Change in tropical cyclone size over the western North Pacific

Change in tropical cyclone size over the western North Pacific | 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 Article Change in tropical cyclone size over the western North Pacific Yalan Zhang, Yixuan Shen, Yuan Sun, Wei Zhong, Zhou Meng, Zhihao Feng This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6898595/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Tropical cyclone (TC) size is a key factor in determining TC destructiveness and a major challenge in understanding changes in TCs. Although much effort has been devoted to investigating TC track and intensity changes, relatively few studies have focused on TC size changes, particularly their responses to natural variability and global warming. Here, we use a metric of TC size in which TCs achieve their lifetime-maximum intensity, which is relatively insensitive to uncertainty in past data. The results show that the Interdecadal Pacific Oscillation (IPO) dominates TC size variability in the western North Pacific, which is the most active region for TCs. Moreover, TC size variability is governed primarily by radial sea surface temperature (SST) gradients rather than absolute SST values. This finding explains not only the difference in TC size between IPO positive and negative years but also the stronger correlation between the IPO and TC size than other climate indices, such as El Niño–Southern Oscillation (ENSO), since the IPO is derived from SST differences and thus gradients, whereas ENSO is determined by absolute SST values. This finding further implies that the pattern rather than the magnitude of the SST change will determine the change in TC size under future global warming. Earth and environmental sciences/Climate sciences/Climate change Earth and environmental sciences/Climate sciences Earth and environmental sciences/Ocean sciences Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction Tropical cyclones (TCs) are among the most severe natural disasters worldwide 1 , 2 , and the western North Pacific (WNP) region is one of the major TC-affected basins globally 3 . TC size has become an increasingly important parameter of concern, as larger TCs can cause more severe wind damage and greater storm surges, posing a greater threat to life and property 4 , 5 , 6 . In the past, the lack of oceanic observational data led to a scarcity of information on TC size. In recent years, as satellite remote sensing technology has advanced, the quality of TC size data has gradually improved 7 , 8 , 9 . The definition of TC size is not unique. The definition of TC size includes the radius of maximum wind (RMW), the radius of strong winds (17 m s − 1 or 30 kt), the radius of closed isobars, and other related parameters 10 , 11 . Since strong winds exceeding 17 m s − 1 or 30 kt are associated with significant destructive power, many organizations, such as the Japan Meteorological Agency (JMA), record the radius of 30 kt wind (R30) in the best track data 12 , 13 . This study places greater emphasis on TC size, which is closely associated with the destructive potential of TCs. Many scholars have focused on the climatological characteristics of the WNP TC size. The WNP TC size is related to atmospheric circulation patterns, according to QuikSCAT wind field data, and the TC size tends to be smaller during La Niña years and larger during El Niño years 14 . The TC size data (17 m s − 1 wind radius) from the National Centers for Environmental Prediction (NCEP) reanalysis for the period of 1980–2010 indicate that the interannual variation in the WNP TC size is significantly correlated with the El Niño‒Southern Oscillation (ENSO) index 8 . The analysis of the RMW indicate a decreasing trend from 1980 to 2016 15 . Using the TC size (radius of 34 kt wind) derived from ERA5 reanalysis data, the long-term trend is generally not obvious, but the interannual variation is apparent in the WNP region from 1979 to 2019 16 . Moreover, the TC size, in terms of the radius of 64 kt winds from HURSAT v3 from 1995 to 2011, similarly exhibited no discernible long-term trend 17 . The TC size is influenced by many factors, including its track, the activity of the subtropical high ridge, and sea surface temperature (SST). Against the background of global warming, analyses of the long-term variation in TC size are often linked to changes in SST. Numerous studies have noted that TC size is highly sensitive to changes in SST, and a warming SST favors an increase in TC size 18 , 19 , 20 . However, these studies are all case studies based on numerical models, making it difficult to distinguish whether the sensitivity of TC size to SST is driven by global warming or natural variability. TC activity exhibits contradictory trends over different periods, indicating that it is more likely influenced by decadal variability rather than long-term trends 21 . Natural variability can influence SSTs by modulating the interactions between the ocean and the atmosphere. The natural variability in SST (i.e., ENSO, Pacific Decadal Oscillation (PDO), and Interdecadal Pacific Oscillation (IPO)) has been extensively studied and shown to be closely related to the formation and characteristics of TCs 22 , 23 , 24 , 25 , 26 , 27 . Among these natural variabilities, the ENSO index reflects the absolute value of the SST (absolute SST) in a region, whereas the IPO index is more indicative of the relative differences in SST across different regions. Sun et al. noted that, compared with the absolute SST differences in the region of TC formation, the changes in TC intensity and the RMW are more sensitive to the differences in the radial SST distribution (e.g., the difference in SST between the TC inner and outer regions, referred to as the relative SST) 28 . To date, few studies have used long-term TC size data to investigate changes in TC size and influential factors, and no research has thoroughly explored the potential relationship between the TC size variation and natural variability. To address this issue further, in this study, data on the TC size obtained from the widely recognized and authoritative JMA are used, its variations and their associations with natural variability, particularly the IPO, are explored, and potential explanations for the underlying causes are provided. 2. Results 2.1 Variation in the WNP TC size and its relationship with natural variability In this study, first, the relationship between the TC size (R30 LMI − annual ) corresponding to the moment of lifetime maximum intensity (LMI) in the JMA data from 1980 to 2023 and the variations in natural variability (i.e., ENSO, IPO, and PDO) are investigated, as shown in Fig. 1 . The patterns of variation between the three natural variabilities and TC size are very similar, particularly at some peak points (e.g., 1982, 1995, 2002, and 2009). Note that although the IPO reflects primarily low-frequency changes at the decadal scale, the IPO index used here is the unfiltered IPO index (IPO unfilt ) and thus still contains interannual variability. For the subsequent analysis of the interdecadal variation in the TC size, the filtered IPO index is used. Further calculation of the correlation coefficients between the time series of R30 LMI − annual and the three natural variabilities reveals that the correlation between R30 LMI − annual and the IPO unfilt index is the strongest, reaching 0.53, and passes the 99.99% confidence level test. The correlation between R30 LMI − annual and ENSO is the next highest, at 0.47, and passes the 99.9% confidence level test. The correlation between R30 LMI − annual and the PDO is the smallest, at only 0.33, and passes only the 95% confidence level test. The above results indicate that the variation in R30 LMI − annual is related to three natural variabilities (i.e., ENSO, PDO, and IPO unfilt ). Notably, according to the JMA data, the TC size exhibits a significant decreasing trend over the past four decades. To validate the robustness of the results, we further select the Joint Typhoon Warning Center (JTWC) and China Meteorological Administration (CMA) data from International Best Track Archive for Climate Stewardship (IBTrACS) for comparison. Unfortunately, a significant declining trend is not observed in either dataset, which may be attributed to the relatively short length of the data records (CMA data span from 1980 to 2020, and JTWC data span from 2000 to 2023) (Fig. S1 ). In fact, our main concern is not the declining trend with high uncertainty because of the large difference in trends among different agencies (i.e., JMA, CMA, and JTWC). This study focuses more on the underlying causes of this change, specifically the main factors that have influenced the variation in the TC size over the past few decades. To further explore the relationship between the change in TC size in the WNP and three different natural variabilities (i.e., ENSO, PDO, and IPO), we apply a regression method 29 , regressing the R30 LMI − annual time series onto the ENSO time series (i.e., Niño-3.4), PDO, and IPO unfilt indices, to decouple the influence of each natural variability on the TC size. The results show that the residual variations after removing the effects of ENSO and the PDO (Fig. 2 a and 2 c) are largely consistent with the trend of the original R30 LMI − annual time series, indicating that the contributions of ENSO and the PDO to TC size changes are limited. For the IPO unfilt , the residual after removing the influence of the IPO (Fig. 2 b) shows a significant change compared with the original R30 LMI − annual time series. Its linear trend weakens from − 18.56 km decade⁻¹ to -7.46 km decade⁻¹, and this trend is not statistically significant (p > 0.05), indicating that the IPO unfilt has a significant impact on the change in TC size. Nevertheless, the residuals after simultaneously removing the effects of ENSO, PDO, and the IPO unfilt (Fig. 2 d) are highly consistent with the results obtained by removing only the IPO unfilt (with a trend of -10.55 km decade⁻¹), further confirming that the IPO unfilt plays a dominant role in determining the interannual variation in the WNP TC size. To further investigate the potential connection between the TC size (R30 LMI − annual ) and natural variability (e.g., ENSO, PDO, and IPO, which are all SST-related indices), we calculated the distribution of the spatial correlation between the R30 LMI−annual and global annual July–November averaged SSTs from the ERA5 reanalysis from 1980 to 2023, as shown in Fig. 3 . The results indicate that the SST change in the Pacific is a key factor influencing the R30 LMI−annual change in the WNP TC over the past few decades. The R30 LMI−annual time series shows a significant negative correlation with the SST in most regions of the WNP and the western South Pacific but a significant positive correlation with the SST in the tropical central eastern Pacific. This spatial distribution closely resembles the distribution of the spatial correlation between the SST and traditional IPO index (Fig. 1 in ref. 30), further confirming the close relationship between the IPO and WNP TC size. 2.2 Possible reasons for the changes in the WNP TC size To capture the decadal characteristics of changes in the TC size and explore the potential relationship between the IPO and R30 LMI − annual , we apply a 12-year Chebyshev lowpass filter to both the IPO unfilt and R30 LMI − annual data. The time series after filtering is adjusted to span from 1986 to 2017. Importantly, the IPO unfilt data reflect interannual variability, whereas the filtered IPO (IPO filt ) data capture interdecadal characteristics. We sort the 32-year July–November average IPO filt values in descending order and define the highest 25% (8 years) as years of positive IPO phases and the lowest 25% (8 years) as years of negative IPO phases, with the remaining years defined as neutral years (Fig. 4 ). Specifically, the years 1986 and 1990–1996 are categorized as years of positive IPO phases, whereas 1999–2000 and 2007–2012 are categorized as years of negative IPO phases. The changes in R30 LMI − annual and IPO are highly consistent, as R30 LMI − annual is typically larger (smaller) in years of positive (negative) IPO phases. Specifically, during the eight positive IPO years, seven present above-average R30 LMI − annual values, with six consecutive years (1991–1996) ranking in the highest 25%. In contrast, six of the eight negative IPO years present below-average R30 LMI − annual values, including five years (2007–2011) in the lowest 25% (Fig. 4 ). This result indicates that the IPO filt plays a critical role in determining the interdecadal variation in the WNP TC size. The TC size is expected to be affected by three factors: TC intensity (maximum wind speed (MWS)) 31 , 32 , RMW 33 , and the rate at which the 10 m tangential wind speed decreases radially outward from the TC 34 . On the basis of ERA5 6-hourly reanalysis data, we distinguish between IPO positive and negative years and calculate the azimuthally averaged SST, sea level pressure (SLP), and radial distribution profile of the 10 m tangential wind at the time when each TC reaches its LMI (Fig. 5 ). In the IPO-positive years, the radial distributions of SSTs for TCs exhibit minimal differences between the outer SST and inner SST, with the outer SST being slightly warmer than the inner SST (Fig. 5 a). In contrast, during the IPO-negative years, the radial SST differences are notable, and the outer SST is notably colder than the inner SST, indicating a larger radial SST gradient. The differences in the SST distributions directly affect the radial SLP gradient, as warmer local SSTs tend to induce stronger local convections and thus stronger local pressure gradients, including local SLP gradients 28 . Previous studies have indicated that, owing to more (less) energy exchanges at the air‒sea interface, the local warmer (colder) SST tends to produce a larger magnitude of local latent heating 35 , 36 , which results in a greater decrease in local pressure through hydrostatic adjustment 37 , 38 , thus causing a larger (less) radial pressure gradient and leading to an increase (decrease) in local wind speed 39 . As a result, compared with that in the negative IPO years, the warmer SST in the outer region in the positive IPO years contributes to the stronger SLP gradient in the outer region (Fig. 5 b). This contribution further leads to stronger surface wind speeds in the outer region in IPO-positive years than in IPO-negative years, as the pressure gradient is closely associated with the wind speed, according to the gradient wind balance theory 40 (Fig. 5 c). Note that ERA5 systematically underestimates the wind speed and thus TC intensity 7 , 41 , 42 , but this systematic bias is not expected to affect our comparative analysis of outer-wind differences between IPO-positive and -negative years (Fig. 5 c). The difference in the TC outer wind speed directly determines the difference in the TC size between the IPO-positive years and the IPO-negative years. Furthermore, although the SST during the positive IPO years is generally lower than that during the negative years, the MWS and RMW do not significantly weaken, and the TC outer wind speed is even notably higher than that during the negative IPO years. This explains the differences in the TC size between the IPO-positive and -negative years. The impacts of SST differences between the positive and negative IPO years on TC wind profiles can be further analyzed from the following two aspects. First, the TC intensity (i.e., MWS at the surface) is relatively insensitive to the SST change when the SST is high (Fig. 5 c). Specifically, there is a case-dependent critical SST threshold for the development of a TC since the TC intensity increases notably (slightly) with the SST when the SST is below (above) the threshold 14 , 28 , 32 . In this study, as the time-averaged SSTs in both IPO-positive years and -negative years are high (i.e., > 29°C), the small difference in the absolute SST (i.e., < 0.5°C) does not lead to a marked difference in TC intensity (i.e., MWS) between IPO-positive years and -negative years (Fig. 5 a). Second, the radial distribution of the TC surface wind speed is affected by that of the SST. Differences in the radial distribution of the SST lead to variations in the SLP gradient, which in turn cause differences in the tangential wind speed (Fig. 5 b and c). Although the average SST in the negative IPO year is lower than that in the positive IPO year, in terms of the radial distribution, the SSTs in the TC outer regions of the positive (negative) IPO year are significantly lower (slightly higher) than those in the inner regions. This difference in SST between the TC inner and outer regions (relative SST) is a key factor contributing to the radial distribution of TC winds, particularly the greater tangential wind speed in the outer regions of TCs during the positive phase than during the negative phase. Overall, under high-SST conditions, the contrast between the relative SST, rather than the absolute value of the SST (absolute SST), plays an important role in determining TC size. This may also explain why the unfiltered IPO index has a greater correlation with TC size than the ENSO index does, as the former is determined by SST differences, which are similar to the relative SST, whereas the latter, the Nino 3.4 index, is determined by the absolute SST. 3. Conclusion and discussion In this study, the changes in the TC size in the WNP over the past few decades and the possible reasons for these changes are investigated. In this study, the TC size is defined as the radius of the strong wind (i.e., R30), and data from the moment when the TC reaches its maximum intensity are used to reduce uncertainty during the weaker intensity stages. The results indicate that natural variability has been the main driver of these significant changes and the cause of the possible downward trend in TC size over the past few decades. Among the three natural variabilities—ENSO, IPO, and PDO—the IPO has the most important impact on TC size changes. In this study, the impact of positive and negative IPO years on TC size (R30 LMI − annual ) is analyzed, and the results indicate that the change in the R30 LMI − annual is closely aligned with the IPO phase: during positive IPO years, the R30 LMI − annual is typically greater, whereas during negative years, the R30 LMI − annual is smaller. Further research indicates that although the difference in the absolute value of the SST (absolute SST) between the positive and negative IPO years is small (< 0.5°C), the radial distribution of the SST (relative SST) is significant, leading to noticeable variations in the radial distribution of the TC wind speed. Specifically, during the negative IPO year, the SST in the outer regions of the TC is significantly lower than that in the core regions, whereas during the positive IPO year, the outer-region SST is slightly higher than that in the core. This contrast in the SST between the inner and outer regions is assumed to affect local energy exchanges and latent heating at the air‒sea interface and thus alter the local pressure and its gradient, ultimately resulting in significantly higher wind speeds in the outer regions of TCs during the positive IPO year than during the negative IPO year. Additionally, the results of this study revealed that under high-SST conditions (> 29°C), the sensitivity of TC intensity (such as the MWS) to SST changes is relatively low, whereas the variation in the TC size depends primarily on the radial contrast in SST (relative SST) rather than the absolute value of SST. Compared with the ENSO index, which is determined by the absolute value of the SST, the IPO index is based on SST differences (relative SST). This mechanism aligns with the primary factors driving changes in TC size, thereby explaining why the correlation between the IPO index and TC size is significantly greater than that with the ENSO index. In summary, the radial distribution of the SST and the contrast between the inner and outer regions play key roles in regulating the distributions of TC wind speeds and size variations. This finding enhances our understanding of the mechanisms driving changes in the TC size and provides new perspectives for future TC activity predictions. In addition, this study focuses on the impact of natural variability on changes in the TC size rather than the contribution of global warming to the trend in changes in the TC size. The role of global warming in the trend of the TC size remains uncertain, primarily due to limitations in the length of high-quality observational records, which restricts our ability to detect robust trends within such relatively short time series. Methods 1. Data Sources and Preprocessing The operational implementation of geostationary satellite monitoring since the 1980s has significantly enhanced the reliability of global tropical cyclone best-track records (Shan et al., 2024). In this study, the TC size information from 1980 to 2023 provided by the JMA in the IBTrACS v4.0 best track dataset 43 are primarily used. The JMA data include the TC position, intensity, and 30-knot wind radius (R30; units: km) recorded every 3 hours, with the most recent TC data available up to 2024. To validate the robustness of the decreasing TC size trend in the JMA data, in this study, the best track dataset and analyzed size data from TC records from the JTWC, as well as the CMA (https://tcdata.typhoon.org.cn/tcsize.html) 44 , were selected. The CMA dataset provides the TC position, intensity, and 34-knot wind radius (R34) recorded every 6 hours for the northwestern Pacific region (including the South China Sea) from 1980 to 2020. The JTWC dataset provides the TC position, intensity, and 34-knot wind radius (R34) data recorded every 3 hours from 2001 to 2024. In this study, the 6-hourly reanalysis data from ERA5 (https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview), including variables such as SST, surface wind speed, and monthly average SST, are also used. Importantly, owing to the relatively low resolution of the ERA5 data (0.25°), the simulated core information of TCs may be inaccurate. However, the resolution is sufficient to identify the outer wind speeds of TCs; thus, the distribution of the TC outer wind speed is relatively reliable. Although the ERA5 data may underestimate the overall TC wind speed due to its coarser resolution, it can still be used to explore differences in TC outer wind speeds between cold and warm years (Bian et al., 2021). Additionally, ERA5 significantly underestimates TC wind speed and thus the TC size, but this does not affect our analysis of the relative differences in the TC size between IPO-positive and -negative years. 2. Definition of TC Size Metrics This study considers only northwestern Pacific TCs (excluding those generated in the South China Sea) during the active TC season (July-November), with maximum wind speeds exceeding 30 kt. In this study, the average size during the lifetime of a single TC is used as an indicator of the TC's outer size, specifically R34 AVE or R30 AVE . The method of calculation is as follows: where τ represents the life cycle duration of a specific TC and R ( t ) represents the specific wind radius at a given time for that TC. The annual average of the TC lifetime average size is expressed as follows: where N denotes the total number of TCs in a given year, I refers to the i -th TC in that year, and R AVE ( i ) is the R AVE for the i-th TC. Additionally, in this study, the TC size at the moment of the LMI is used as another indicator of TC size, specifically R34 LMI or R30 LMI . The annual average TC size at the moment of the LMI is expressed as follows: where N represents the total number of TCs in a given year, i refers to the i -th TC in that year, and R LMI ( i ) is the specific wind radius ( R LMI ) at the moment when the i -th TC reaches its LMI. There are significant differences in the annual variations in the lifetime average TC size between different agencies (Fig. S1). This may be because the moment of TC formation is determined when the initial storm intensity reaches a specified threshold. Since a storm’s intensity is relatively low in the early stages of development, accurately estimating its strength is difficult, leading to considerable uncertainty. As a result, there are discrepancies in the recorded locations of TC initiation and life cycle durations across different observational datasets. Additionally, although the TC size is weakly correlated with TC intensity, uncertainties in intensity estimates can also affect TC size information. Therefore, using the average TC size of the life cycle to study characteristics of TC size variation may introduce more uncertainty and error. Moreover, Kim et al. mentioned that differences in the TC size estimates across different agencies may arise because some agencies consider only the storm's wind when estimating the wind radius, whereas others incorporate the effects of other midlatitude weather systems 45 . Compared with the lifetime average TC size, the error in identifying the storm intensity at the moment of the maximum life cycle intensity (LMI) is smaller, and the uncertainty in determining the TC location is reduced. This is because, at the LMI moment, the TC size is generally larger within its lifetime, making it easier to identify. Additionally, the determination of the LMI moment is solely related to the relative intensity of the TC's life cycle. Therefore, although the life cycle intensity recorded in different observational datasets may vary, the LMI moment is less affected by the absolute intensity. Figures S1(d-f) show the trends in the annual average TC size at the moment of maximum intensity (R30 LMI-annual or R34 LMI-annual ) for TCs in the WNP from the CMA, JMA, and JTWC datasets between 2001 and 2020. The values of R34 LMI-annual (and R30 LMI-annual ) are generally slightly larger than those of R34 AVE-annual (and R30 AVE-annual ), but in some years, the former is smaller than the latter. This result is because the TC size and intensity are not strictly linearly related, so in these years, in some TCs, the size at the moment of the LMI is smaller than that at other points in their life cycle. However, the overall trends of both variables are quite similar. In years when the R30 AVE-annual is significantly larger (or smaller) than usual (e.g., 2002, 2010), the R30 LMI-annual also follows a similar pattern, being larger (or smaller) in these years. The correlation coefficients between the three datasets, as shown in Table S1, increased significantly compared with the life cycle average TC size, and the differences in changes in the trends among the three datasets decreased. Therefore, compared with R AVE , R LMI reduces, to some extent, the differences in uncertainty caused by varying data sources. For these reasons, and because the JMA dataset spans a longer period, this study focuses on using the TC size at the moment of maximum life cycle intensity (R30 LMI-annual ) from the JMA dataset to investigate the associated patterns of size variation. 3. Natural variability and Decadal Signal Extraction To analyze the impact of natural variability on changes in the TC size, in this study, the monthly indices of the El Niño‒Southern Oscillation (ENSO) (Niño-3.4), Pacific Decadal Oscillation (PDO), and Tripole Index for the Interdecadal Pacific Oscillation (TPI, which is the index reflecting the IPO) from 1980 to 2023, which are all provided by the Physical Sciences Division of the Earth System Research Laboratory at the National Oceanic and Atmospheric Administration (NOAA) (https://psl.noaa.gov/gcos_wgsp/Timeseries/), are used. The Niño-3.4 index is calculated as the monthly average SST of the region bounded by 5°N-5°S and 170°W-120°W and is based on data from HadISST1. The PDO index is derived from a time series of spatially averaged monthly SSTs in the North Pacific (north of 20°N) and is calculated using the SST time covariance matrix from 1900–1993. The IPO index (i.e., TPI) is defined as the difference between the SST anomaly averaged over the central equatorial Pacific (10°S-10°N, 170°E-90°W) and the average SST anomalies in the WNP (25°N-45°N, 140°E-145°W) and southwestern Pacific (50°S-15°S, 150°E-160°W). The data used here are in unfiltered form, but in the later parts of the analysis, both the IPO and R30 data are low-pass filtered to extract the decadal signals of climate indices. This study follows the IPO filtering method provided by the NOAA website and uses a Chebyshev low-pass filter. The specific parameters are as follows: a 13-point weighting coefficient, a low-frequency cutoff frequency of 1/12, and standard smoothing parameters. This zero-phase filter effectively separates the low-frequency components of natural variability, retaining decadal signal features, such as the IPO, while effectively filtering out higher-frequency interannual fluctuations. In subsequent analyses of the interdecadal characteristics of TC size, to maintain consistency with the filtered IPO index, the R30 index was processed using an identical filtering approach. 4. Statistical Framework for Attribution of the TC Size The annual average ENSO, PDO, and IPO indices mentioned in this study are based on July to November averages, which is a period considered the active season for TCs in the WNP. To quantitatively analyze the independent effects of natural variability on changes in the TC size, in this study, a linear regression framework is employed, starting from the time series of the TC size: where X(t) represents the time series of standardized indices for each climate mode (which may also refer to multiple natural variabilities) and where is the TC size regression residual after removing the linear influence of the corresponding mode. The statistical significance of each regression coefficient ( ) is rigorously evaluated using a two-tailed Student's t test (confidence level of 95%, α = 0.05). Declarations Author Contribution Conceptualization: YS, YZ, WZ, ZMMethodology: YS, YS, YZInvestigation: YZ, YS, ZFVisualization: YZSupervision: YS, WZ, ZMWriting—original draft: YS, YZWriting—review & editing: YS, YZ Data Availability All data, code, and materials used in the analyses are publicly available through Figshare (DOI: 10.6084/m9.figshare.29219018.v1). This dataset includes [briefly describe the core contents, e.g., 'tropical cyclone size measurements over the western North Pacific from 1980-2020, along with associated metadata and analysis scripts']. No special restrictions apply to these materials beyond the CC BY 4.0 license terms. The raw data supporting the conclusions are provided in full within this repository, and processed data are available in the main text or supplementary materials of this article. The authors will make any additional methodological details available upon reasonable request. References Peduzzi, P. et al. O. Global trends in tropical cyclone risk. Nat. Clim. Change 2, 289–294 (2012). Jing, R. et al. Global population profile of tropical cyclone exposure from 2002 to 2019. Nature. 626, 549–554 (2024). Kim, H.-K. & Seo, K.-H. Cluster analysis of tropical cyclone tracks over the western North Pacific using a self-organizing map. J. Clim. 29, 3731–3751 (2016). Rappaport, E. N. Loss of life in the United States associated with recent Atlantic tropical cyclones: bulletin of the American meteorological society. Bull. Am. Meteorol. Soc. 81, 2065–2073 (2000). Chen, L., Ma, Z. & Fei, J. How different is tropical cyclone precipitation over land and ocean?. npj Clim Atmos Sci. 8, 70 (2025). Emanuel, K. Increasing destructiveness of tropical cyclones over the past 30 years. Nature. 436, 686-688 (2005). Bian, G. F., Nie, G. Z. & Qiu, X. How well is outer tropical cyclone size represented in the ERA5 reanalysis dataset? Atmos. Res. 249, 105339 (2021). Mok, D., Chan, J. C. L. & Chan, K. T. F. A 31-year climatology of tropical cyclone size from the NCEP Climate Forecast System Reanalysis. Int. J. Climatol. 38, 5669-5685 (2018). Schenkel, B. A. et al. Evaluating outer tropical cyclone size in reanalysis datasets using QuikSCAT data. J. Clim. 30, 8745-8762 (2017). Merrill, R. T. A comparison of large and small tropical cyclones. Mon. Weather Rev. 112, 1408 (1984). Weatherford, C. L. & Gray, W. M. Typhoon structure as revealed by aircraft reconnaissance. Part I: Data analysis and climatology. Mon. Weather Rev. 116, 1032–1043 (1988). Wu, L. et al. Implications of the observed relationship between tropical cyclone size and intensity over the Western North Pacific. J. Clim. 28, 9501-9506 (2015). Gori, A., Lin, N., Schenkel, B. & Chavas, D. North Atlantic tropical cyclone size and storm surge reconstructions from 1950–present. J. Geophys. Res. Atmos. 128, e2022JD037312 Chan, J. C. L. & Yip, C. K. M. Interannual variations of tropical cyclone size over the western North Pacific. Geophys. Res. Lett. 30, 665-678 (2003). Shou, H. et al. Decreasing trend of western North Pacific tropical cyclone inner-core size over the past decades. J. Meteorol. Res. 35, 635-645 (2021). Zhang, K. & Chan, K. T. F. An ERA5 global climatology of tropical cyclone size asymmetry. Int. J. Climatol. 43, 950-963 (2022). Knaff, J. A., Longmore, S. P. & Molenar, D. A. An objective satellite-based tropical cyclone size climatology. J. Clim. 27, 455-476 (2014). Sun, Y. et al. Impact of ocean warming on tropical cyclone track over the western North Pacific: A numerical investigation based on two case studies. J. Geophys. Res. Atmos. 122, 6476-6492 (2017a). Wang, S. & Toumi, R. Reduced sensitivity of tropical cyclone intensity and size to sea surface temperature in a radiative-convective equilibrium environment. Adv. Atmos. Sci. 35, 981-993 (2018). Xu, Z. et al. Tropical cyclone size change under ocean warming and associated responses of tropical cyclone destructiveness: idealized experiments. J. Meteorol. Res. 34, 163-175 (2020). Shan, K. et al. How does global accumulated tropical cyclone energy vary in response to a changing climate? Sci. Bull. 69, 2053-2064 (2024). Patricola, C. M. et al. The influence of ENSO flavors on western North Pacific tropical cyclone activity. J. Clim. 31, 5395-5416 (2018). Scoccimarro, E. et al. The Pacific Decadal Oscillation modulates tropical cyclone days on the interannual timescale in the North Pacific Ocean. J. Geophys. Res. Atmos. 126, e2021JD034988 (2021). Zhao, J. et al. Untangling impacts of global warming and interdecadal Pacific oscillation on long-term variability of North Pacific tropical cyclone track density. Sci. Adv. 6, eaba6813 (2020). Chan, J. C. L. Tropical cyclone activity over the western North Pacific associated with El Niño and La Niña events. J. Clim. 13, 2960–2972 (2000). Lander, M. A. An exploratory analysis of the relationship between tropical storm formation in the western North Pacific and ENSO. Mon. Weather Rev. 122, 636-651 (1994). Cai, Y. et al. Enhanced predictability of rapidly intensifying tropical cyclones over the western North Pacific associated with snow depth changes over the Tibetan Plateau. J. Clim . 35, 2093–2110 (2022). Sun, Y. et al. The opposite effects of inner and outer sea surface temperature on tropical cyclone intensity. J. Geophys. Res. Atmos . 119, 2193-2208 (2014). Kossin, J., Emanuel, K. & Vecchi, G. The poleward migration of the location of tropical cyclone maximum intensity. Nature. 509, 349–352 (2014). Henley, B.J. et al. A tripole index for the Interdecadal Pacific Oscillation. Clim. Dyn. 45, 3077–3090 (2015). Smith, R. K., Montgomery, M. T., & Schmidt, C. Dynamical constraints on the intensity and size of tropical cyclones. Quarterly Journal of the Royal Meteorological Society 137, 1841 - 1855 (2010). Weatherford, C. L. & Gray, W. M. Typhoon structure as revealed by aircraft reconnaissance. Part II: structural variability. Mon. Weather Rev. 116, 1044-1056 (1988). Ruan, Z. & Wu, Q. Relationship between size and intensity in North Atlantic tropical cyclones with steady radii of maximum wind. Geophys. Res. Lett. 49, e2021GL095632 (2022). Sun, Y. et al. Impact of ocean warming on tropical cyclone size and its destructiveness. Sci. Rep. 7, 8154 (2017). Emanuel, K. A. An air-sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci. 43, 585–604 (1986). Rotunno, R. & Emanuel, K. A. An air-sea interaction theory for tropical cyclones. Part II: Evolutionary study using a nonhydrostatic axisymmetric numerical model. J. Atmos. Sci. 44, 542-561 (1987). Raymond, D. J. Convective processes and tropical atmospheric circulations. Q. J. R. Meteorol. Soc. 120, 1431-1455 (1994). Xu, J. & Wang, Y. Sensitivity of tropical cyclone inner-core size and intensity to the radial distribution of surface entropy flux. J. Atmos. Sci. 67, 1831–1852 (2010). Fierro, A. O., Rogers, R. F. & Marks, F. D. The impact of horizontal grid spacing on the microphysical and kinematic structures of strong tropical cyclones simulated with the WRF-ARW Model. Mon. Weather Rev . 137, 3717-3743 (2009). Brill, K. F. Revisiting an Old Concept: The Gradient Wind*. Mon. Weather Rev . 142, 1460 - 1471 (2014). Dulac, W., Cattiaux, J., Chauvin, F., Bourdin, S. & Fromang, S. Assessing the representation of tropical cyclones in ERA5 with the CNRM tracker. Clim. Dyn. 62, 223–238 (2024). Xu, Z. et al. Global tropical cyclone size and intensity reconstruction dataset for 1959-2022 based on IBTrACS and ERA5 data. Earth Syst. Sci. Data 16, 5753-5766 (2024). Knapp, K. R. et al. International Best Track Archive for Climate Stewardship (IBTrACS) Project, Version 4. NOAA National Centers for Environmental Information (2018). Ying, M. et al. An overview of the China Meteorological Administration tropical cyclone database. J. Atmos. Ocean. Technol. 31, 287-301 (2014). Kim, H.-J., Moon, I.-J. & Oh, I. Comparison of tropical cyclone wind radius estimates between the KMA, RSMC Tokyo, and JTWC. Asia-Pac. J. Atmos. Sci . 58, 563-576 (2022). Additional Declarations No competing interests reported. Supplementary Files supplementarymaterials.docx Cite Share Download PDF Status: Posted Version 1 posted 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-6898595","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":474800996,"identity":"e29d04fb-8033-4194-8d70-abee125d6374","order_by":0,"name":"Yalan Zhang","email":"","orcid":"","institution":"National University of Defense Technology","correspondingAuthor":false,"prefix":"","firstName":"Yalan","middleName":"","lastName":"Zhang","suffix":""},{"id":474800997,"identity":"4275b92e-660c-4325-9ccb-ca89fc12d050","order_by":1,"name":"Yixuan Shen","email":"","orcid":"","institution":"National University of Defense Technology","correspondingAuthor":false,"prefix":"","firstName":"Yixuan","middleName":"","lastName":"Shen","suffix":""},{"id":474800998,"identity":"42804535-d3c7-40f6-b514-6a2f77907414","order_by":2,"name":"Yuan Sun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4ElEQVRIie3PMQrCMBiG4b9EOv3SNVKxV4gIglDwKhEhkxXHDm5CO4jOXbyDRygGnCpdO3RQhE4ddBFcxLo5NR0F824J3wMJgE73w6FFYHLmvtucdDshXNk5Ec2Jy1Kj6FyCg3rqRN7RxmWOIInwuRmDFa55LWHZXNh4LNBYmSLjmANNTvt6QmdDG02JhGBFaFHdePXEiT7kJdEk1mPBmVQTyCrSDiQiAQGcNyAsKQej3VYiJTClPBao/IsTzvpZ+ZDjcRpP7s+X27PCjeJhAC36fULV/BO5NVnpdDrdH/cGXP1Fqmo0MaQAAAAASUVORK5CYII=","orcid":"","institution":"National University of Defense Technology","correspondingAuthor":true,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Sun","suffix":""},{"id":474800999,"identity":"637f5596-a64d-411b-ba62-a1bc4951e165","order_by":3,"name":"Wei Zhong","email":"","orcid":"","institution":"National University of Defense Technology","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Zhong","suffix":""},{"id":474801000,"identity":"e0656c07-634e-4ceb-954e-f50b8a302836","order_by":4,"name":"Zhou Meng","email":"","orcid":"","institution":"National University of Defense Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhou","middleName":"","lastName":"Meng","suffix":""},{"id":474801001,"identity":"2d0f6881-50cf-4143-b762-89933fa1efb1","order_by":5,"name":"Zhihao Feng","email":"","orcid":"","institution":"National University of Defense Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhihao","middleName":"","lastName":"Feng","suffix":""}],"badges":[],"createdAt":"2025-06-15 13:38:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6898595/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6898595/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":85236038,"identity":"4ff02a2d-04f8-4d27-925c-f2a53a54f907","added_by":"auto","created_at":"2025-06-23 16:59:40","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":162467,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRelationship between the TC size and natural variabilities. \u003c/strong\u003eTime series of the WNP TC annual average size (R30\u003csub\u003eLMI-annual\u003c/sub\u003e, km) corresponding to the moment of the LMI in the JMA data from 1980 to 2023 alongside the annual average ENSO index (a; °C), PDO index (b; dimensionless), and IPO\u003csub\u003eunfilt\u003c/sub\u003e index (c; °C). The \u003cem\u003ex\u003c/em\u003e-axis represents the years, the solid red line (left \u003cem\u003ey\u003c/em\u003e-axis) represents the annual average TC size corresponding to the moment of the LMI, the dashed red line represents the linear regression curve for the trend in the TC size variation, and the red shading indicates the 95% two-sided confidence interval of the regression curve.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/1c5a72cc067cc948e85bbd66.png"},{"id":85236039,"identity":"908ade18-4aff-4de3-9d1f-926ea4fb8839","added_by":"auto","created_at":"2025-06-23 16:59:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":100271,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eReduced variations in the WNP TC annual average size with natural variabilities. \u003c/strong\u003eTime series of R30\u003csub\u003eLMI-annual\u003c/sub\u003e calculated from the JMA with (a) ENSO, (b) IPO, (c) PDO, and (d) ENSO, IPO, and PDO variability reduced from 1980 to 2023. The values are calculated from the residuals of the regression of R30\u003csub\u003eLMI-annual\u003c/sub\u003e onto each index. Shading represents the 95% two-sided confidence interval of the trend.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/a6cc53cad9be4e8c8596a336.png"},{"id":85236040,"identity":"b6898271-c141-426d-a9a1-f9536a1c7d71","added_by":"auto","created_at":"2025-06-23 16:59:41","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":263555,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIPO pattern-dependent relationship of the SST–TC size. \u003c/strong\u003eSpatial distribution of the correlation coefficients between the global annual average SST (July to November) from the ERA5 reanalysis dataset and the R30\u003csub\u003eLMI-annual\u003c/sub\u003e data from the JMA dataset corresponding to the moment of the LMI from 1980 to 2023. Different colors represent the magnitude of the correlation coefficients, and black dots indicate regions where the correlation coefficients are statistically significant at the 95% confidence level.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/8fea2c7e18277d45b34e2be1.png"},{"id":85237265,"identity":"4da599fe-43d1-4090-bf81-1b8f6672c22e","added_by":"auto","created_at":"2025-06-23 17:15:41","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":34473,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTC size anomalies modulated by IPO phases. \u003c/strong\u003eFiltered annual IPO (°C) and R30\u003csub\u003eLMI-annual\u003c/sub\u003e anomalies (km) over the WNP from 1986–2017 (the reference for the R30\u003csub\u003eLMI-annual\u003c/sub\u003e anomalies is 388.0 km). The blue bars denote negative IPO years, the red bars denote positive IPO years, and the white bars denote neutral IPO years. The horizontal lines represent the 25th percentile (blue dashed line) and 75th percentile (red dashed line).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/6dc3cb5614bea0c8b1219ce5.png"},{"id":85236422,"identity":"74a66a71-b44f-4e34-94f7-c98d1d3a4bfc","added_by":"auto","created_at":"2025-06-23 17:07:41","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":57183,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMechanism of the effect of the IPO on TC size. \u003c/strong\u003eThe azimuthally averaged radial distributions of (a) SST (°C), (b) SLP gradient, and (c) tangential wind at a height of 10 m for all TCs during the positive and negative IPO years from 1986 to 2017 in the ERA5 reanalysis. The blue curve represents positive IPO years, and the red curve represents negative IPO years. The x-axis represents the distance from the TC center (°).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/b0926ead2103b132f1dc043e.png"},{"id":86315997,"identity":"d7503dde-1d02-48ee-8453-147d850628c0","added_by":"auto","created_at":"2025-07-09 08:54:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1133071,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/c4c83062-0f54-4cdf-b935-48dbf7a04832.pdf"},{"id":85236048,"identity":"aacb2917-b052-4021-831f-dbb677cf3e03","added_by":"auto","created_at":"2025-06-23 16:59:41","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1072462,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-6898595/v1/666a399bbdc67b2feaafc2da.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Change in tropical cyclone size over the western North Pacific","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eTropical cyclones (TCs) are among the most severe natural disasters worldwide\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, and the western North Pacific (WNP) region is one of the major TC-affected basins globally\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. TC size has become an increasingly important parameter of concern, as larger TCs can cause more severe wind damage and greater storm surges, posing a greater threat to life and property\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. In the past, the lack of oceanic observational data led to a scarcity of information on TC size. In recent years, as satellite remote sensing technology has advanced, the quality of TC size data has gradually improved\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. The definition of TC size is not unique. The definition of TC size includes the radius of maximum wind (RMW), the radius of strong winds (17 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e or 30 kt), the radius of closed isobars, and other related parameters\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Since strong winds exceeding 17 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e or 30 kt are associated with significant destructive power, many organizations, such as the Japan Meteorological Agency (JMA), record the radius of 30 kt wind (R30) in the best track data\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. This study places greater emphasis on TC size, which is closely associated with the destructive potential of TCs.\u003c/p\u003e \u003cp\u003eMany scholars have focused on the climatological characteristics of the WNP TC size. The WNP TC size is related to atmospheric circulation patterns, according to QuikSCAT wind field data, and the TC size tends to be smaller during La Ni\u0026ntilde;a years and larger during El Ni\u0026ntilde;o years\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. The TC size data (17 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e wind radius) from the National Centers for Environmental Prediction (NCEP) reanalysis for the period of 1980\u0026ndash;2010 indicate that the interannual variation in the WNP TC size is significantly correlated with the El Ni\u0026ntilde;o‒Southern Oscillation (ENSO) index\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. The analysis of the RMW indicate a decreasing trend from 1980 to 2016\u003csup\u003e15\u003c/sup\u003e. Using the TC size (radius of 34 kt wind) derived from ERA5 reanalysis data, the long-term trend is generally not obvious, but the interannual variation is apparent in the WNP region from 1979 to 2019\u003csup\u003e16\u003c/sup\u003e. Moreover, the TC size, in terms of the radius of 64 kt winds from HURSAT v3 from 1995 to 2011, similarly exhibited no discernible long-term trend\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. The TC size is influenced by many factors, including its track, the activity of the subtropical high ridge, and sea surface temperature (SST). Against the background of global warming, analyses of the long-term variation in TC size are often linked to changes in SST. Numerous studies have noted that TC size is highly sensitive to changes in SST, and a warming SST favors an increase in TC size\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. However, these studies are all case studies based on numerical models, making it difficult to distinguish whether the sensitivity of TC size to SST is driven by global warming or natural variability. TC activity exhibits contradictory trends over different periods, indicating that it is more likely influenced by decadal variability rather than long-term trends\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Natural variability can influence SSTs by modulating the interactions between the ocean and the atmosphere. The natural variability in SST (i.e., ENSO, Pacific Decadal Oscillation (PDO), and Interdecadal Pacific Oscillation (IPO)) has been extensively studied and shown to be closely related to the formation and characteristics of TCs\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Among these natural variabilities, the ENSO index reflects the absolute value of the SST (absolute SST) in a region, whereas the IPO index is more indicative of the relative differences in SST across different regions. Sun et al. noted that, compared with the absolute SST differences in the region of TC formation, the changes in TC intensity and the RMW are more sensitive to the differences in the radial SST distribution (e.g., the difference in SST between the TC inner and outer regions, referred to as the relative SST)\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. To date, few studies have used long-term TC size data to investigate changes in TC size and influential factors, and no research has thoroughly explored the potential relationship between the TC size variation and natural variability. To address this issue further, in this study, data on the TC size obtained from the widely recognized and authoritative JMA are used, its variations and their associations with natural variability, particularly the IPO, are explored, and potential explanations for the underlying causes are provided.\u003c/p\u003e"},{"header":"2. Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Variation in the WNP TC size and its relationship with natural variability\u003c/h2\u003e \u003cp\u003eIn this study, first, the relationship between the TC size (R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e) corresponding to the moment of lifetime maximum intensity (LMI) in the JMA data from 1980 to 2023 and the variations in natural variability (i.e., ENSO, IPO, and PDO) are investigated, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The patterns of variation between the three natural variabilities and TC size are very similar, particularly at some peak points (e.g., 1982, 1995, 2002, and 2009). Note that although the IPO reflects primarily low-frequency changes at the decadal scale, the IPO index used here is the unfiltered IPO index (IPO\u003csub\u003eunfilt\u003c/sub\u003e) and thus still contains interannual variability. For the subsequent analysis of the interdecadal variation in the TC size, the filtered IPO index is used. Further calculation of the correlation coefficients between the time series of R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e and the three natural variabilities reveals that the correlation between R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e and the IPO\u003csub\u003eunfilt\u003c/sub\u003e index is the strongest, reaching 0.53, and passes the 99.99% confidence level test. The correlation between R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e and ENSO is the next highest, at 0.47, and passes the 99.9% confidence level test. The correlation between R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e and the PDO is the smallest, at only 0.33, and passes only the 95% confidence level test. The above results indicate that the variation in R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e is related to three natural variabilities (i.e., ENSO, PDO, and IPO\u003csub\u003eunfilt\u003c/sub\u003e). Notably, according to the JMA data, the TC size exhibits a significant decreasing trend over the past four decades.\u003c/p\u003e \u003cp\u003eTo validate the robustness of the results, we further select the Joint Typhoon Warning Center (JTWC) and China Meteorological Administration (CMA) data from International Best Track Archive for Climate Stewardship (IBTrACS) for comparison. Unfortunately, a significant declining trend is not observed in either dataset, which may be attributed to the relatively short length of the data records (CMA data span from 1980 to 2020, and JTWC data span from 2000 to 2023) (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). In fact, our main concern is not the declining trend with high uncertainty because of the large difference in trends among different agencies (i.e., JMA, CMA, and JTWC). This study focuses more on the underlying causes of this change, specifically the main factors that have influenced the variation in the TC size over the past few decades.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo further explore the relationship between the change in TC size in the WNP and three different natural variabilities (i.e., ENSO, PDO, and IPO), we apply a regression method\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, regressing the R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e time series onto the ENSO time series (i.e., Ni\u0026ntilde;o-3.4), PDO, and IPO\u003csub\u003eunfilt\u003c/sub\u003e indices, to decouple the influence of each natural variability on the TC size. The results show that the residual variations after removing the effects of ENSO and the PDO (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ea and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ec) are largely consistent with the trend of the original R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e time series, indicating that the contributions of ENSO and the PDO to TC size changes are limited. For the IPO\u003csub\u003eunfilt\u003c/sub\u003e, the residual after removing the influence of the IPO (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eb) shows a significant change compared with the original R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e time series. Its linear trend weakens from \u0026minus;\u0026thinsp;18.56 km decade⁻\u0026sup1; to -7.46 km decade⁻\u0026sup1;, and this trend is not statistically significant (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05), indicating that the IPO\u003csub\u003eunfilt\u003c/sub\u003e has a significant impact on the change in TC size. Nevertheless, the residuals after simultaneously removing the effects of ENSO, PDO, and the IPO\u003csub\u003eunfilt\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ed) are highly consistent with the results obtained by removing only the IPO\u003csub\u003eunfilt\u003c/sub\u003e (with a trend of -10.55 km decade⁻\u0026sup1;), further confirming that the IPO\u003csub\u003eunfilt\u003c/sub\u003e plays a dominant role in determining the interannual variation in the WNP TC size.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo further investigate the potential connection between the TC size (R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e) and natural variability (e.g., ENSO, PDO, and IPO, which are all SST-related indices), we calculated the distribution of the spatial correlation between the R30 \u003csub\u003eLMI\u0026minus;annual\u003c/sub\u003e and global annual July\u0026ndash;November averaged SSTs from the ERA5 reanalysis from 1980 to 2023, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The results indicate that the SST change in the Pacific is a key factor influencing the R30 \u003csub\u003eLMI\u0026minus;annual\u003c/sub\u003e change in the WNP TC over the past few decades. The R30 \u003csub\u003eLMI\u0026minus;annual\u003c/sub\u003e time series shows a significant negative correlation with the SST in most regions of the WNP and the western South Pacific but a significant positive correlation with the SST in the tropical central eastern Pacific. This spatial distribution closely resembles the distribution of the spatial correlation between the SST and traditional IPO index (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e in ref. 30), further confirming the close relationship between the IPO and WNP TC size.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Possible reasons for the changes in the WNP TC size\u003c/h2\u003e \u003cp\u003eTo capture the decadal characteristics of changes in the TC size and explore the potential relationship between the IPO and R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e, we apply a 12-year Chebyshev lowpass filter to both the IPO\u003csub\u003eunfilt\u003c/sub\u003e and R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e data. The time series after filtering is adjusted to span from 1986 to 2017. Importantly, the IPO\u003csub\u003eunfilt\u003c/sub\u003e data reflect interannual variability, whereas the filtered IPO (IPO\u003csub\u003efilt\u003c/sub\u003e) data capture interdecadal characteristics. We sort the 32-year July\u0026ndash;November average IPO\u003csub\u003efilt\u003c/sub\u003e values in descending order and define the highest 25% (8 years) as years of positive IPO phases and the lowest 25% (8 years) as years of negative IPO phases, with the remaining years defined as neutral years (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Specifically, the years 1986 and 1990\u0026ndash;1996 are categorized as years of positive IPO phases, whereas 1999\u0026ndash;2000 and 2007\u0026ndash;2012 are categorized as years of negative IPO phases. The changes in R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e and IPO are highly consistent, as R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e is typically larger (smaller) in years of positive (negative) IPO phases. Specifically, during the eight positive IPO years, seven present above-average R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e values, with six consecutive years (1991\u0026ndash;1996) ranking in the highest 25%. In contrast, six of the eight negative IPO years present below-average R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e values, including five years (2007\u0026ndash;2011) in the lowest 25% (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This result indicates that the IPO\u003csub\u003efilt\u003c/sub\u003e plays a critical role in determining the interdecadal variation in the WNP TC size.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe TC size is expected to be affected by three factors: TC intensity (maximum wind speed (MWS))\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, RMW\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, and the rate at which the 10 m tangential wind speed decreases radially outward from the TC\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. On the basis of ERA5 6-hourly reanalysis data, we distinguish between IPO positive and negative years and calculate the azimuthally averaged SST, sea level pressure (SLP), and radial distribution profile of the 10 m tangential wind at the time when each TC reaches its LMI (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003e). In the IPO-positive years, the radial distributions of SSTs for TCs exhibit minimal differences between the outer SST and inner SST, with the outer SST being slightly warmer than the inner SST (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). In contrast, during the IPO-negative years, the radial SST differences are notable, and the outer SST is notably colder than the inner SST, indicating a larger radial SST gradient. The differences in the SST distributions directly affect the radial SLP gradient, as warmer local SSTs tend to induce stronger local convections and thus stronger local pressure gradients, including local SLP gradients\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Previous studies have indicated that, owing to more (less) energy exchanges at the air‒sea interface, the local warmer (colder) SST tends to produce a larger magnitude of local latent heating\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e, which results in a greater decrease in local pressure through hydrostatic adjustment\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, thus causing a larger (less) radial pressure gradient and leading to an increase (decrease) in local wind speed\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. As a result, compared with that in the negative IPO years, the warmer SST in the outer region in the positive IPO years contributes to the stronger SLP gradient in the outer region (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). This contribution further leads to stronger surface wind speeds in the outer region in IPO-positive years than in IPO-negative years, as the pressure gradient is closely associated with the wind speed, according to the gradient wind balance theory\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ec). Note that ERA5 systematically underestimates the wind speed and thus TC intensity\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, but this systematic bias is not expected to affect our comparative analysis of outer-wind differences between IPO-positive and -negative years (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ec). The difference in the TC outer wind speed directly determines the difference in the TC size between the IPO-positive years and the IPO-negative years. Furthermore, although the SST during the positive IPO years is generally lower than that during the negative years, the MWS and RMW do not significantly weaken, and the TC outer wind speed is even notably higher than that during the negative IPO years. This explains the differences in the TC size between the IPO-positive and -negative years. The impacts of SST differences between the positive and negative IPO years on TC wind profiles can be further analyzed from the following two aspects.\u003c/p\u003e \u003cp\u003eFirst, the TC intensity (i.e., MWS at the surface) is relatively insensitive to the SST change when the SST is high (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ec). Specifically, there is a case-dependent critical SST threshold for the development of a TC since the TC intensity increases notably (slightly) with the SST when the SST is below (above) the threshold\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In this study, as the time-averaged SSTs in both IPO-positive years and -negative years are high (i.e., \u0026gt;\u0026thinsp;29\u0026deg;C), the small difference in the absolute SST (i.e., \u0026lt; 0.5\u0026deg;C) does not lead to a marked difference in TC intensity (i.e., MWS) between IPO-positive years and -negative years (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003eSecond, the radial distribution of the TC surface wind speed is affected by that of the SST. Differences in the radial distribution of the SST lead to variations in the SLP gradient, which in turn cause differences in the tangential wind speed (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eb and c). Although the average SST in the negative IPO year is lower than that in the positive IPO year, in terms of the radial distribution, the SSTs in the TC outer regions of the positive (negative) IPO year are significantly lower (slightly higher) than those in the inner regions. This difference in SST between the TC inner and outer regions (relative SST) is a key factor contributing to the radial distribution of TC winds, particularly the greater tangential wind speed in the outer regions of TCs during the positive phase than during the negative phase.\u003c/p\u003e \u003cp\u003eOverall, under high-SST conditions, the contrast between the relative SST, rather than the absolute value of the SST (absolute SST), plays an important role in determining TC size. This may also explain why the unfiltered IPO index has a greater correlation with TC size than the ENSO index does, as the former is determined by SST differences, which are similar to the relative SST, whereas the latter, the Nino 3.4 index, is determined by the absolute SST.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3. Conclusion and discussion","content":"\u003cp\u003eIn this study, the changes in the TC size in the WNP over the past few decades and the possible reasons for these changes are investigated. In this study, the TC size is defined as the radius of the strong wind (i.e., R30), and data from the moment when the TC reaches its maximum intensity are used to reduce uncertainty during the weaker intensity stages. The results indicate that natural variability has been the main driver of these significant changes and the cause of the possible downward trend in TC size over the past few decades. Among the three natural variabilities\u0026mdash;ENSO, IPO, and PDO\u0026mdash;the IPO has the most important impact on TC size changes. In this study, the impact of positive and negative IPO years on TC size (R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e) is analyzed, and the results indicate that the change in the R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e is closely aligned with the IPO phase: during positive IPO years, the R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e is typically greater, whereas during negative years, the R30\u003csub\u003eLMI\u0026thinsp;\u0026minus;\u0026thinsp;annual\u003c/sub\u003e is smaller. Further research indicates that although the difference in the absolute value of the SST (absolute SST) between the positive and negative IPO years is small (\u0026lt;\u0026thinsp;0.5\u0026deg;C), the radial distribution of the SST (relative SST) is significant, leading to noticeable variations in the radial distribution of the TC wind speed. Specifically, during the negative IPO year, the SST in the outer regions of the TC is significantly lower than that in the core regions, whereas during the positive IPO year, the outer-region SST is slightly higher than that in the core. This contrast in the SST between the inner and outer regions is assumed to affect local energy exchanges and latent heating at the air‒sea interface and thus alter the local pressure and its gradient, ultimately resulting in significantly higher wind speeds in the outer regions of TCs during the positive IPO year than during the negative IPO year.\u003c/p\u003e \u003cp\u003eAdditionally, the results of this study revealed that under high-SST conditions (\u0026gt;\u0026thinsp;29\u0026deg;C), the sensitivity of TC intensity (such as the MWS) to SST changes is relatively low, whereas the variation in the TC size depends primarily on the radial contrast in SST (relative SST) rather than the absolute value of SST. Compared with the ENSO index, which is determined by the absolute value of the SST, the IPO index is based on SST differences (relative SST). This mechanism aligns with the primary factors driving changes in TC size, thereby explaining why the correlation between the IPO index and TC size is significantly greater than that with the ENSO index.\u003c/p\u003e \u003cp\u003eIn summary, the radial distribution of the SST and the contrast between the inner and outer regions play key roles in regulating the distributions of TC wind speeds and size variations. This finding enhances our understanding of the mechanisms driving changes in the TC size and provides new perspectives for future TC activity predictions. In addition, this study focuses on the impact of natural variability on changes in the TC size rather than the contribution of global warming to the trend in changes in the TC size. The role of global warming in the trend of the TC size remains uncertain, primarily due to limitations in the length of high-quality observational records, which restricts our ability to detect robust trends within such relatively short time series.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003e1. Data Sources and Preprocessing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe operational implementation of geostationary satellite monitoring since the 1980s has significantly enhanced the reliability of global tropical cyclone best-track records (Shan et al., 2024). In this study, the TC size information from 1980 to 2023 provided by the JMA in the IBTrACS v4.0 best track dataset \u003csup\u003e43\u003c/sup\u003e are primarily used. The JMA data include the TC position, intensity, and 30-knot wind radius (R30; units: km) recorded every 3 hours, with the most recent TC data available up to 2024. To validate the robustness of the decreasing TC size trend in the JMA data, in this study, the best track dataset and analyzed size data from TC records from the JTWC, as well as the CMA (https://tcdata.typhoon.org.cn/tcsize.html)\u003csup\u003e44\u003c/sup\u003e, were selected. The CMA dataset provides\u0026nbsp;the\u0026nbsp;TC position, intensity, and 34-knot wind radius (R34) recorded every 6 hours for the\u0026nbsp;northwestern\u0026nbsp;Pacific region (including the South China Sea) from 1980 to 2020. The JTWC dataset provides the TC position, intensity, and 34-knot wind radius (R34)\u0026nbsp;data\u0026nbsp;recorded every 3 hours from 2001 to 2024.\u003c/p\u003e\n\u003cp\u003eIn this study, the 6-hourly reanalysis data from ERA5 (https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview), including variables such as SST, surface wind speed, and monthly average SST, are also used. Importantly, owing to the relatively low resolution of the ERA5 data (0.25\u0026deg;), the simulated core information of TCs may be inaccurate. However, the resolution is sufficient to identify the outer wind speeds of TCs; thus, the distribution of the TC outer wind speed is relatively reliable. Although the ERA5 data may underestimate the overall TC wind speed due to its coarser resolution, it can still be used to explore differences in TC outer wind speeds between cold and warm years (Bian et al., 2021). Additionally, ERA5 significantly underestimates TC wind speed and thus the TC size, but this does not affect our analysis of the relative differences in the TC size between IPO-positive and -negative years.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2. Definition of TC Size Metrics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study considers only northwestern Pacific TCs (excluding those generated in the South China Sea) during the active TC season (July-November), with maximum wind speeds exceeding 30 kt. In this study, the average size during the lifetime of a single TC is used as an indicator of the TC\u0026apos;s outer size, specifically R34\u003csub\u003eAVE\u003c/sub\u003e or R30\u003csub\u003eAVE\u003c/sub\u003e. The method of calculation is as follows:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"469\" height=\"85\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003e\u0026tau;\u0026nbsp;\u003c/em\u003erepresents the life cycle duration of a specific TC and\u0026nbsp;\u003cem\u003eR\u003c/em\u003e(\u003cem\u003et\u003c/em\u003e) represents the specific wind radius at a given time for that TC.\u003c/p\u003e\n\u003cp\u003eThe annual average of the TC lifetime average size is expressed as follows:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"486\" height=\"76\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003eN\u003c/em\u003e denotes the total number of TCs in a given year, \u003cem\u003eI\u0026nbsp;\u003c/em\u003erefers to the \u003cem\u003ei\u003c/em\u003e-th TC in that year, and \u003cem\u003eR\u003c/em\u003e\u003csub\u003eAVE\u003c/sub\u003e(\u003cem\u003ei\u003c/em\u003e) is the \u003cem\u003eR\u003c/em\u003e\u003csub\u003eAVE\u003c/sub\u003e for the i-th TC.\u003c/p\u003e\n\u003cp\u003eAdditionally, in this study, the TC size at the moment of the LMI is used as another indicator of TC size, specifically R34\u003csub\u003eLMI\u003c/sub\u003e or R30\u003csub\u003eLMI\u003c/sub\u003e. The annual average TC size at the moment of the LMI is expressed as follows:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"385\" height=\"79\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003eN\u003c/em\u003e represents the total number of TCs in a given year, \u003cem\u003ei\u003c/em\u003e refers to the \u003cem\u003ei\u003c/em\u003e-th TC in that year, and \u003cem\u003eR\u003c/em\u003e\u003csub\u003eLMI\u003c/sub\u003e(\u003cem\u003ei\u003c/em\u003e) is the specific wind radius (\u003cem\u003eR\u003c/em\u003e\u003csub\u003eLMI\u003c/sub\u003e) at the moment when the \u003cem\u003ei\u003c/em\u003e-th TC reaches its LMI.\u003c/p\u003e\n\u003cp\u003eThere are significant differences in the annual variations in the lifetime average TC size between different agencies (Fig. S1). This may be because the moment of TC formation is determined when the initial storm intensity reaches a specified threshold. Since a storm\u0026rsquo;s intensity is relatively low in the early stages of development, accurately estimating its strength is difficult, leading to considerable uncertainty. As a result, there are discrepancies in the recorded locations of TC initiation and life cycle durations across different observational datasets. Additionally, although the TC size is weakly correlated with TC intensity, uncertainties in intensity estimates can also affect TC size information. Therefore, using the average TC size of the life cycle to study characteristics of TC size variation may introduce more uncertainty and error. Moreover, Kim et al. mentioned that differences in the TC size estimates across different agencies may arise because some agencies consider only the storm\u0026apos;s wind when estimating the wind radius, whereas others incorporate the effects of other midlatitude weather systems\u003csup\u003e45\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eCompared with the lifetime average TC size, the error in identifying the storm intensity at the moment of the maximum life cycle intensity (LMI) is smaller, and the uncertainty in determining the TC location is reduced. This is because, at the LMI moment, the TC size is generally larger within its lifetime, making it easier to identify. Additionally, the determination of the LMI moment is solely related to the relative intensity of the TC\u0026apos;s life cycle. Therefore, although the life cycle intensity recorded in different observational datasets may vary, the LMI moment is less affected by the absolute intensity.\u003c/p\u003e\n\u003cp\u003eFigures S1(d-f) show the trends in the annual average TC size at the moment of maximum intensity (R30\u003csub\u003eLMI-annual\u003c/sub\u003e or R34\u003csub\u003eLMI-annual\u003c/sub\u003e) for TCs in the WNP from the CMA, JMA, and JTWC datasets between 2001 and 2020. The values of R34\u003csub\u003eLMI-annual\u003c/sub\u003e (and R30\u003csub\u003eLMI-annual\u003c/sub\u003e) are generally slightly larger than those of R34\u003csub\u003eAVE-annual\u003c/sub\u003e (and R30\u003csub\u003eAVE-annual\u003c/sub\u003e), but in some years, the former is smaller than the latter. This result is because the TC size and intensity are not strictly linearly related, so in these years, in some TCs, the size at the moment of the LMI is smaller than that at other points in their life cycle.\u0026nbsp;However, the overall trends of both variables are quite similar. In years when the R30\u003csub\u003eAVE-annual\u003c/sub\u003e is significantly larger (or smaller) than usual (e.g., 2002, 2010), the R30\u003csub\u003eLMI-annual\u003c/sub\u003e also follows a similar pattern, being larger (or smaller) in these years. The correlation coefficients between the three datasets, as shown in Table S1, increased significantly compared with the life cycle average TC size, and the differences in changes in the trends among the three datasets decreased. Therefore, compared with R\u003csub\u003eAVE\u003c/sub\u003e, R\u003csub\u003eLMI\u003c/sub\u003e reduces, to some extent, the differences in uncertainty caused by varying data sources. For these reasons, and because the JMA dataset spans a longer period, this study focuses on using the TC size at the moment of maximum life cycle intensity (R30\u003csub\u003eLMI-annual\u003c/sub\u003e) from the JMA dataset to investigate the associated patterns of size variation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3. Natural variability and Decadal Signal Extraction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo analyze the impact of natural variability on changes in the TC size, in this study, the monthly indices of the El Ni\u0026ntilde;o‒Southern Oscillation (ENSO) (Ni\u0026ntilde;o-3.4), Pacific Decadal Oscillation (PDO), and Tripole Index for the Interdecadal Pacific Oscillation (TPI, which is the index reflecting the IPO) from 1980 to 2023, which are all provided by the Physical Sciences Division of the Earth System Research Laboratory at the National Oceanic and Atmospheric Administration (NOAA) (https://psl.noaa.gov/gcos_wgsp/Timeseries/), are used. The Ni\u0026ntilde;o-3.4 index is calculated as the monthly average SST of the region bounded by 5\u0026deg;N-5\u0026deg;S and 170\u0026deg;W-120\u0026deg;W and is based on data from HadISST1. The PDO index is derived from a time series of spatially averaged monthly SSTs in the North Pacific (north of 20\u0026deg;N) and is calculated using the SST time covariance matrix from 1900\u0026ndash;1993. The IPO index (i.e., TPI) is defined as the difference between the SST anomaly averaged over the central equatorial Pacific (10\u0026deg;S-10\u0026deg;N, 170\u0026deg;E-90\u0026deg;W) and the average SST anomalies in the WNP (25\u0026deg;N-45\u0026deg;N, 140\u0026deg;E-145\u0026deg;W) and southwestern Pacific (50\u0026deg;S-15\u0026deg;S, 150\u0026deg;E-160\u0026deg;W). The data used here are in unfiltered form, but in the later parts of the analysis, both the IPO and R30 data are low-pass filtered to extract the decadal signals of climate indices.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study follows the IPO filtering method provided by the NOAA website and uses a Chebyshev low-pass filter. The specific parameters are as follows: a 13-point weighting coefficient, a low-frequency cutoff frequency of 1/12, and standard smoothing parameters. This zero-phase filter effectively separates the low-frequency components of natural variability, retaining decadal signal features, such as the IPO, while effectively filtering out higher-frequency interannual fluctuations. In subsequent analyses of the interdecadal characteristics of TC size, to maintain consistency with the filtered IPO index, the R30 index was processed using an identical filtering approach.\u003c/p\u003e\n\u003cp\u003e4. \u003cstrong\u003eStatistical Framework for Attribution of the TC Size\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe annual average ENSO, PDO, and IPO indices mentioned in this study are based on July to November averages, which is a period considered the active season for TCs in the WNP. To quantitatively analyze the independent effects of natural variability on changes in the TC size, in this study, a linear regression framework is employed, starting from the time series of the TC size:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"529\" height=\"34\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003eX(t)\u003c/em\u003e represents the time series of standardized indices for each climate mode (which may also refer to multiple natural variabilities) and where \u0026nbsp; is the TC size regression residual after removing the linear influence of the corresponding mode. The statistical significance of each regression coefficient ( ) is rigorously evaluated using a two-tailed Student\u0026apos;s t test (confidence level of 95%, \u0026alpha; = 0.05).\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConceptualization: YS, YZ, WZ, ZMMethodology: YS, YS, YZInvestigation: YZ, YS, ZFVisualization: YZSupervision: YS, WZ, ZMWriting\u0026mdash;original draft: YS, YZWriting\u0026mdash;review \u0026amp; editing: YS, YZ\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll data, code, and materials used in the analyses are publicly available through Figshare (DOI: 10.6084/m9.figshare.29219018.v1). This dataset includes [briefly describe the core contents, e.g., 'tropical cyclone size measurements over the western North Pacific from 1980-2020, along with associated metadata and analysis scripts']. No special restrictions apply to these materials beyond the CC BY 4.0 license terms. The raw data supporting the conclusions are provided in full within this repository, and processed data are available in the main text or supplementary materials of this article. The authors will make any additional methodological details available upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePeduzzi, P. et al. O. Global trends in tropical cyclone risk. \u003cem\u003eNat. Clim. Change\u003c/em\u003e 2, 289\u0026ndash;294 (2012).\u003c/li\u003e\n\u003cli\u003eJing, R. et al. Global population profile of tropical cyclone exposure from 2002 to 2019. \u003cem\u003eNature.\u003c/em\u003e 626, 549\u0026ndash;554 (2024).\u003c/li\u003e\n\u003cli\u003eKim, H.-K. \u0026amp; Seo, K.-H. Cluster analysis of tropical cyclone tracks over the western North Pacific using a self-organizing map. \u003cem\u003eJ. Clim.\u003c/em\u003e 29, 3731\u0026ndash;3751 (2016).\u003c/li\u003e\n\u003cli\u003eRappaport, E. N. Loss of life in the United States associated with recent Atlantic tropical cyclones: bulletin of the American meteorological society. \u003cem\u003eBull. Am. Meteorol. Soc.\u003c/em\u003e 81, 2065\u0026ndash;2073 (2000).\u003c/li\u003e\n\u003cli\u003eChen, L., Ma, Z. \u0026amp; Fei, J. How different is tropical cyclone precipitation over land and ocean?. \u003cem\u003enpj Clim Atmos Sci.\u003c/em\u003e 8, 70 (2025).\u003c/li\u003e\n\u003cli\u003eEmanuel, K. Increasing destructiveness of tropical cyclones over the past 30 years. \u003cem\u003eNature.\u003c/em\u003e 436, 686-688 (2005).\u003c/li\u003e\n\u003cli\u003eBian, G. F., Nie, G. Z. \u0026amp; Qiu, X. How well is outer tropical cyclone size represented in the ERA5 reanalysis dataset? \u003cem\u003eAtmos. Res.\u003c/em\u003e 249, 105339 (2021).\u003c/li\u003e\n\u003cli\u003eMok, D., Chan, J. C. L. \u0026amp; Chan, K. T. F. A 31-year climatology of tropical cyclone size from the NCEP Climate Forecast System Reanalysis. \u003cem\u003eInt. J. Climatol.\u003c/em\u003e 38, 5669-5685 (2018).\u003c/li\u003e\n\u003cli\u003eSchenkel, B. A. et al. Evaluating outer tropical cyclone size in reanalysis datasets using QuikSCAT data. \u003cem\u003eJ. Clim.\u003c/em\u003e 30, 8745-8762 (2017).\u003c/li\u003e\n\u003cli\u003eMerrill, R. T. A comparison of large and small tropical cyclones. \u003cem\u003eMon. Weather Rev.\u003c/em\u003e 112, 1408 (1984).\u003c/li\u003e\n\u003cli\u003eWeatherford, C. L. \u0026amp; Gray, W. M. Typhoon structure as revealed by aircraft reconnaissance. Part I: Data analysis and climatology. \u003cem\u003eMon. Weather Rev.\u003c/em\u003e 116, 1032\u0026ndash;1043 (1988).\u003c/li\u003e\n\u003cli\u003eWu, L. et al. Implications of the observed relationship between tropical cyclone size and intensity over the Western North Pacific. \u003cem\u003eJ. Clim.\u003c/em\u003e 28, 9501-9506 (2015).\u003c/li\u003e\n\u003cli\u003eGori, A., Lin, N., Schenkel, B. \u0026amp; Chavas, D. North Atlantic tropical cyclone size and storm surge reconstructions from 1950\u0026ndash;present. \u003cem\u003eJ. Geophys. Res. Atmos.\u003c/em\u003e 128, e2022JD037312 \u003c/li\u003e\n\u003cli\u003eChan, J. C. L. \u0026amp; Yip, C. K. M. Interannual variations of tropical cyclone size over the western North Pacific. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e 30, 665-678 (2003).\u003c/li\u003e\n\u003cli\u003eShou, H. et al. Decreasing trend of western North Pacific tropical cyclone inner-core size over the past decades. \u003cem\u003eJ. Meteorol. Res.\u003c/em\u003e 35, 635-645 (2021).\u003c/li\u003e\n\u003cli\u003eZhang, K. \u0026amp; Chan, K. T. F. An ERA5 global climatology of tropical cyclone size asymmetry. \u003cem\u003eInt. J. Climatol.\u003c/em\u003e 43, 950-963 (2022).\u003c/li\u003e\n\u003cli\u003eKnaff, J. A., Longmore, S. P. \u0026amp; Molenar, D. A. An objective satellite-based tropical cyclone size climatology. \u003cem\u003eJ. Clim.\u003c/em\u003e 27, 455-476 (2014).\u003c/li\u003e\n\u003cli\u003eSun, Y. et al. Impact of ocean warming on tropical cyclone track over the western North Pacific: A numerical investigation based on two case studies. \u003cem\u003eJ. Geophys. Res. Atmos.\u003c/em\u003e 122, 6476-6492 (2017a).\u003c/li\u003e\n\u003cli\u003eWang, S. \u0026amp; Toumi, R. Reduced sensitivity of tropical cyclone intensity and size to sea surface temperature in a radiative-convective equilibrium environment. \u003cem\u003eAdv. Atmos. Sci.\u003c/em\u003e 35, 981-993 (2018).\u003c/li\u003e\n\u003cli\u003eXu, Z. et al. Tropical cyclone size change under ocean warming and associated responses of tropical cyclone destructiveness: idealized experiments. \u003cem\u003eJ. Meteorol. Res.\u003c/em\u003e 34, 163-175 (2020).\u003c/li\u003e\n\u003cli\u003eShan, K. et al. How does global accumulated tropical cyclone energy vary in response to a changing climate? \u003cem\u003eSci. Bull.\u003c/em\u003e 69, 2053-2064 (2024).\u003c/li\u003e\n\u003cli\u003ePatricola, C. M. et al. The influence of ENSO flavors on western North Pacific tropical cyclone activity. \u003cem\u003eJ. Clim.\u003c/em\u003e 31, 5395-5416 (2018).\u003c/li\u003e\n\u003cli\u003eScoccimarro, E. et al. The Pacific Decadal Oscillation modulates tropical cyclone days on the interannual timescale in the North Pacific Ocean. \u003cem\u003eJ. Geophys. Res. Atmos.\u003c/em\u003e 126, e2021JD034988 (2021).\u003c/li\u003e\n\u003cli\u003eZhao, J. et al. Untangling impacts of global warming and interdecadal Pacific oscillation on long-term variability of North Pacific tropical cyclone track density. \u003cem\u003eSci. Adv.\u003c/em\u003e 6, eaba6813 (2020).\u003c/li\u003e\n\u003cli\u003eChan, J. C. L. Tropical cyclone activity over the western North Pacific associated with El Ni\u0026ntilde;o and La Ni\u0026ntilde;a events. \u003cem\u003eJ. Clim.\u003c/em\u003e 13, 2960\u0026ndash;2972 (2000).\u003c/li\u003e\n\u003cli\u003eLander, M. A. An exploratory analysis of the relationship between tropical storm formation in the western North Pacific and ENSO. \u003cem\u003eMon. Weather Rev.\u003c/em\u003e 122, 636-651 (1994).\u003c/li\u003e\n\u003cli\u003eCai, Y. et al. Enhanced predictability of rapidly intensifying tropical cyclones over the western North Pacific associated with snow depth changes over the Tibetan Plateau. \u003cem\u003eJ. Clim\u003c/em\u003e. 35, 2093\u0026ndash;2110 (2022).\u003c/li\u003e\n\u003cli\u003eSun, Y. et al. The opposite effects of inner and outer sea surface temperature on tropical cyclone intensity. \u003cem\u003eJ. Geophys.\u003c/em\u003e\u003cem\u003eRes. Atmos\u003c/em\u003e. 119, 2193-2208 (2014).\u003c/li\u003e\n\u003cli\u003eKossin, J., Emanuel, K. \u0026amp; Vecchi, G. The poleward migration of the location of tropical cyclone maximum intensity. \u003cem\u003eNature.\u003c/em\u003e 509, 349\u0026ndash;352 (2014).\u003c/li\u003e\n\u003cli\u003eHenley, B.J. et al. A tripole index for the Interdecadal Pacific Oscillation. \u003cem\u003eClim. Dyn.\u003c/em\u003e 45, 3077\u0026ndash;3090 (2015).\u003c/li\u003e\n\u003cli\u003eSmith, R. K., Montgomery, M. T., \u0026amp; Schmidt, C. Dynamical constraints on the intensity and size of tropical cyclones. \u003cem\u003eQuarterly Journal of the Royal Meteorological Society\u003c/em\u003e 137, 1841 - 1855 (2010).\u003c/li\u003e\n\u003cli\u003eWeatherford, C. L. \u0026amp; Gray, W. M. Typhoon structure as revealed by aircraft reconnaissance. Part II: structural variability. \u003cem\u003eMon. Weather Rev.\u003c/em\u003e 116, 1044-1056 (1988).\u003c/li\u003e\n\u003cli\u003eRuan, Z. \u0026amp; Wu, Q. Relationship between size and intensity in North Atlantic tropical cyclones with steady radii of maximum wind. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e 49, e2021GL095632 (2022).\u003c/li\u003e\n\u003cli\u003eSun, Y. et al. Impact of ocean warming on tropical cyclone size and its destructiveness. \u003cem\u003eSci. Rep.\u003c/em\u003e 7, 8154 (2017).\u003c/li\u003e\n\u003cli\u003eEmanuel, K. A. An air-sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. \u003cem\u003eJ. Atmos. Sci.\u003c/em\u003e 43, 585\u0026ndash;604 (1986).\u003c/li\u003e\n\u003cli\u003eRotunno, R. \u0026amp; Emanuel, K. A. An air-sea interaction theory for tropical cyclones. Part II: Evolutionary study using a nonhydrostatic axisymmetric numerical model. \u003cem\u003eJ. Atmos. Sci.\u003c/em\u003e 44, 542-561 (1987).\u003c/li\u003e\n\u003cli\u003eRaymond, D. J. Convective processes and tropical atmospheric circulations. \u003cem\u003eQ. J. R. Meteorol. Soc.\u003c/em\u003e 120, 1431-1455 (1994).\u003c/li\u003e\n\u003cli\u003eXu, J. \u0026amp; Wang, Y. Sensitivity of tropical cyclone inner-core size and intensity to the radial distribution of surface entropy flux. \u003cem\u003eJ. Atmos. Sci.\u003c/em\u003e 67, 1831\u0026ndash;1852 (2010).\u003c/li\u003e\n\u003cli\u003eFierro, A. O., Rogers, R. F. \u0026amp; Marks, F. D. The impact of horizontal grid spacing on the microphysical and kinematic structures of strong tropical cyclones simulated with the WRF-ARW Model. \u003cem\u003eMon. Weather Rev\u003c/em\u003e. 137, 3717-3743 (2009).\u003c/li\u003e\n\u003cli\u003eBrill, K. F. Revisiting an Old Concept: The Gradient Wind*. \u003cem\u003eMon. Weather Rev\u003c/em\u003e. 142, 1460 - 1471 (2014).\u003c/li\u003e\n\u003cli\u003eDulac, W., Cattiaux, J., Chauvin, F., Bourdin, S. \u0026amp; Fromang, S. Assessing the representation of tropical cyclones in ERA5 with the CNRM tracker. \u003cem\u003eClim. Dyn.\u003c/em\u003e 62, 223\u0026ndash;238 (2024).\u003c/li\u003e\n\u003cli\u003eXu, Z. et al. Global tropical cyclone size and intensity reconstruction dataset for 1959-2022 based on IBTrACS and ERA5 data. \u003cem\u003eEarth Syst. Sci. Data\u003c/em\u003e 16, 5753-5766 (2024).\u003c/li\u003e\n\u003cli\u003eKnapp, K. R. et al. International Best Track Archive for Climate Stewardship (IBTrACS) Project, Version 4. NOAA National Centers for Environmental Information (2018).\u003c/li\u003e\n\u003cli\u003eYing, M. et al. An overview of the China Meteorological Administration tropical cyclone database.\u003cem\u003e J. Atmos. Ocean. Technol.\u003c/em\u003e 31, 287-301 (2014).\u003c/li\u003e\n\u003cli\u003eKim, H.-J., Moon, I.-J. \u0026amp; Oh, I. Comparison of tropical cyclone wind radius estimates between the KMA, RSMC Tokyo, and JTWC. \u003cem\u003eAsia-Pac. J. Atmos. Sci\u003c/em\u003e. 58, 563-576 (2022).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6898595/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6898595/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTropical cyclone (TC) size is a key factor in determining TC destructiveness and a major challenge in understanding changes in TCs. Although much effort has been devoted to investigating TC track and intensity changes, relatively few studies have focused on TC size changes, particularly their responses to natural variability and global warming. Here, we use a metric of TC size in which TCs achieve their lifetime-maximum intensity, which is relatively insensitive to uncertainty in past data. The results show that the Interdecadal Pacific Oscillation (IPO) dominates TC size variability in the western North Pacific, which is the most active region for TCs. Moreover, TC size variability is governed primarily by radial sea surface temperature (SST) gradients rather than absolute SST values. This finding explains not only the difference in TC size between IPO positive and negative years but also the stronger correlation between the IPO and TC size than other climate indices, such as El Ni\u0026ntilde;o\u0026ndash;Southern Oscillation (ENSO), since the IPO is derived from SST differences and thus gradients, whereas ENSO is determined by absolute SST values. This finding further implies that the pattern rather than the magnitude of the SST change will determine the change in TC size under future global warming.\u003c/p\u003e","manuscriptTitle":"Change in tropical cyclone size over the western North Pacific","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-23 16:59:36","doi":"10.21203/rs.3.rs-6898595/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ad7fdddc-2779-46e6-90ce-f5e7332e402a","owner":[],"postedDate":"June 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":50412995,"name":"Earth and environmental sciences/Climate sciences/Climate change"},{"id":50412996,"name":"Earth and environmental sciences/Climate sciences"},{"id":50412997,"name":"Earth and environmental sciences/Ocean sciences"}],"tags":[],"updatedAt":"2025-07-09T08:53:38+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-23 16:59:36","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6898595","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6898595","identity":"rs-6898595","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}