\u003c/li\u003e\n \u003cli\u003e(2021b) Global Volcanism Program | Kadovar. https://volcano.si.edu/volcano.cfm?vn=251002. Accessed 8 May 2021\u003c/li\u003e\n \u003cli\u003e(2021c) Global Volcanism Program | Ulawun. In: Smithson. Inst. Glob. Volcanism Program. https://volcano.si.edu/volcano.cfm?vn=252120. Accessed 8 May 2021\u003c/li\u003e\n \u003cli\u003e(2022) MIROVA. https://www.mirovaweb.it/?action=volcanoDetails\u0026amp;volcano_id=251020#lrp. Accessed 27 May 2021\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 and 2 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bulletin-of-volcanology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"buvo","sideBox":"Learn more about [Bulletin of Volcanology](http://link.springer.com/journal/445)","snPcode":"445","submissionUrl":"https://www.editorialmanager.com/buvo/default2.aspx","title":"Bulletin of Volcanology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"magma recharge, open vent, remote sensing, thermal anomalies, SO2 degassing","lastPublishedDoi":"10.21203/rs.3.rs-3903120/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3903120/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eManam is one of the most frequently active volcanoes in Papua New Guinea and is a top contributor to global volcanic volatile emissions due to its persistent open vent degassing. Here, we present a multi-year time series (2018-2021) of thermal and SO\u003csub\u003e2\u003c/sub\u003e emissions for Manam from satellite remote sensing, which we interpret in the context of open vent feedbacks between magma supply, reservoir pressure, and outgassing. We classify the time series into four phases based on the varying SO\u003csub\u003e2\u003c/sub\u003e flux and observe a transient, yet substantial, increase in time-averaged SO\u003csub\u003e2\u003c/sub\u003e flux from background levels of ~0.6 kt day\u003csup\u003e-1\u003c/sup\u003e to ~4.72 kt day\u003csup\u003e-1\u003c/sup\u003e between March and July 2019. We also identify a transition from temporally-coupled to decoupled gas and thermal emissions during this period which we explain in the context of a magma recharge event that supplied new, volatile-rich magma to the shallow plumbing system beneath Manam. We infer that the arrival of this recharge magma triggered the series of eruptions between August 2018 and March 2019. These explosive events collectively removed 0.18 km\u003csup\u003e3 \u003c/sup\u003eof degassed residual magma and signalled the onset of a renewed period of unrest that ultimately culminated in a major eruption on 28 June 2019. We quantify the magnitude of “excess” degassing at Manam after the removal of the inferred residual magma. SO\u003csub\u003e2\u003c/sub\u003e emissions reveal that ~0.18 km\u003csup\u003e3\u003c/sup\u003e of magma was supplied but only ~0.08km\u003csup\u003e3\u003c/sup\u003e was erupted between April 2019 and December 2021. We highlight how multi-parameter remote sensing observations over months to years enables interpretation of open vent processes that may be missed by short duration campaign measurements.\u003c/p\u003e","manuscriptTitle":"Magma recharge at Manam volcano, Papua New Guinea, identified through thermal and SO2 satellite remote sensing of open vent emissions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-03-04 18:41:39","doi":"10.21203/rs.3.rs-3903120/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2024-03-04T17:07:09+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-02-27T12:55:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Bulletin of Volcanology","date":"2024-01-26T07:29:01+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bulletin-of-volcanology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"buvo","sideBox":"Learn more about [Bulletin of Volcanology](http://link.springer.com/journal/445)","snPcode":"445","submissionUrl":"https://www.editorialmanager.com/buvo/default2.aspx","title":"Bulletin of Volcanology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"a2afbbf8-465b-4fef-93c8-f82eaedc5471","owner":[],"postedDate":"March 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-10-14T16:06:44+00:00","versionOfRecord":{"articleIdentity":"rs-3903120","link":"https://doi.org/10.1007/s00445-024-01772-2","journal":{"identity":"bulletin-of-volcanology","isVorOnly":false,"title":"Bulletin of Volcanology"},"publishedOn":"2024-10-07 15:58:05","publishedOnDateReadable":"October 7th, 2024"},"versionCreatedAt":"2024-03-04 18:41:39","video":"","vorDoi":"10.1007/s00445-024-01772-2","vorDoiUrl":"https://doi.org/10.1007/s00445-024-01772-2","workflowStages":[]},"version":"v1","identity":"rs-3903120","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3903120","identity":"rs-3903120","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}