Detecting Agricultural and Meteorological Drought With Gross Primary Production Recovery Including Spatiotemporal Statistical Analysis in East Africa's Lake Victoria Basin

Detecting Agricultural and Meteorological Drought With Gross Primary Production Recovery Including Spatiotemporal Statistical Analysis in East Africa's Lake Victoria Basin | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Detecting Agricultural and Meteorological Drought With Gross Primary Production Recovery Including Spatiotemporal Statistical Analysis in East Africa's Lake Victoria Basin Yuefeng Hao, Jongjin Baik, Sseguya Fred, Minha Choi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-774555/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Drought imposes severe, long-term effects on global environments and ecosystems. A better understanding of how long it takes a region to recover to pre-drought conditions after drought is essential for addressing future ecology risks. In this study, drought-related variables were obtained using remote sensing and reanalysis products for 2003 to 2016. The meteorological drought index (standardized precipitation evapotranspiration index [SPEI]) and agricultural drought index (vegetation condition index [VCI]) were employed to estimate drought duration time (DDT) and drought recovery time (DRT). To the basin’s west, decreasing rainfall and increasing potential evapotranspiration led to decreasing SPEI. On the east side, decreasing soil moisture from each depth effects vegetation condition, which results in a decreasing gross primary productivity and VCI. Extreme meteorological drought events are likely to occur in the basin’s northeastern and middle western areas, while the southern basin is more likely to suffer from extreme agricultural drought events. The mean SPEI-based DDT (2.45 months) was smaller than the VCI-based DDT (2.97 months); the average SPEI-based DRT (2.02 months) was larger than the VCI-based DRT (1.63 months). Most of the area needs 1 or 2 months to recover from drought except for the basin’s northwestern area, where the DRT is more than 8 months. DDT is the most important parameter in determining DRT. These results provide useful information about regional drought recovery that will help local governments looking to mitigate potential environmental risks and formulate appropriate agricultural policies in Lake Victoria Basin. Environmental Policy Drought Drought recovery time SPEI VCI GPP Lake Victoria Basin Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1. Introduction As a devastating hazard, drought not only imposes extensive and long-term effects on the global environmental system, it also causes significant losses to the economy and human life (Mishra and Singh, 2010 ; Sternberg, 2011 ; Van Dijk et al., 2013 ). To reduce drought risk and prevent drought under global climate change and global warming (Trenberth et al., 2014 ), several research contributions have been made in the fields of drought monitoring (Aadhar and Mishra, 2017 ; West et al., 2019 ) and prediction (AghaKouchak, 2015 ; Hao et al., 2018 , 2014 ). Those studies considered hydrological and meteorological variables to better understand regional drought event causes and quantification. Other studies use ecological variables to identify this natural hazard at the ecosystem level (Anderegg et al., 2018 ; Banerjee et al., 2013 ; Schwalm et al., 2017 ; Van der Molen et al., 2011 ). In particular, drought recovery time (DRT) has received growing research attention in recent years (Ahmadi et al., 2019 ; Ahmadi and Moradkhani, 2019 ; He et al., 2018 ; Huang et al., 2021 ; Liu et al., 2019 ; Schwalm et al., 2017 ; Seneviratne and Ciais, 2017 ; Zhang et al., 2019 ). DRT will likely become longer than the time between drought events (Schwalm et al., 2017 ). Regional drought effects are compounded if a new drought event occurs before recovery from a preceding drought event is complete (Seneviratne and Ciais, 2017 ). Therefore, accurate DRT assessments are essential for understanding possible ecological risks. DRT is defined as the time required for a region to fully return to its pre-drought conditions (Schwalm et al., 2017 ). Drought identification and recovery parameter selection are the key steps in DRT calculations. To identify a drought event, previous studies used the meteorological drought index (Standardized Precipitation Evapotranspiration Index, SPEI) and Drought Severity Index (DSI) (Liu et al., 2019 ; Schwalm et al., 2017 ; Yu et al., 2017 ). Climate data-based SPEI is easy to estimate for long-term analysis but it is not linked to plant condition (Vicente-Serrano et al., 2010 ). Satellite-based DSI includes greenness information with high spatial resolution. However, DSI data is only available from 2000 to 2011 with uncertainties from the satellite (cloud cover, atmospheric aerosols, and low solar illumination) and single input data (Mu et al., 2013 ). Many factors such as water quantity (streamflow and total water storage), water quality (water temperature, turbidity, and dissolved oxygen), ecosystem fluxes (carbon and energy fluxes), and gross primary productivity (GPP) are used in drought recovery assessments (Ahmadi et al., 2019 ; He et al., 2018 ; Schwalm et al., 2017 ; Seneviratne and Ciais, 2017 ; Zhang et al., 2019 ). For example, the change in total water storage was used to identify hydrological DRT (between 3.6 and 5.7 months) for the Yangtze River in China (Zhang et al., 2019 ). Ahmadi et al. ( 2019 ) analyzed drought recovery by considering both streamflow and water quality parameter changes. They found that the average recovery time in the contiguous United States is around 1.2 months. The time required for carbon and energy fluxes to recover from the 2012 U.S. drought (0.5–2 months) and the 2003 European drought (1–2 months) were examined by He et al. ( 2018 ). Based on changes to GPP, global spatiotemporal DRT patterns were examined by Schwalm et al. ( 2017 ) and Yu et al. ( 2017 ). All of these studies focused on the DRT length and response functions. Among the factors, GPP is the most impactful due to its high drought sensitivity and spatiotemporal patterns. It can accommodate increasingly fine spatial resolution and frequent repeat measurements (Schwalm et al., 2017 ). Schwalm et al. ( 2017 ) estimated that DRT can be determined at 0.5° spatial resolution and 6 months of temporal resolution from 1901 to 2010 around the world. They found that most of the world can recover from a drought in less than six months. Unlike a conventional drought, a flash drought typically occurs during warm seasons and can occur more frequently - in one or two months (Ford and Labosier, 2017 ). Therefore, DRT should be examined in more detailed studies at a high spatiotemporal resolution. Yu et al. ( 2017 ) calculated global DRT from 2000 to 2011 at 0.5° spatial resolution and 1 month temporal resolution, but the results were different for Schwalm et al. ( 2017 ) in terms of spatial pattern and DRT length. As reported by Liu et al. ( 2019 ), using different methods to define drought events and recovery levels are the key factors contributing to the contradictory conclusions. Despite the large amount of research on meteorological and hydrological DRT, there has been very little work integrating long-term agricultural DRT at a high spatiotemporal resolution in dry regions such as East Africa, where people are largely dependent on rain-fed agriculture and livestock farming (Gebremeskel et al., 2019 ). To better understand drought recovery in the Lake Victoria Basin, our objectives are 1) check the seasonal patterns and trends of drought-related variables; 2) capture drought events using meteorological (SPEI) and agricultural (vegetation condition index [VCI]) drought indices from 2003 to 2016; 3) investigate SPEI based- and VCI based-DRT for high spatiotemporal resolution; and 4) examine parameter importance for determining DRT across the Lake Victoria Basin. To the best of our knowledge, this is the first comprehensive study to quantify agricultural DRT at a high spatiotemporal resolution. 2. Study Area And Dataset 2.1 Study area As the second largest freshwater lake in the world, Lake Victoria supports one of the world’s poorest and densest populations. Its catchment covers a 194,200 km 2 area and has a total population of 30 million people (Mailu, 2001 ). Basin agriculture supports more than 70% of the local population (Zhou et al., 2014 ). Figure 1 shows the land cover distribution in the Lake Victoria Basin. Grassland (16%) is predominantly in the west, open water (26%) in the middle, and cropland (36%) in the east. Maize is the main crop (Zhou et al., 2014 ). The Victoria Basin is shared by 5 agricultural East African nations: Kenya, Uganda, Tanzania, Rwanda, and Burundi. Kenya is in the northeastern part of the basin, and contains three provinces: Nyanza, Western, and Rift Valley. They are predominantly cropland. The northwest comprises central Ugandan districts Masaka and Mpigi, which are characterized by large farming communities. Tanzania, located in the basin’s south, also has a large cropland area. Most of the farmland is located in the Mwanza and Mara provinces, though it also covers the Kagera region, a key region for food production and distribution whose landcover is a mixture of cropland and grassland. Rwanda is located in the western part of the study area and is mainly covered by cropland. Burundi is also in the western basin and is half cropland. According to the Köppen-Geiger climate classification (Peel et al., 2007 ), Lake Victoria Basin’s climate zone is tropical with four distinct seasons: hot dry (Dec. to Feb., DJF), major rainy (Mar. to May, MMA), short dry (Jun. to Aug., JJA) and short rainy (Sept. to Nov., SON) (Awange JL, 2006 ). Annual precipitation ranges between 670 and 2,200 mm, with a mean value of 1,202 mm (Kizza et al., 2009 ). Several studies discuss droughts in the Lake Victoria Basin during the twentieth century’s last two decades (Funk et al., 2014 ; Gebremeskel et al., 2019 ; Lyon and Dewitt, 2012 ). Since 2003, drought has continued to occur every 3 years on average, with a significant increase in severity and frequency (Funk et al., 2014 ). Drought re-emergence has had devastating effects on local people, flora, and fauna. 2.2 Dataset To estimate SPEI, we used global hydrological datasets (Global Land Data Assimilation System Noah Land Surface Model L4 monthly 0.25° × 0.25° Version 2.1 [GLDAS_NOAH025_M _V2.1] and Tropical Rainfall Measurement Mission Rainfall Estimate L3 1 month 0.25-degree x 0.25-degree V7 [TRMM-3B43]). First, we used the GLDAS-2.1 net radiation flux, ground heat flux, air temperature, and surface pressure to calculate potential evapotranspiration (PET) based on the Priestley and Taylor (PT) method. Next, we used the GLDAS-based PET and rainfall data from TRMM-3B43 to compute SPEI. In this study, we used the GLDAS_NOAH025_M _V2.1 dataset (Beaudoing and Rodell, 2020) for soil moisture (0–10cm, 10–40cm, 40–100cm, and 100–200cm) data, net radiation flux, ground heat flux, air temperature, and surface pressure. GLDAS combines ground and satellite data via assimilation methods along with land surface models. The National Centers for Environmental Prediction (NCEP), National Aeronautics and Space Administration (NASA), and National Oceanic and Atmospheric Administration (NOAA) created GLDAS (Rodell et al., 2004 ) ( http://disc.sci.gsfc.nasa.gov/hydrology/data-holdings ). TRMM is a two-country collaboration between NASA and Japan’s National Space Development Agency that gathers information related to tropical and subtropical precipitation (Tropical Rainfall Measuring Mission. 2011). The TRMM mission became defunct in 2015 and was succeeded by the Global Precipitation Mission (GPM), with some TRMM products continuing with GPM, such as TRMM-3B43 (Huffman et al., 2007 ). TRMM-3B43-Monthly gives the best latitudinal precipitation estimate and is available at 0.25° spatial resolution with a 1 month granule size. Data has been collected from 1998 to the present and covers 50° N to 50° S ( https://disc.gsfc.nasa.gov/datasets/TRMM_3B43_7/summary ). Kogan (1990) proposed and modified NDVI to VCI to normalize NDVI values relative to their minimum and maximum. We estimated VCI from NDVI using one of Terra Moderate Resolution Imaging Spectroradiometer (MODIS)’s standard products, the MODIS Vegetation Indices (MOD13C2). It has a temporal and spatial resolution of 1 month and 500 m, respectively. TRMM data and MOD13C2 have the same temporal coverage. The MODIS NDVIs are processed from atmospherically rectified two-way surface reflectance values that have been suppressed for cloud shadows, heavy aerosols, clouds, and water (Didan, 2015). The MODIS NDVI product can be accessed from the Oak Ridge National Laboratory Distributed Active Center ( http://daac.ornl.gov/MODIS/ ). To estimate DRT, we used the Global Monthly GPP from an Improved Light Use Efficiency (LUE) Model (GMPILUEM) between 1982–2016 with an 8 km spatial resolution (Madani and Parazoo, 2020 ). This GPP product uses climate data from the Modern-Era Retrospective Analysis for Research and Applications Version 2, canopy and fraction of photosynthetically active radiation data from Global Inventory Modeling and Mapping Studies (GIMMS 3g), and improved LUE based on flux tower data. The dataset can be downloaded from https://daac.ornl.gov/cgi-bin/dsviewer.pl?ds_id=1789 . All dataset spatiotemporal resolution was converted to 0.05° × 0.05° (monthly). Detailed information about this study’s dataset is shown in Table 1. 3. Methods 3.1 Potential evapotranspiration In this study, we calculated PET using the PT method (Eq. ( 1 )) (Priestley and Taylor, 1972 ) because it only requires wind and relative humidity climate data. Additionally, PT model PET estimates in tropical areas were found to be acceptable in Gunston and Batchelor, ( 1983 ). The PET is estimated as shown below: $$\lambda \times PET=\alpha \times \frac{\varDelta }{\varDelta +\gamma }\times ({R}_{n}-G) 1$$ where Δ is the slope vapor pressure curve (kPa/˚C); R n is net radiation (MJ/m 2 /day); γ is a psychrometric constant (kPa/˚C); G is the soil heat flux (MJ/m 2 /day); λ is the latent heat of vaporization (MJ/kg); and α is the PT coefficient, usually with a default value of 1.26 (Priestley and Taylor, 1972 ). 3.2 Drought indices Kogan (1990) proposed the VCI equation using current, minimum, and maximum NDVI values for each pixel: $${VCI}_{amy}=\frac{{NDVI}_{amy}-{NDVI}_{a,min}}{{NDVI}_{a,max}{-NDVI}_{a,min}}\times 100 2$$ where VCI amy is the value of VCI assigned to pixel a for the duration of month m for year y ; NDVI amy is a value of the monthly NDVI for pixel a for month m and year y ; and NDVI a, min and NDVI a, max are the multiyear minimum and maximum NDVI, respectively, corresponding to pixel a . The observed values’ resulting percentage is positioned between the maximum and minimum values of prior years. Higher and lower values signify good and bad vegetation state conditions, respectively. SPEI is a meteorological drought index based on the log-logistic distribution of the difference between precipitation and PET (Vicente-Serrano et al., 2010 ). To estimate SPEI, the monthly difference between precipitation and PET is calculated first: $${d}_{i}={P}_{i}-{PET}_{i} 2$$ where d i is the difference between precipitation (P i ) and PET in month i . Next, the probability density function of the log-logistic distributed variable f(x), which has three parameters (Singh et al., 1993 ), is shown as: $$f\left(x\right)=\frac{\theta }{\omega }{\left(\frac{x-\phi }{\omega }\right)}^{\theta -1}{\left[{(\frac{x-\phi }{\omega }+1)}^{\theta }\right]}^{-2} 4$$ $$\theta =\frac{2{w}_{1}-{w}_{0}}{6{w}_{1}-{w}_{0}-6{w}_{2}} 5$$ $$\omega =\frac{({w}_{0}-{2w}_{1})\theta }{\varGamma (1+{\theta }^{-1})\varGamma (1-{\theta }^{-1})} 6$$ $$\phi ={w}_{0}- \omega \varGamma (1+{\theta }^{-1})\varGamma (1-{\theta }^{-1}) 7$$ $${w}_{i}=\frac{\sum _{i=1}^{n}{d}_{i}{(1-\frac{i-0.35}{n})}^{i}}{n} 8$$ where \(\omega\) , \(\theta\) , and \(\phi\) are scale, shape, and origin parameters, respectively, for d values in the range ( \(\theta\) >d<∞); \(\varGamma\) is the gamma function; \({w}_{i}(i=\text{0,1},2\dots )\) are probability-weighted moments for order i ; and n is sample size. Following this, we can estimate the probability distribution function of d as: $$F\left(x\right)={\left[{\left(\frac{\omega }{x-\phi }\right)}^{\theta }+1\right]}^{-1} 9$$ Finally, SPEI can be estimated by converting F(x) into corresponding SPEI 1-month Z-standardized normal values (Abramowitz and Stegun, 1965 ): \(SPEI=\sqrt{-2\text{ln}(1-F(x\left)\right)}+\frac{2.515517+0.802853+0.010328{W}^{2}}{1+1.432788W+0.189269{W}^{2}+0.001308{W}^{3}}\) \(F\left(x\right)\ge 0.5\) (10) \(SPEI=\sqrt{-2\text{ln}\left(F\right(x\left)\right)}+\frac{2.515517+0.802853+0.010328{W}^{2}}{1+1.432788W+0.189269{W}^{2}+0.001308{W}^{3}}\) \(F\left(x\right)\le 0.5\) (11) According to SPEI classification (McKee et al., 1993 ), the drought severity scale from − 2 to 2 is divided into 7 levels, with less than − 2 indicating extreme drought and larger than 2 indicating extreme wet conditions. With the same drought severity scale as SPEI, VCI has only 5 levels (Kogan, 1995 ). Detailed information about their classifications can be found in Table 2. 3.3 Drought Recovery Time (DRT) We determined DRT using a combination of monthly drought index (SPEI or VCI) and GPP values for each pixel in the 14-year period from 2003 to 2016. A drought starts when SPEI \(\le\) − 1 (VCI \(\le\) 30), and ends when SPEI \(>\) − 1 (VCI \(>\) 30). These conditions have to persist for at least 3 months for it to be considered a drought event. After defining a drought event using SPEI or VCI, DRT can be estimated based on pre-drought and post-drought GPP values. The pre-drought GPP is described as the 14-year GPP average for a particular month without considering drought events (Liu et al., 2019 ). It shows the basic metabolism conditions of an ecosystem without drought events. To determine pre-drought GPP, we removed all values in tandem with an SPEI \(\le\) − 1 (VCI \(\le\) 30), and only used the mean of the remaining monthly values. For example, the January pre-drought GPP is represented by a single mean value for every January without a drought over the 14-year period. That pattern is repeated for all 12 months across the same time frame. Post-drought GPP refers to real GPP values including all drought events. In contrast with the pre-drought GPP, in which the mean monthly value is the same for each year, post-drought GPP indicates actual values (Liu et al., 2019 ; Schwalm et al., 2017 ). DRT is the time taken after a drought event has concluded in tandem with the time taken for the post-drought GPP to exceed the pre-drought GPP. It is the difference between the drought recovery ending time (post-drought GPP returning to pre-drought GPP) and the drought ending time (Liu et al., 2019 ; Schwalm et al., 2017 ). 3.4 Boruta algorithm We used the Boruta algorithm to examine which parameters are most important when determining DRT. The Boruta method selects variables and ranks them in order of importance while rejecting parameters that do not improve - or adversely affect - the model’s accuracy. Boruta operates by initially adding randomness to a dataset by creating shuffled duplicates of all parameters. These are termed ‘shadow parameters’ (Kursa and Rudnicki, 2010 ). The extended dataset is then trained with a random forest classifier using decision trees to select appropriate class. The appropriate class is reached by applying a measure that determines each parameter’s importance. A higher result translates to more important class. The algorithm performs iterations where it checks whether the real parameter has a higher importance than the best shadow feature at every stage. This is done by comparing “z scores”; the real parameter must have a higher z score than the maximum shadow parameter z score. In this process, parameters considered unimportant are eliminated. The algorithm terminates when all parameters are confirmed or rejected or when random forest runs are exhausted. The equation and further technical breakdown can be found in Prasad et al. ( 2019 ). 4. Results And Discussion 4.1 Seasonal patterns and trends for drought-related variables To elucidate seasonal drought variation in Lake Victoria basin, we used six variables: rainfall (mm/month); PET (mm/month); GPP (g/m 2 /d); SM (m 3 /m 3 ) for depths of 0–10 cm, 10–40 cm, 40–100 cm, and 100–200 cm; VCI; and SPEI (Fig. 2). We studied four seasons, DJF, MMA, JJA, and SON, and used seasonal mean values from 2003 to 2016 (Table 3). As shown in Fig. 2, we found a clear seasonal rainfall pattern, with large amounts of rainfall (151.79 mm/month) in MAM and the least (53.67 mm/month) in JJA. Although the different season PET did not vary greatly (175.00 to 180.11 mm/month), it increased significantly in the northeastern part of the basin during DJF. GPP was the same for DJF (6.69 g/m 2 /d) and MMA (6.79 g/m 2 /d) though it dropped significantly in JJA (5.46 g/m 2 /d) and SON (5.38 g/m 2 /d), particularly in the basin’s southern and eastern areas. 0 to 100 cm SM did not change significantly except for JJA, when the entire basin experienced very low SM levels (0.20 m 3 /m 3 ) and MMA, when the soil moisture was quite high (0.27 m 3 /m 3 ). From 100 to 200 cm, SM was high for all seasons (0.27 m 3 /m 3 ) compared to the average SM across seasons from 0–100 cm (0.23 m 3 /m 3 ). SM from 100 to 200 cm remained similar for all seasons, including JJA, even though that season had the lowest soil moisture from 0–100 cm. VCI was significantly high and did not vary greatly between DJF (63%) and MMA (69%). JJA and SON had very low VCI, with SON (44%) having a much lower VCI than JJA (46%). VCI was generally low in the southern and eastern parts of the basin during JJA and SON. SPEI was high during MMA (0.47) and SON (0.23). A significant SPEI drop was observed for DJF (-0.11), and it was low throughout the entire basin in JJA (-0.90). Generally, JJA showed low rainfall, GPP, SM, VCI, and SPEI, the driest season. SM from 0 to 100 cm increased when rainfall increased and decreased when there was less rainfall. In contrast, SM from 100cm to 200cm was significantly high for each season. This difference could be attributed to maize’s root depth. Its fibrous roots reach a depth of up to 100 cm and draws most of its soil moisture from that depth (Leenaars et al., 2018 ). VCI and SPEI showed similar results, with very low values in JJA. SPEI was quite high in SON, however, whereas VCI was quite low. This can be explained by VCI only considering a greenness index, whereas SPEI also accounts for climate factors (rainfall and PET). While precipitation and PET have cumulative effects on vegetation conditions, these is a time lag (Gebrehiwot et al., 2011 ; ZHANG et al., 2013 ). Figure 3 shows Lake Victoria Basin climate variable trends from 2003 to 2016. The blue and red colors represent decreasing and increasing trends, respectively. The dark points mark significant (p < 0.1) decreased or increased areas. Rainfall decreased from 2003 to 2006 in the west basin along with a significant (p < 0.1) decrease trend in the mid-west area. A rainfall increase was found in the east with additional pockets showing significant increases in the east and southeast. PET increased significantly in the eastern part of the basin while the far west and southwestern basin area have pockets indicating decreasing PET. GPP in the northern portions of the basin increased while it decreased in the southern area. Soil moisture from 0–10 cm, 10–40 cm and 40–100 cm decreased, with a significant incremental decrease in the east. There is an overall significant (p < 0.1) soil moisture increase, however, from 100–200 cm throughout the basin. Similar to GPP, VCI’s trends are evenly distributed throughout the basin with scattered significant (p < 0.1) increases in the western and northeastern areas. SPEI decreases over almost the basin’s entirety with a significant trend decrease (p < 0.1) in the northwest and western portions. Overall, decreasing rainfall and increasing PET in the western basin led to decreasing SPEI. Although rainfall in the eastern basin increased from 2003 to 2006, the increasing PET and decreasing SM from each depth effects vegetation condition and results in a decreasing GPP and VCI. This is especially true in the basin’s southeast area. 4.2 Drought characterization 4.2.1 Annual drought conditions To assess drought conditions in the study area, we estimated SPEI and VCI annual spatial distribution (Fig. 4). Generally, mean SPEI and VCI values indicate that near-normal conditions imply a potentially stable canopy cover and greenness. As shown in Fig. 4a, SPEI mean annual spatial distribution showed near-normal conditions with very moderate drought tendencies. For example, in 2007, the entire basin was almost normal. But drought prevalence has progressed since then, with conditions peaking in 2016 when the majority of the basin experienced moderate to severe drought conditions. VCI mean annual spatial distribution shows almost no drought, with light to moderate drought occurring in the southeastern part of the basin (Fig. 4b). 2004, 2005, 2006, and 2016 had the largest spatial coverage (almost the entirety of the basin) for drought. In other years, however, the western side was not affected. 2005, 2009, 2011, and 2016 have the lowest VCI values, indicating moderate to severe drought conditions on the eastern outskirts of the study area. 2016 was the driest year, with both the lowest VCI value and largest drought spatial coverage. The highest VCI value and largest non-drought spatial coverage was in 2007. This is consistent with the observed SPEI drought conditions, which indicates that the driest and wettest years were 2016 and 2007, respectively. SPEI and VCI differed spatially in 2006 and 2015, presumably because of a potential temporal lag between VCI and SPEI. For example, 2006 SPEI has high values even though VCI in the same year is low. This is because VCI is an indicator of greenness, whereas SPEI considers PET and rainfall climate factors. That fundamental difference in approach implies a time lag in the results (Fig. 4); the low 2005 SPEI value led to the low 2006 VCI value. The same phenomenon can be found in 2015, when the high VCI value can be explained by the high SPEI value in 2014, not the SPEI value in 2015. 4.2.2 Monthly drought conditions We calculated monthly SPEI and VCI spatial distributions to assess detailed monthly drought variation with a scale denoting drought severity from “extreme wet” to “extreme drought” (Fig. 5). The results shown in Fig. 2, Fig. 4, and Fig. 5 are consistent with the evidence from several reports referenced in Table 4. From 2003 to 2016, July has consistently experienced the most severe to extreme drought conditions. Generally, JJA has the most significant drought conditions, which are consistent with the results shown in Fig. 2, showing the most drought severity occurring in JJA. DJF also show significant drought conditions, especially in the northern parts of the basin. MAM and SON show near normal conditions with occasional oddly scattered spatial instances of drought. 2016 and 2009 had the highest drought severity conditions, which is consistent with average SPEI 2016 values and average 2009 VCI values, respectively. This result could be due to the different times in which the basin receives rainfall, as a rainfall deficit tends to correlate with drought conditions and vice versa. To evaluate drought condition results, we analyzed recorded drought years for different countries in the study area, as shown in Table 4. Burundi, located in the southwestern part of the area, has almost 4 million people in the basin, and 95% of the country’s population is rural. Thus, drought impacts are quite severe (Gebremeskel et al., 2019 ; Yao et al., 2014 ). According to drought event records (Table 4), the drought progressed from 2003 to 2005 and again from 2008 to 2010, mostly affecting the northeastern parts of Burundi (East African Community (EAC), 2010 ). This finding matches the data shown in both Fig. 4 and Fig. 5, demonstrating moderate to extreme drought conditions in the southwestern part of the study area. For Western Kenya, located in the eastern and northeastern parts of the study area, drought events occurred in 2004 and affected 3.3 million Kenyans. Furthermore, as shown in both Fig. 4 and Fig. 5, Western Kenya suffered repeated drought events in 2005, 2008, 2010, 2012, 2014, and 2016 (Agutu et al., 2017 ; Ayugi et al., 2020 ; East African Community (EAC), 2010 ; Gebremeskel et al., 2019 ; Nyaoro et al., 2016 ; Opiyo et al., 2015 ; Schmidt et al., 2017 ). As a result, three major provinces, Nyanza, Western, and Rift Valley, bore the brunt of the drought impact (Awange et al., 2013 ). The 2003 drought event in Rwanda in the western basin was characterized by below-average rainfall. It disproportionately affected rural areas in Rwanda’s semi-arid east, where people are much more dependent on rainfall for both crops and animal husbandry (East African Community (EAC), 2010 ). This is consistent with the data in Fig. 4 and Fig. 5, which show moderate to extreme drought events in the western basin. The drought record in Tanzania, the southern part of the basin, shows drought first occurred in 2003, affecting the regions of Kagera, Mwanza, and Mara. In addition, as shown in Fig. 4 and Fig. 5, drought conditions in this region recurred in 2004, 2006, and 2011 but not in 2007. The failure of short-season rains caused severe drought in late 2005 and early 2006 (Bhaga et al., 2020 ). In early 2007, this region experienced heavy rains, (Fig. 5), with moderate wet conditions in January (Bhaga et al., 2020 ). Uganda lies in the northern portion of the study area, with some parts in the western, central, and eastern areas. More than 600,000 people were affected by drought as of 2005 in the Masaka and Mpigi districts (Hakuza and Waita, 2008 ). Droughts recurred in 2008 and 2010 and affected similar areas. For this reason, farming communities in central, eastern and southwestern Uganda suffered economic and financial damage (Hakuza and Waita, 2008 ). The drought variation’s spatial extent is shown in Fig. 5. It is notable that Fig. 5 shows a one-month delay in ‘dry’ conditions predominant from June to August using the SPEI index and from July to September, even extending into October, with the VCI index. This is because SPEI relies on meteorological data whereas VCI relies on vegetation conditions, implying drought is first observed meteorologically before it’s observed in vegetation. 4.3 Drought duration time (DDT) and drought recovery time 4.3.1 Drought event identification based on drought severity To understand drought event history based on SPEI and VCI results, we calculated the percentage of different drought events based on severity (Table 3). Spatial distribution is shown in Fig. 6. The most moderate SPEI drought events ( \(-1.5cript>\) ) were in the southern part of the study area (Fig. 6). Most of the study area has up to 70% possibility of experiencing moderate drought and a 10 to 20% possibility of experiencing a severe drought ( \(-2cript>\) ). Extreme drought events ( \(\text{S}\text{P}\text{E}\text{I}\le -2)\) mostly occurred in the eastern and northeastern tip of the study area (Fig. 6). Overall trends indicate that the basin has a higher spatial area percentage prone to moderate drought. Using VCI data, 50 to 80% of the study area may experience a moderate drought ( \(20cript>\) ) except for regions in the southeast and southern tip, where the percentage is only 10 to 30%. VCI-predicted severe drought ( \(10cript>\) is similar to SPEI. The majority of the study area has a 10 to 20% chance of experiencing extreme drought ( \(0cript>\) ) except for the southeast region, which shows a 40 to 70% likelihood (Fig. 6). Both VCI- and SPEI-based percentage coverage trends for moderate (M), severe (S), and extreme (E) drought are similar where the spatial drought coverage percentage decreases from moderate to severe. There are some differences, however, in the southeast region for SPEI_M, VCI_M, and SPEI_E, VCI_E which show an inverse trend. This is because VCI indicates drought based on actual vegetation conditions whereas SPEI indicates drought based on meteorological conditions. Overall, more moderate and severe drought events occurred in the southern part of the study area; the eastern and western parts of the basin experienced extreme drought events. As previously discussed, the western area is mostly farming communities in Uganda, Tanzania, Rwanda, and Burundi, including large food crop (maize) belts in central and eastern Uganda. The areas prone to moderate, severe and extreme drought events (Fig. 6) should be considered high-risk drought areas and likely future bottlenecks for water balance. Drought mitigation strategies are imperative in these areas. 4.3.2 Drought duration and recovery time based on SPEI and VCI Using SPEI and VCI drought indices, we calculated and compared mean DDT and DRT spatial distributions in the Lake Victoria Basin from 2003 to 2016 (Fig. 7). The mean SPEI-DDT ranges from 2 to 4 months with a 2.45-day average value. The northeastern and southwestern study area were most likely to suffer from meteorological drought (i.e., SPEI) for at least 3 months, compared to 2 months in other areas. The VCI-DDT spatial distribution (2.97 months) was similar to SPEI-DDT (2.45 months) except for the southeast study area, where VCI-DDT is 2 months longer than SPEI-DDT. This means the southeastern part of the basin is more likely to suffer a long agricultural drought. The SPEI-DRT mean value is 2.02 months, though there are a few instances in the northwest basin area where SPEI-DRT took more than 8 months (black). The VCI-DRT mean value is 1.63 months, which is 0.39 months shorter than the SPEI-DRT mean value. The study area’s mean DDT is higher than the mean DRT, though the mean DDT and mean DRT are almost the same in the southern area, particularly in the southwest. The mean DRT significantly exceeded the mean DDT in the western and northwestern parts of the basin (in black), which show a mean DRT of 8 months compared to three months for DDT. Our results are consistent with Schwalm et al. ( 2017 ), who showed that most drought events in the study area required less than 6 months to recover, though parts of the western basin needed 7–12 months. This long mean DRT puts the western and northwestern basin regions at risk of excessively long drought periods. These include the semi-arid eastern parts of Rwanda, northeastern regions of Burundi, communities in the Kagera region of Tanzania, and central, and southwestern parts of Uganda. It is imperative to pay attention to areas with long DRTs because the northeastern (i.e., Nyanza, Western, and Rift Valley provinces) and southwestern (i.e., Uganda, Tanzania, Rwanda, and Burundi) study areas are mostly cropland (Fig. 1 ). These regions take 4 to 5 months to recover from drought events. Several sub-basins in the western and northern study area create microclimates that potentially affect land–atmosphere interactions. For example, the catchment’s western portion is in the purview of the Kagera sub-basin, which encompasses Uganda, Tanzania, Kenya, Rwanda, and Burundi. Countries such as Uganda, Rwanda, and Burundi lie entirely in the Nile basin, which includes the Victoria basin. The various microclimates created by the sub-basins influence drought characteristics. The Great Rift Valley in the western part of the basin has several lakes and rivers, such as Lake Albert, Lake Edward, and Lake Kivu. These are connected by river systems west of the Victoria basin and affect the climate. The Great Rift Valley also passes close to the eastern side of the basin, but has no significant water bodies. 4.4 Factors affecting DRT We ranked the importance of DDT, SM depths, PET, GPP, rainfall, and drought indices on DRT (Fig. 8). We selected and ranked these factors using a “z score” for SPEI-DRT and VCI-DRT, where a higher value means the factor is more important. The shadow Min, shadow Mean and shadow Max (blue box plots) add randomness to the data to allow for more precise parameter ranking. The green box plots show that each features’ z score is higher than the shadow value, which means they are all significantly important in determining SPEI-DRT and VCI-DRT. DDT is the most important parameter for determining DRT for SPEI-DRT followed by GPP, PET, SPEI, and rainfall. Soil moisture, though still relevant, is the least important parameter for DRT determination. DDT is also the most crucial parameter for determining DRT for VCI-DRT followed by VCI and rainfall. Soil moisture is again classified as the least important parameter, though still important. Based on these results, DDT is the most important factor for determining DRT, regardless of drought indices. We examined DRT spatial distribution after 1, 2-, 3-, 4-, and 5-month drought events to understand the DDT and DRT relationship (Fig. 9). After a one-month drought, SPEI-DRT (SPEI_DRT_1) shows that the entire basin needs to recover from the drought event, with DRTs ranging from 1 to 4 months with some scattered pockets requiring up to 6 months. For a 2-month drought, SPEI_DRT_2, less areas need to recover from drought than for the SPEI_DRT_1 scenario, though there are some isolated pockets in the far east study area that require more than 8 months. SPEI_DRT_3 indicates similar DRTs to SPEI_DRT_2, but the area requiring longer time was larger than for SPEI_DRT_1 or SPEI_DRT_2. As larger basin percentages do not experience 4- or 5-month drought events, SPEI_DRT_4 shows only the southwest and northeast portions of the study area requiring drought recovery, and SPEI_DRT_5 indicates that almost all of the study area does not require any DRT other than sparce pockets in the south. VCI-DRT results are similar to SPEI-DRT except for the DRTs after 4- and 5-month droughts. Results show a significant number of pockets in the northwest part of the basin that need to recover from drought. These were not present in SPEI_DRT_4. VCI_DRT_5 shows the southeastern parts of the study area require time to recover from drought whereas none is needed for SPEI_DRT_5. In the western portion of the study area, SPEI_DRT_1 to SPEI_DRT_4 trends indicate that as the number of drought months increase, drought recovery time also increases. Schwalm et al. ( 2017 ) found similar results. Long droughts affect the water balance, hydrometeorological cycles, and natural land–atmosphere relationships. Thus, it makes sense that long droughts have long recovery times. High PET correlates with a low DRT. This can be attributed to Lake Victoria’s natural hydrological cycle, where a high PET yields high precipitation levels and wards off drought; low PET will allow little to no precipitation to reach the basin (Avanzi et al., 2019 ). As noted in Schwalm et al. ( 2017 ), low SPEI values correlate with a high DRT. Low SPEI values indicate severe drought conditions from which it often takes a long time to recover. Low GPP implies long DRTs because it indicates an unhealthy plant eco-system requiring a long recovery time. Rainfall has a negative relationship with DRT because rainfall encourages plant growth; thus, more rainfall results in a shorter DRT. He et al. ( 2018 ) also found that more rainfall shortened the ecosystem’s DRT. Overall, increasing DDT is the most important factor influencing DRT. Tropical environments take a long time to recover after droughts in dry conditions. Our results are consistent with Wright et al. ( 2002 ), who found that drought recovery is more likely to occur in wet conditions. Furthermore, wet conditions with high PET and high levels of rainfall shortened DRT, whereas dry conditions, with a large SPEI values and long DDTs, led to low GPP and lengthened DRT. This also agrees with Schwalm et al. ( 2017 ). 5. Conclusions Access to accurate drought duration and recovery time information is vitally important in drought-prone areas used for agricultural purposes. Using grid-based data with high resolution, we investigated seasonal patterns in drought-related variables, identified drought events, and examined both SPEI- and VCI-based DRTs in the Lake Victoria Basin from 2003 to 2016. This study is the first to determine meteorological and agricultural DRTs for the Lake Victoria Basin on a 0.05° monthly scale. Of the four seasons, JJA was the driest, with lower values of rainfall, GPP, SM, VCI, and SPEI. Decreasing GPP and VCI is caused by increasing PET and decreasing SM from each depth in the basin’s eastern area. Decreasing SPEI is due to decreasing rainfall and increasing PET in the basin’s western area. Drought indices SPEI and VCI showed that 2016 and 2007 were Lake Victoria Basin’s driest and wettest years in the study range. Meteorological drought calculations showed that moderate droughts occurred at higher frequency in the southeastern part of the basin, whereas the northeastern and mid-western areas were more likely to suffer from extreme drought events. Agricultural drought measurements showed that extreme drought events occurred at higher frequency in the basin’s southern areas. On average, SPEI-based DRT (2.02 months) was longer than VCI-based DRT (1.63 months). SPEI-DRT and VCI-DRT showed similar spatial distribution though SPEI-based DRT (2.02 months) was longer than VCI-based DRT (1.63 months) on average. DDT is the most important parameter for determining DRT, though regions with higher PET, SPEI, GPP, and precipitation values are also associated with shorter recovery times. These results improve understanding of drought on an ecosystem level. Nevertheless, a global DRT product with high accuracy and good spatial and temporal resolution remains challenging, and requires additional investigation. Declarations Acknowledgments We acknowledge the satellite and reanalysis data freely provided by the Land Cover (LC) project of the European Space Agency (ESA) Climate Change Initiative (CCI), the NASA Goddard Earth Sciences Data and Information Services Center (GES DISC), the Oak Ridge National Laboratory Distributed Active Archive Center (ORNL DAAC), and the Land Processes Distributed Active Archive Center (LP DAAC) from NASA Earth Observing System Data and Information System (EOSDIS). This research was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (NRF-2019R1A2B5B01070196). References Aadhar S, Mishra V (2017) Data Descriptor: High-resolution near real-time drought monitoring in South Asia. Sci Data 4:1–14. https://doi.org/10.1038/sdata.2017.145 Abramowitz M, Stegun IA (1965) Handbook of Mathematical Functions, with Formulas, Graphs, and Mathematical Tables. Dover Publications AghaKouchak A (2015) A multivariate approach for persistence-based drought prediction: Application to the 2010–2011 East Africa drought. J Hydrol 526:127–135. https://doi.org/10.1016/j.jhydrol.2014.09.063 Agutu NO, Awange JL, Zerihun A, Ndehedehe CE, Kuhn M, Fukuda Y (2017) Assessing multi-satellite remote sensing, reanalysis, and land surface models’ products in characterizing agricultural drought in East Africa. Remote Sens Environ 194:287–302. https://doi.org/10.1016/j.rse.2017.03.041 Ahmadi B, Ahmadalipour A, Moradkhani H (2019) Hydrological drought persistence and recovery over the CONUS: A multi-stage framework considering water quantity and quality. Water Res 150:97–110. https://doi.org/10.1016/j.watres.2018.11.052 Ahmadi B, Moradkhani H (2019) Revisiting hydrological drought propagation and recovery considering water quantity and quality. Hydrol Process 33:1492–1505. https://doi.org/10.1002/hyp.13417 Anderegg WRL, Konings AG, Trugman AT, Yu K, Bowling DR, Gabbitas R, Karp DS, Pacala S, Sperry JS, Sulman BN, Zenes N (2018) Hydraulic diversity of forests regulates ecosystem resilience during drought. Nature 561:538–541. https://doi.org/10.1038/s41586-018-0539-7 Avanzi F, Rungee J, Maurer T, Bales R, Ma Q, Glaser S, Conklin M (2019) Evapotranspiration feedbacks shift annual precipitation-runoff relationships during multi-year droughts in a Mediterranean mixed rain-snow climate. Hydrol Earth Syst Sci Discuss 1–35. https://doi.org/10.5194/hess-2019-377 Awange JL, Anyah R, Agola N, Forootan E, Omondi P (2013) Potential impacts of climate and environmental change on the stored water of Lake Victoria Basin and economic implications. Water Resour Res 49:8160–8173. https://doi.org/https://doi.org/10.1002/2013WR014350 Awange JL, O.O (2006) Lake Victoria: Ecology, Resources, Environment. Springer-Verlag, Heidelberg. https://doi.org/10.1007/978-1-4020-9726-3_12 Ayugi B, Tan G, Rouyun N, Zeyao D, Ojara M, Mumo L, Babaousmail H, Ongoma V (2020) Evaluation of meteorological drought and flood scenarios over Kenya, East Africa. Atmosphere (Basel). 11. https://doi.org/10.3390/atmos11030307 Banerjee O, Bark R, Connor J, Crossman ND (2013) An ecosystem services approach to estimating economic losses associated with drought. Ecol Econ 91:19–27. https://doi.org/10.1016/j.ecolecon.2013.03.022 Bhaga TD, Dube T, Shekede MD, Shoko C (2020) Impacts of climate variability and drought on surface water resources in sub-saharan africa using remote sensing: A review. Remote Sens 12:1–34. https://doi.org/10.3390/rs12244184 East African Community (EAC) (2010) Lake Victoria Water Supply and Sanitation Program Phase II_Appraisal Ford TW, Labosier CF (2017) Meteorological conditions associated with the onset of flash drought in the Eastern United States. Agric For Meteorol 247:414–423. https://doi.org/https://doi.org/10.1016/j.agrformet.2017.08.031 Funk C, Hoell A, Shukla S, Bladé I, Liebmann B, Roberts JB, Robertson FR, Husak G (2014) Predicting East African spring droughts using Pacific and Indian Ocean sea surface temperature indices. Hydrol Earth Syst Sci 18:4965–4978. https://doi.org/10.5194/hess-18-4965-2014 Gebrehiwot T, van der Veen A, Maathuis B (2011) Spatial and temporal assessment of drought in the Northern highlands of Ethiopia. Int J Appl Earth Obs Geoinf 13:309–321. https://doi.org/https://doi.org/10.1016/j.jag.2010.12.002 Gebremeskel G, Tang Q, Sun S, Huang Z, Zhang X, Liu X (2019) Droughts in East Africa: Causes, impacts and resilience. Earth-Science Rev 193:146–161. https://doi.org/10.1016/j.earscirev.2019.04.015 Gunston H, Batchelor CH (1983) A comparison of the Priestley-Taylor and Penman methods for estimating reference crop evapotranspiration in tropical countries. Agric Water Manag 6:65–77. https://doi.org/10.1016/0378-3774(83)90026-4 Hakuza A, Waita J (2008) REVIEW AND ANALYSIS OF EXISTING DROUGHT RISK REDUCTION POLICIES AND PROGRAMMES IN UGANDA Hao Z, AghaKouchak A, Nakhjiri N, Farahmand A (2014) Global integrated drought monitoring and prediction system. Sci data 1:140001. https://doi.org/10.1038/sdata.2014.1 Hao Z, Singh VP, Xia Y (2018) Seasonal Drought Prediction: Advances, Challenges, and Future Prospects. Rev Geophys 56:108–141. https://doi.org/10.1002/2016RG000549 He B, Liu J, Guo L, Wu X, Xie X, Zhang Y, Chen C, Zhong Z, Chen Z (2018) Recovery of Ecosystem Carbon and Energy Fluxes From the 2003 Drought in Europe and the 2012 Drought in the United States. Geophys Res Lett 45:4879–4888. https://doi.org/10.1029/2018GL077518 Huang L, Zhou P, Cheng L, Liu Z (2021) Dynamic drought recovery patterns over the Yangtze River Basin. Catena 201. https://doi.org/10.1016/j.catena.2021.105194 Huffman GJ, Adler RF, Bolvin DT, Gu G, Nelkin EJ, Bowman KP, Hong Y, Stocker EF, Wolff DB (2007) Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J Hydrometeorol 8:38–55. https://doi.org/10.1175/JHM560.1 The TRMM Multisatellite Precipitation Analysis (TMPA) Kizza M, Rodhe A, Xu CY, Ntale HK, Halldin S (2009) Temporal rainfall variability in the Lake Victoria Basin in East Africa during the twentieth century. Theor Appl Climatol 98:119–135. https://doi.org/10.1007/s00704-008-0093-6 Kogan FN (1995) Application of vegetation index and brightness temperature for drought detection. Adv Sp Res 15:91–100. https://doi.org/https://doi.org/10.1016/0273-1177(95)00079-T KOGAN FN (1990) Remote sensing of weather impacts on vegetation in non-homogeneous areas. Int J Remote Sens 11:1405–1419. https://doi.org/10.1080/01431169008955102 Kursa MB, Rudnicki WR (2010) Feature selection with the boruta package. J Stat Softw 36:1–13. https://doi.org/10.18637/jss.v036.i11 Leenaars JGB, Claessens L, Heuvelink GBM, Hengl T, Ruiperez González M, van Bussel LGJ, Guilpart N, Yang H, Cassman KG (2018) Mapping rootable depth and root zone plant-available water holding capacity of the soil of sub-Saharan Africa. Geoderma 324:18–36. https://doi.org/10.1016/j.geoderma.2018.02.046 Liu L, Gudmundsson L, Hauser M, Qin D, Li S, Seneviratne SI (2019) Revisiting assessments of ecosystem drought recovery. Environ Res Lett 14:114028. https://doi.org/10.1088/1748-9326/ab4c61 Lyon B, Dewitt DG (2012) A recent and abrupt decline in the East African long rains. Geophys Res Lett 39. https://doi.org/10.1029/2011GL050337 Madani N, Parazoo NC (2020) Global Monthly GPP from an Improved Light Use Efficiency Model, 1982–2016. https://doi.org/10.3334/ORNLDAAC/1789 Mailu A (2001) Preliminary assessment of the social, economic and environmental impacts of water hyacinth in Lake Victoria Basin and status of control McKee TB, Doesken NJ, Kleist J (1993) The relationship of drought frequency and duration to time scales, in: Proceedings of the 8th Conference on Applied Climatology. Boston, pp. 179–183 Mishra AK, Singh VP (2010) A review of drought concepts. J Hydrol 391:202–216. https://doi.org/10.1016/j.jhydrol.2010.07.012 Mu Q, Zhao M, Kimball JS, McDowell NG, Running SW (2013) A Remotely Sensed Global Terrestrial Drought Severity Index. Bull Am Meteorol Soc 94:83–98. https://doi.org/10.1175/BAMS-D-11-00213.1 Nyaoro D, Schade J, Schmidt K (2016) Assessing the Evidence: Migration, Environment and Climate Change in Kenya Opiyo F, Wasonga O, Nyangito M, Schilling J, Munang R (2015) Drought Adaptation and Coping Strategies Among the Turkana Pastoralists of Northern Kenya. Int J Disaster Risk Sci 6:295–309. https://doi.org/10.1007/s13753-015-0063-4 Peel MC, Finlayson BL, McMahon TA (2007) Updated world map of the Köppen-Geiger climate classification. Hydrol Earth Syst Sci 11:1633–1644. https://doi.org/10.5194/hess-11-1633-2007 Prasad R, Deo RC, Li Y, Maraseni T (2019) Weekly soil moisture forecasting with multivariate sequential, ensemble empirical mode decomposition and Boruta-random forest hybridizer algorithm approach. Catena 177:149–166. https://doi.org/10.1016/j.catena.2019.02.012 Priestley, C.H.B., Taylor, R.J., 1972. On the Assessment of Surface Heat Flux and Evaporation Using Large-Scale Parameters. Mon. Weather Rev. https://doi.org/10.1175/1520-0493(1972)1002.3.CO;2 Rodell M, Houser PR, Jambor U, Gottschalck J, Mitchell K, Meng CJ, Arsenault K, Cosgrove B, Radakovich J, Bosilovich M, Entin JK, Walker JP, Lohmann D, Toll D (2004) The Global Land Data Assimilation System. Bull Am Meteorol Soc 85, 381–394. https://doi.org/10.1175/BAMS-85-3-381 Schmidt W, Peter Uhe A, Kimutai J, Otto F, Cullen H (2017) The Drought in Kenya, 2016–2017, The Drought in Kenya, 2016–2017. Climate and Development Knowledge Network and World Weather Attribution Initiative Schwalm CR, Anderegg WRL, Michalak AM, Fisher JB, Biondi F, Koch G, Litvak M, Ogle K, Shaw JD, Wolf A, Huntzinger DN, Schaefer K, Cook R, Wei Y, Fang Y, Hayes D, Huang M, Jain A, Tian H (2017) Global patterns of drought recovery. Nature 548:202–205. https://doi.org/10.1038/nature23021 Seneviratne SI, Ciais P (2017) Trends in ecosystem recovery from drought. Nature 548:164–165. https://doi.org/10.1038/548164a Singh VP, Guo H, Yu FX (1993) Parameter estimation for 3-parameter log-logistic distribution (LLD3) by Pome. Stoch Hydrol Hydraul 7:163–177. https://doi.org/10.1007/BF01585596 Sternberg T (2011) Regional drought has a global impact. Nature 472:169–169. https://doi.org/10.1038/472169d Trenberth KE, Dai A, Van Der Schrier G, Jones PD, Barichivich J, Briffa KR, Sheffield J (2014) Global warming and changes in drought. Nat Clim Chang 4:17–22. https://doi.org/10.1038/nclimate2067 Van der Molen MK, Dolman AJ, Ciais P, Eglin T, Gobron N, Law BE, Meir P, Peters W, Phillips OL, Reichstein M, Chen T, Dekker SC, Doubková M, Friedl MA, Jung M, van den Hurk BJJM, de Jeu RAM, Kruijt B, Ohta T, Rebel KT, Plummer S, Seneviratne SI, Sitch S, Teuling AJ, van der Werf GR, Wang G (2011) Drought and ecosystem carbon cycling. Agric For Meteorol 151:765–773. https://doi.org/10.1016/j.agrformet.2011.01.018 Van Dijk AIJM, Beck HE, Crosbie RS, De Jeu RAM, Liu YY, Podger GM, Timbal B, Viney NR (2013) The Millennium Drought in southeast Australia (2001–2009): Natural and human causes and implications for water resources, ecosystems, economy, and society. Water Resour Res 49:1040–1057. https://doi.org/10.1002/wrcr.20123 Vicente-Serrano SM, Beguería S, López-Moreno JI (2010) A Multiscalar Drought Index Sensitive to Global Warming: The Standardized Precipitation Evapotranspiration Index. J Clim 23:1696–1718. https://doi.org/10.1175/2009JCLI2909.1 West H, Quinn N, Horswell M (2019) Remote sensing for drought monitoring & impact assessment: Progress, past challenges and future opportunities. Remote Sens Environ 232:111291. https://doi.org/10.1016/j.rse.2019.111291 Wright JF, Gunn RJM, Winder JM, Wiggers R, Vowles K, Clarke RT, Harris I (2002) A comparison of the macrophyte cover and macroinvertebrate fauna at three sites on the River Kennet in the mid 1970s and late 1990s. Sci Total Environ 282–283, 121–142. https://doi.org/https://doi.org/10.1016/S0048-9697(01)00948-2 Yao Y, Liang S, Li X, Hong Y, Fisher JB, Zhang N, Chen J, Cheng J, Zhao S, Zhang X, Jiang B, Sun L, Jia K, Wang K, Chen Y, Mu Q, Feng F (2014) Bayesian multimodel estimation of global terrestrial latent heat flux from eddy covariance, meteorological, and satellite observations. J Geophys Res Atmos 119:4521–4545. https://doi.org/10.1002/2013JD020864 Yu Z, Wang J, Liu S, Rentch JS, Sun P, Lu C (2017) Global gross primary productivity and water use efficiency changes under drought stress. Environ Res Lett 12. https://doi.org/10.1088/1748-9326/aa5258 Zhang D, Liu X, Bai P (2019) Assessment of hydrological drought and its recovery time for eight tributaries of the Yangtze River (China) based on downscaled GRACE data. J Hydrol 568:592–603. https://doi.org/10.1016/j.jhydrol.2018.11.030 ZHANG F, ZHANG L, WANG X, HUNG J (2013) Detecting Agro-Droughts in Southwest of China Using MODIS Satellite Data. J Integr Agric 12:159–168. https://doi.org/https://doi.org/10.1016/S2095-3119(13)60216-6 Zhou M, Brandt P, Pelster D, Rufino MC, Robinson T, Butterbach-Bahl K (2014) Regional nitrogen budget of the Lake Victoria Basin, East Africa: Syntheses, uncertainties and perspectives. Environ Res Lett 9. https://doi.org/10.1088/1748-9326/9/10/105009 Tables Due to technical limitations, table 1 to 4 is only available as a download in the Supplemental Files section. Supplementary Files Table.pdf Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 05 Aug, 2021 Reviewers invited by journal 02 Aug, 2021 Editor invited by journal 02 Aug, 2021 Editor assigned by journal 02 Aug, 2021 First submitted to journal 01 Aug, 2021 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-774555","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":43357718,"identity":"890a04e8-f33a-4138-96d0-c94acf0ca489","order_by":0,"name":"Yuefeng Hao","email":"","orcid":"","institution":"Sungkyunkwan University - Suwon Campus: Sungkyunkwan University - Natural Sciences Campus","correspondingAuthor":false,"prefix":"","firstName":"Yuefeng","middleName":"","lastName":"Hao","suffix":""},{"id":43357719,"identity":"2e0f69c3-97d1-46fc-a8ab-996dd0fb41b6","order_by":1,"name":"Jongjin Baik","email":"","orcid":"","institution":"Chung-Ang University","correspondingAuthor":false,"prefix":"","firstName":"Jongjin","middleName":"","lastName":"Baik","suffix":""},{"id":43357720,"identity":"4377a9d3-e8b8-4a17-9ae6-fff66afa9be0","order_by":2,"name":"Sseguya Fred","email":"","orcid":"","institution":"Sungkyunkwan University - Suwon Campus: Sungkyunkwan University - Natural Sciences Campus","correspondingAuthor":false,"prefix":"","firstName":"Sseguya","middleName":"","lastName":"Fred","suffix":""},{"id":43357721,"identity":"60f71349-7df1-4354-90ee-909be9f247c7","order_by":3,"name":"Minha Choi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAo0lEQVRIiWNgGAWjYDACCQY2BgYDGwYJHhCPjXgtaSRrYThMghbd2c3PHnwoOG8v2XPGgOFD2WHCWszuHDM3nGFwO3E2b48B44xzxGi5kcMmzWNwO0GOn8eAmbeNeC3n7MFa/pKg5QAjyGHMjMRpSTOTnGGQnDiz51jBwZ5z6cRoSX4m8eGPnb3EmeSND36UWRPWggIOkKh+FIyCUTAKRgEuAAAyQTWxx7+62wAAAABJRU5ErkJggg==","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":true,"prefix":"","firstName":"Minha","middleName":"","lastName":"Choi","suffix":""}],"badges":[],"createdAt":"2021-08-02 03:44:37","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-774555/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-774555/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":12153967,"identity":"c8426fa1-1752-42a8-ad71-1a7ff7e2f558","added_by":"auto","created_at":"2021-08-05 16:37:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1479034,"visible":true,"origin":"","legend":"Land cover map for the Lake Victoria Basin. Bold lines represent country boundaries and dotted lines represent key districts for countries in the basin. Land cover data is from the ‘S2 prototype LC map at 20m of Africa 2016’ (CCI Land Cover. 2017). The map can be downloaded from http://2016africalandcover20m.esrin.esa.int/.","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/1222b889d79758226c6874d5.png"},{"id":12153966,"identity":"489625d9-8298-4d86-912d-b6fea7ea6628","added_by":"auto","created_at":"2021-08-05 16:37:26","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":194002,"visible":true,"origin":"","legend":"Seasonal spatial patterns of rainfall (mm/month); PET (mm/month); GPP (g/m2/d); SM (m3/m3) for depths of 0–10 cm, 10–40 cm, 40–100 cm, and 100–200 cm; VCI; and SPEI in the Lake Victoria Basin based on seasonal mean data from 2003 to 2016.","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/d8b56afa7f601e2680bb7cca.png"},{"id":12153973,"identity":"7df246d6-0a6d-4a1b-9d85-46d834b6428d","added_by":"auto","created_at":"2021-08-05 16:37:27","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":377263,"visible":true,"origin":"","legend":"Spatial trends of climate variables including rainfall, PET, GPP, SM for depths [0-10 cm, 10-40 cm, 40-100 cm, 100-200 cm], VCI, and SPEI) in the Lake Victoria Basin from 2003 to 2016.","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/2975731cb4322ee7b45d68be.png"},{"id":12154218,"identity":"4c1dd075-d847-4415-9a93-2cbd2a7c48c9","added_by":"auto","created_at":"2021-08-05 16:40:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":208645,"visible":true,"origin":"","legend":"Annual spatial distribution and drought characterization using SPEI (a) and VCI (b) indices in the Lake Victoria Basin from 2003 to 2016. Calculations include the mean for all years.","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/f1e050b9436fc92d71117499.png"},{"id":12153972,"identity":"f110942a-f34f-4451-a168-22356e813384","added_by":"auto","created_at":"2021-08-05 16:37:27","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":336780,"visible":true,"origin":"","legend":"Spatial extent and drought characterization using the SPEI and VCI indices for the Lake Victoria Basin from 2003 to 2016 for each month in each year.","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/9aca44516087edc7766c6596.png"},{"id":12153970,"identity":"afa91d16-ac81-4933-ac0a-6da28ebef7f9","added_by":"auto","created_at":"2021-08-05 16:37:27","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":91591,"visible":true,"origin":"","legend":"SPEI- and VCI-based drought calculations. Percentage of the spatial extent indicates moderate (M), severe(S) and extreme(E) drought.","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/21203b7e8b80ab5f4a2ac27e.png"},{"id":12154217,"identity":"dd8b912c-7311-4e30-ad4d-cbf0042c243b","added_by":"auto","created_at":"2021-08-05 16:40:27","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":137630,"visible":true,"origin":"","legend":"Spatial coverage of SPEI- and VCI-based mean drought duration and recovery times from 2003 to 2016 in the Lake Victoria Basin.","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/c6e1a8a4a674190828588b3d.png"},{"id":12153974,"identity":"a46f6962-2b24-4311-98b6-e7034f0b5f17","added_by":"auto","created_at":"2021-08-05 16:37:27","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":131177,"visible":true,"origin":"","legend":"Parameter ranking for SPEI_DRT and VCI_DRT determination using the Boruta Random Forest Parameter Selection Algorithm","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/28e03694921d04a7fca96c62.png"},{"id":13707687,"identity":"76087310-0d17-41a2-b2a0-9227140882d6","added_by":"auto","created_at":"2021-09-17 14:04:14","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3577802,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/985c1f08-f24f-46d5-883d-c68eb9e1318d.pdf"},{"id":12153968,"identity":"9d30cd66-f2f3-4851-9a1d-d09e52efc766","added_by":"auto","created_at":"2021-08-05 16:37:27","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":107129,"visible":true,"origin":"","legend":"","description":"","filename":"Table.pdf","url":"https://assets-eu.researchsquare.com/files/rs-774555/v1/17bcb4f9f99fbb3e7afd0078.pdf"}],"financialInterests":"","formattedTitle":"\u003cp\u003eDetecting Agricultural and Meteorological Drought With Gross Primary Production Recovery Including Spatiotemporal Statistical Analysis in East Africa's Lake Victoria Basin\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAs a devastating hazard, drought not only imposes extensive and long-term effects on the global environmental system, it also causes significant losses to the economy and human life (Mishra and Singh, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Sternberg, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Van Dijk et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). To reduce drought risk and prevent drought under global climate change and global warming (Trenberth et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), several research contributions have been made in the fields of drought monitoring (Aadhar and Mishra, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; West et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) and prediction (AghaKouchak, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Hao et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Those studies considered hydrological and meteorological variables to better understand regional drought event causes and quantification.\u003c/p\u003e \u003cp\u003eOther studies use ecological variables to identify this natural hazard at the ecosystem level (Anderegg et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Banerjee et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Van der Molen et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In particular, drought recovery time (DRT) has received growing research attention in recent years (Ahmadi et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Ahmadi and Moradkhani, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; He et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Huang et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Liu et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Seneviratne and Ciais, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). DRT will likely become longer than the time between drought events (Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Regional drought effects are compounded if a new drought event occurs before recovery from a preceding drought event is complete (Seneviratne and Ciais, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Therefore, accurate DRT assessments are essential for understanding possible ecological risks.\u003c/p\u003e \u003cp\u003eDRT is defined as the time required for a region to fully return to its pre-drought conditions (Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Drought identification and recovery parameter selection are the key steps in DRT calculations. To identify a drought event, previous studies used the meteorological drought index (Standardized Precipitation Evapotranspiration Index, SPEI) and Drought Severity Index (DSI) (Liu et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Yu et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Climate data-based SPEI is easy to estimate for long-term analysis but it is not linked to plant condition (Vicente-Serrano et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Satellite-based DSI includes greenness information with high spatial resolution. However, DSI data is only available from 2000 to 2011 with uncertainties from the satellite (cloud cover, atmospheric aerosols, and low solar illumination) and single input data (Mu et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eMany factors such as water quantity (streamflow and total water storage), water quality (water temperature, turbidity, and dissolved oxygen), ecosystem fluxes (carbon and energy fluxes), and gross primary productivity (GPP) are used in drought recovery assessments (Ahmadi et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; He et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Seneviratne and Ciais, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). For example, the change in total water storage was used to identify hydrological DRT (between 3.6 and 5.7 months) for the Yangtze River in China (Zhang et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Ahmadi et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) analyzed drought recovery by considering both streamflow and water quality parameter changes. They found that the average recovery time in the contiguous United States is around 1.2 months. The time required for carbon and energy fluxes to recover from the 2012 U.S. drought (0.5\u0026ndash;2 months) and the 2003 European drought (1\u0026ndash;2 months) were examined by He et al. (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Based on changes to GPP, global spatiotemporal DRT patterns were examined by Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) and Yu et al. (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). All of these studies focused on the DRT length and response functions. Among the factors, GPP is the most impactful due to its high drought sensitivity and spatiotemporal patterns. It can accommodate increasingly fine spatial resolution and frequent repeat measurements (Schwalm et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) estimated that DRT can be determined at 0.5\u0026deg; spatial resolution and 6 months of temporal resolution from 1901 to 2010 around the world. They found that most of the world can recover from a drought in less than six months. Unlike a conventional drought, a \u003cem\u003eflash drought\u003c/em\u003e typically occurs during warm seasons and can occur more frequently - in one or two months (Ford and Labosier, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Therefore, DRT should be examined in more detailed studies at a high spatiotemporal resolution. Yu et al. (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) calculated global DRT from 2000 to 2011 at 0.5\u0026deg; spatial resolution and 1 month temporal resolution, but the results were different for Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) in terms of spatial pattern and DRT length. As reported by Liu et al. (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), using different methods to define drought events and recovery levels are the key factors contributing to the contradictory conclusions. Despite the large amount of research on meteorological and hydrological DRT, there has been very little work integrating long-term agricultural DRT at a high spatiotemporal resolution in dry regions such as East Africa, where people are largely dependent on rain-fed agriculture and livestock farming (Gebremeskel et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo better understand drought recovery in the Lake Victoria Basin, our objectives are 1) check the seasonal patterns and trends of drought-related variables; 2) capture drought events using meteorological (SPEI) and agricultural (vegetation condition index [VCI]) drought indices from 2003 to 2016; 3) investigate SPEI based- and VCI based-DRT for high spatiotemporal resolution; and 4) examine parameter importance for determining DRT across the Lake Victoria Basin. To the best of our knowledge, this is the first comprehensive study to quantify agricultural DRT at a high spatiotemporal resolution.\u003c/p\u003e"},{"header":"2. Study Area And Dataset","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study area\u003c/h2\u003e \u003cp\u003eAs the second largest freshwater lake in the world, Lake Victoria supports one of the world\u0026rsquo;s poorest and densest populations. Its catchment covers a 194,200 km\u003csup\u003e2\u003c/sup\u003e area and has a total population of 30\u0026nbsp;million people (Mailu, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). Basin agriculture supports more than 70% of the local population (Zhou et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the land cover distribution in the Lake Victoria Basin. Grassland (16%) is predominantly in the west, open water (26%) in the middle, and cropland (36%) in the east. Maize is the main crop (Zhou et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe Victoria Basin is shared by 5 agricultural East African nations: Kenya, Uganda, Tanzania, Rwanda, and Burundi. Kenya is in the northeastern part of the basin, and contains three provinces: Nyanza, Western, and Rift Valley. They are predominantly cropland. The northwest comprises central Ugandan districts Masaka and Mpigi, which are characterized by large farming communities. Tanzania, located in the basin\u0026rsquo;s south, also has a large cropland area. Most of the farmland is located in the Mwanza and Mara provinces, though it also covers the Kagera region, a key region for food production and distribution whose landcover is a mixture of cropland and grassland. Rwanda is located in the western part of the study area and is mainly covered by cropland. Burundi is also in the western basin and is half cropland. According to the K\u0026ouml;ppen-Geiger climate classification (Peel et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), Lake Victoria Basin\u0026rsquo;s climate zone is tropical with four distinct seasons: hot dry (Dec. to Feb., DJF), major rainy (Mar. to May, MMA), short dry (Jun. to Aug., JJA) and short rainy (Sept. to Nov., SON) (Awange JL, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Annual precipitation ranges between 670 and 2,200 mm, with a mean value of 1,202 mm (Kizza et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Several studies discuss droughts in the Lake Victoria Basin during the twentieth century\u0026rsquo;s last two decades (Funk et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Gebremeskel et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Lyon and Dewitt, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Since 2003, drought has continued to occur every 3 years on average, with a significant increase in severity and frequency (Funk et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Drought re-emergence has had devastating effects on local people, flora, and fauna.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Dataset\u003c/h2\u003e \u003cp\u003eTo estimate SPEI, we used global hydrological datasets (Global Land Data Assimilation System Noah Land Surface Model L4 monthly 0.25\u0026deg; \u0026times; 0.25\u0026deg; Version 2.1 [GLDAS_NOAH025_M _V2.1] and Tropical Rainfall Measurement Mission Rainfall Estimate L3 1 month 0.25-degree x 0.25-degree V7 [TRMM-3B43]). First, we used the GLDAS-2.1 net radiation flux, ground heat flux, air temperature, and surface pressure to calculate potential evapotranspiration (PET) based on the Priestley and Taylor (PT) method. Next, we used the GLDAS-based PET and rainfall data from TRMM-3B43 to compute SPEI.\u003c/p\u003e \u003cp\u003eIn this study, we used the GLDAS_NOAH025_M _V2.1 dataset (Beaudoing and Rodell, 2020) for soil moisture (0\u0026ndash;10cm, 10\u0026ndash;40cm, 40\u0026ndash;100cm, and 100\u0026ndash;200cm) data, net radiation flux, ground heat flux, air temperature, and surface pressure. GLDAS combines ground and satellite data via assimilation methods along with land surface models. The National Centers for Environmental Prediction (NCEP), National Aeronautics and Space Administration (NASA), and National Oceanic and Atmospheric Administration (NOAA) created GLDAS (Rodell et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://disc.sci.gsfc.nasa.gov/hydrology/data-holdings\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTRMM is a two-country collaboration between NASA and Japan\u0026rsquo;s National Space Development Agency that gathers information related to tropical and subtropical precipitation (Tropical Rainfall Measuring Mission. 2011). The TRMM mission became defunct in 2015 and was succeeded by the Global Precipitation Mission (GPM), with some TRMM products continuing with GPM, such as TRMM-3B43 (Huffman et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). TRMM-3B43-Monthly gives the best latitudinal precipitation estimate and is available at 0.25\u0026deg; spatial resolution with a 1 month granule size. Data has been collected from 1998 to the present and covers 50\u0026deg; N to 50\u0026deg; S (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://disc.gsfc.nasa.gov/datasets/TRMM_3B43_7/summary\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eKogan (1990) proposed and modified NDVI to VCI to normalize NDVI values relative to their minimum and maximum. We estimated VCI from NDVI using one of Terra Moderate Resolution Imaging Spectroradiometer (MODIS)\u0026rsquo;s standard products, the MODIS Vegetation Indices (MOD13C2). It has a temporal and spatial resolution of 1 month and 500 m, respectively. TRMM data and MOD13C2 have the same temporal coverage. The MODIS NDVIs are processed from atmospherically rectified two-way surface reflectance values that have been suppressed for cloud shadows, heavy aerosols, clouds, and water (Didan, 2015). The MODIS NDVI product can be accessed from the Oak Ridge National Laboratory Distributed Active Center (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://daac.ornl.gov/MODIS/\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo estimate DRT, we used the Global Monthly GPP from an Improved Light Use Efficiency (LUE) Model (GMPILUEM) between 1982\u0026ndash;2016 with an 8 km spatial resolution (Madani and Parazoo, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This GPP product uses climate data from the Modern-Era Retrospective Analysis for Research and Applications Version 2, canopy and fraction of photosynthetically active radiation data from Global Inventory Modeling and Mapping Studies (GIMMS 3g), and improved LUE based on flux tower data. The dataset can be downloaded from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://daac.ornl.gov/cgi-bin/dsviewer.pl?ds_id=1789\u003c/span\u003e\u003c/span\u003e. All dataset spatiotemporal resolution was converted to 0.05\u0026deg; \u0026times; 0.05\u0026deg; (monthly). Detailed information about this study\u0026rsquo;s dataset is shown in Table\u0026nbsp;1.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Methods","content":"\u003cdiv class=\"Section2\" id=\"Sec6\"\u003e\n \u003ch2\u003e\u003cem\u003e3.1 Potential evapotranspiration\u003c/em\u003e\u003c/h2\u003e\n \u003cp\u003eIn this study, we calculated PET using the PT method (Eq.\u0026nbsp;(\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e)) (Priestley and Taylor, \u003cspan class=\"CitationRef\"\u003e1972\u003c/span\u003e) because it only requires wind and relative humidity climate data. Additionally, PT model PET estimates in tropical areas were found to be acceptable in Gunston and Batchelor, (\u003cspan class=\"CitationRef\"\u003e1983\u003c/span\u003e). The PET is estimated as shown below:\u003c/p\u003e\n \u003cdiv class=\"Equation\" id=\"Equ1\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e$$\\lambda \\times PET=\\alpha \\times \\frac{\\varDelta }{\\varDelta +\\gamma }\\times ({R}_{n}-G) 1$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003ewhere \u003cem\u003e\u0026Delta;\u003c/em\u003e is the slope vapor pressure curve (kPa/˚C); \u003cem\u003eR\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e is net radiation (MJ/m\u003csup\u003e2\u003c/sup\u003e/day); \u0026gamma; is a psychrometric constant (kPa/˚C); \u003cem\u003eG\u003c/em\u003e is the soil heat flux (MJ/m\u003csup\u003e2\u003c/sup\u003e/day); \u003cem\u003e\u0026lambda;\u003c/em\u003e is the latent heat of vaporization (MJ/kg); and \u0026alpha; is the PT coefficient, usually with a default value of 1.26 (Priestley and Taylor, \u003cspan class=\"CitationRef\"\u003e1972\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Section2\" id=\"Sec7\"\u003e\n \u003ch2\u003e\u003cem\u003e3.2 Drought indices\u003c/em\u003e\u003c/h2\u003e\n \u003cp\u003eKogan (1990) proposed the VCI equation using current, minimum, and maximum NDVI values for each pixel:\u003c/p\u003e\n \u003cdiv class=\"Equation\" id=\"Equ2\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e$${VCI}_{amy}=\\frac{{NDVI}_{amy}-{NDVI}_{a,min}}{{NDVI}_{a,max}{-NDVI}_{a,min}}\\times 100 \u0026nbsp;2$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003ewhere \u003cem\u003eVCI\u003c/em\u003e\u003csub\u003e\u003cem\u003eamy\u003c/em\u003e\u003c/sub\u003e is the value of VCI assigned to pixel \u003cem\u003ea\u003c/em\u003e for the duration of month \u003cem\u003em\u003c/em\u003e for year \u003cem\u003ey\u003c/em\u003e; \u003cem\u003eNDVI\u003c/em\u003e\u003csub\u003e\u003cem\u003eamy\u003c/em\u003e\u003c/sub\u003e is a value of the monthly NDVI for pixel \u003cem\u003ea\u003c/em\u003e for month \u003cem\u003em\u003c/em\u003e and year \u003cem\u003ey\u003c/em\u003e; and \u003cem\u003eNDVI\u003c/em\u003e\u003csub\u003e\u003cem\u003ea, min\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eNDVI\u003c/em\u003e\u003csub\u003e\u003cem\u003ea, max\u003c/em\u003e\u003c/sub\u003e are the multiyear minimum and maximum NDVI, respectively, corresponding to pixel \u003cem\u003ea\u003c/em\u003e. The observed values\u0026rsquo; resulting percentage is positioned between the maximum and minimum values of prior years. Higher and lower values signify good and bad vegetation state conditions, respectively.\u003c/p\u003e\n \u003cp\u003eSPEI is a meteorological drought index based on the log-logistic distribution of the difference between precipitation and PET (Vicente-Serrano et al., \u003cspan class=\"CitationRef\"\u003e2010\u003c/span\u003e). To estimate SPEI, the monthly difference between precipitation and PET is calculated first:\u003c/p\u003e\n \u003cdiv class=\"Equation\" id=\"Equ3\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e$${d}_{i}={P}_{i}-{PET}_{i} \u0026nbsp;2$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003ewhere d\u003csub\u003ei\u003c/sub\u003e is the difference between precipitation (P\u003csub\u003ei\u003c/sub\u003e) and PET in month \u003cem\u003ei\u003c/em\u003e. Next, the probability density function of the log-logistic distributed variable f(x), which has three parameters (Singh et al., \u003cspan class=\"CitationRef\"\u003e1993\u003c/span\u003e), is shown as:\u003c/p\u003e\n \u003cdiv class=\"Equation\" id=\"Equ4\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e$$f\\left(x\\right)=\\frac{\\theta }{\\omega }{\\left(\\frac{x-\\phi }{\\omega }\\right)}^{\\theta -1}{\\left[{(\\frac{x-\\phi }{\\omega }+1)}^{\\theta }\\right]}^{-2} \u0026nbsp;4$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Equation\" id=\"Equ5\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ5\" name=\"EquationSource\"\u003e$$\\theta =\\frac{2{w}_{1}-{w}_{0}}{6{w}_{1}-{w}_{0}-6{w}_{2}} \u0026nbsp;5$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Equation\" id=\"Equ6\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ6\" name=\"EquationSource\"\u003e$$\\omega =\\frac{({w}_{0}-{2w}_{1})\\theta }{\\varGamma (1+{\\theta }^{-1})\\varGamma (1-{\\theta }^{-1})} \u0026nbsp;6$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Equation\" id=\"Equ7\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ7\" name=\"EquationSource\"\u003e$$\\phi ={w}_{0}- \\omega \\varGamma (1+{\\theta }^{-1})\\varGamma (1-{\\theta }^{-1}) \u0026nbsp;7$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv class=\"Equation\" id=\"Equ8\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ8\" name=\"EquationSource\"\u003e$${w}_{i}=\\frac{\\sum _{i=1}^{n}{d}_{i}{(1-\\frac{i-0.35}{n})}^{i}}{n} \u0026nbsp;8$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\omega\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\theta\\)\u003c/span\u003e\u003c/span\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\phi\\)\u003c/span\u003e\u003c/span\u003e are scale, shape, and origin parameters, respectively, for d values in the range (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\theta\\)\u003c/span\u003e\u003c/span\u003e \u0026gt;d\u0026lt;\u0026infin;); \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\varGamma\\)\u003c/span\u003e\u003c/span\u003e is the gamma function; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({w}_{i}(i=\\text{0,1},2\\dots )\\)\u003c/span\u003e\u003c/span\u003e are probability-weighted moments for order \u003cem\u003ei\u003c/em\u003e; and n is sample size. Following this, we can estimate the probability distribution function of d as:\u003c/p\u003e\n \u003cdiv class=\"Equation\" id=\"Equ9\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ9\" name=\"EquationSource\"\u003e$$F\\left(x\\right)={\\left[{\\left(\\frac{\\omega }{x-\\phi }\\right)}^{\\theta }+1\\right]}^{-1} \u0026nbsp;9$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003eFinally, SPEI can be estimated by converting \u003cem\u003eF(x)\u003c/em\u003e into corresponding SPEI 1-month Z-standardized normal values (Abramowitz and Stegun, \u003cspan class=\"CitationRef\"\u003e1965\u003c/span\u003e):\u003c/p\u003e\n \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u0026nbsp;\u003cspan class=\"mathinline\"\u003e\\(SPEI=\\sqrt{-2\\text{ln}(1-F(x\\left)\\right)}+\\frac{2.515517+0.802853+0.010328{W}^{2}}{1+1.432788W+0.189269{W}^{2}+0.001308{W}^{3}}\\)\u003c/span\u003e\u0026nbsp;\u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(F\\left(x\\right)\\ge 0.5\\)\u003c/span\u003e\u003c/span\u003e (10)\u003c/p\u003e\n \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u0026nbsp;\u003cspan class=\"mathinline\"\u003e\\(SPEI=\\sqrt{-2\\text{ln}\\left(F\\right(x\\left)\\right)}+\\frac{2.515517+0.802853+0.010328{W}^{2}}{1+1.432788W+0.189269{W}^{2}+0.001308{W}^{3}}\\)\u003c/span\u003e\u0026nbsp;\u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(F\\left(x\\right)\\le 0.5\\)\u003c/span\u003e\u003c/span\u003e (11)\u003c/p\u003e\n \u003cp\u003eAccording to SPEI classification (McKee et al., \u003cspan class=\"CitationRef\"\u003e1993\u003c/span\u003e), the drought severity scale from \u0026minus;\u0026thinsp;2 to 2 is divided into 7 levels, with less than \u0026minus;\u0026thinsp;2 indicating extreme drought and larger than 2 indicating extreme wet conditions. With the same drought severity scale as SPEI, VCI has only 5 levels (Kogan, \u003cspan class=\"CitationRef\"\u003e1995\u003c/span\u003e). Detailed information about their classifications can be found in Table\u0026nbsp;2.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Section2\" id=\"Sec8\"\u003e\n \u003ch2\u003e3.3 Drought Recovery Time (DRT)\u003c/h2\u003e\n \u003cp\u003eWe determined DRT using a combination of monthly drought index (SPEI or VCI) and GPP values for each pixel in the 14-year period from 2003 to 2016. A drought starts when SPEI \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\le\\)\u003c/span\u003e\u003c/span\u003e \u0026minus;\u0026thinsp;1 (VCI \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\le\\)\u003c/span\u003e\u003c/span\u003e 30), and ends when SPEI \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\u0026gt;\\)\u003c/span\u003e\u003c/span\u003e \u0026minus;\u0026thinsp;1 (VCI \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\u0026gt;\\)\u003c/span\u003e\u003c/span\u003e 30). These conditions have to persist for at least 3 months for it to be considered a drought event. After defining a drought event using SPEI or VCI, DRT can be estimated based on pre-drought and post-drought GPP values. The pre-drought GPP is described as the 14-year GPP average for a particular month without considering drought events (Liu et al., \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e). It shows the basic metabolism conditions of an ecosystem without drought events. To determine pre-drought GPP, we removed all values in tandem with an SPEI \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\le\\)\u003c/span\u003e\u003c/span\u003e \u0026minus;\u0026thinsp;1 (VCI \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\le\\)\u003c/span\u003e\u003c/span\u003e 30), and only used the mean of the remaining monthly values. For example, the January pre-drought GPP is represented by a single mean value for every January without a drought over the 14-year period. That pattern is repeated for all 12 months across the same time frame. Post-drought GPP refers to real GPP values including all drought events. In contrast with the pre-drought GPP, in which the mean monthly value is the same for each year, post-drought GPP indicates actual values (Liu et al., \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Schwalm et al., \u003cspan class=\"CitationRef\"\u003e2017\u003c/span\u003e). DRT is the time taken after a drought event has concluded in tandem with the time taken for the post-drought GPP to exceed the pre-drought GPP. It is the difference between the drought recovery ending time (post-drought GPP returning to pre-drought GPP) and the drought ending time (Liu et al., \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Schwalm et al., \u003cspan class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Section2\" id=\"Sec9\"\u003e\n \u003ch2\u003e3.4 Boruta algorithm\u003c/h2\u003e\n \u003cp\u003eWe used the Boruta algorithm to examine which parameters are most important when determining DRT. The Boruta method selects variables and ranks them in order of importance while rejecting parameters that do not improve - or adversely affect - the model\u0026rsquo;s accuracy. Boruta operates by initially adding randomness to a dataset by creating shuffled duplicates of all parameters. These are termed \u0026lsquo;shadow parameters\u0026rsquo; (Kursa and Rudnicki, \u003cspan class=\"CitationRef\"\u003e2010\u003c/span\u003e). The extended dataset is then trained with a random forest classifier using decision trees to select appropriate class. The appropriate class is reached by applying a measure that determines each parameter\u0026rsquo;s importance. A higher result translates to more important class. The algorithm performs iterations where it checks whether the real parameter has a higher importance than the best shadow feature at every stage. This is done by comparing \u0026ldquo;z scores\u0026rdquo;; the real parameter must have a higher z score than the maximum shadow parameter z score. In this process, parameters considered unimportant are eliminated. The algorithm terminates when all parameters are confirmed or rejected or when random forest runs are exhausted. The equation and further technical breakdown can be found in Prasad et al. (\u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Results And Discussion","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Seasonal patterns and trends for drought-related variables\u003c/h2\u003e \u003cp\u003eTo elucidate seasonal drought variation in Lake Victoria basin, we used six variables: rainfall (mm/month); PET (mm/month); GPP (g/m\u003csup\u003e2\u003c/sup\u003e/d); SM (m\u003csup\u003e3\u003c/sup\u003e/m\u003csup\u003e3\u003c/sup\u003e) for depths of 0\u0026ndash;10 cm, 10\u0026ndash;40 cm, 40\u0026ndash;100 cm, and 100\u0026ndash;200 cm; VCI; and SPEI (Fig.\u0026nbsp;2). We studied four seasons, DJF, MMA, JJA, and SON, and used seasonal mean values from 2003 to 2016 (Table\u0026nbsp;3).\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;2, we found a clear seasonal rainfall pattern, with large amounts of rainfall (151.79 mm/month) in MAM and the least (53.67 mm/month) in JJA. Although the different season PET did not vary greatly (175.00 to 180.11 mm/month), it increased significantly in the northeastern part of the basin during DJF. GPP was the same for DJF (6.69 g/m\u003csup\u003e2\u003c/sup\u003e/d) and MMA (6.79 g/m\u003csup\u003e2\u003c/sup\u003e/d) though it dropped significantly in JJA (5.46 g/m\u003csup\u003e2\u003c/sup\u003e/d) and SON (5.38 g/m\u003csup\u003e2\u003c/sup\u003e/d), particularly in the basin\u0026rsquo;s southern and eastern areas. 0 to 100 cm SM did not change significantly except for JJA, when the entire basin experienced very low SM levels (0.20 m\u003csup\u003e3\u003c/sup\u003e/m\u003csup\u003e3\u003c/sup\u003e) and MMA, when the soil moisture was quite high (0.27 m\u003csup\u003e3\u003c/sup\u003e/m\u003csup\u003e3\u003c/sup\u003e). From 100 to 200 cm, SM was high for all seasons (0.27 m\u003csup\u003e3\u003c/sup\u003e/m\u003csup\u003e3\u003c/sup\u003e) compared to the average SM across seasons from 0\u0026ndash;100 cm (0.23 m\u003csup\u003e3\u003c/sup\u003e/m\u003csup\u003e3\u003c/sup\u003e). SM from 100 to 200 cm remained similar for all seasons, including JJA, even though that season had the lowest soil moisture from 0\u0026ndash;100 cm. VCI was significantly high and did not vary greatly between DJF (63%) and MMA (69%). JJA and SON had very low VCI, with SON (44%) having a much lower VCI than JJA (46%). VCI was generally low in the southern and eastern parts of the basin during JJA and SON. SPEI was high during MMA (0.47) and SON (0.23). A significant SPEI drop was observed for DJF (-0.11), and it was low throughout the entire basin in JJA (-0.90).\u003c/p\u003e \u003cp\u003eGenerally, JJA showed low rainfall, GPP, SM, VCI, and SPEI, the driest season. SM from 0 to 100 cm increased when rainfall increased and decreased when there was less rainfall. In contrast, SM from 100cm to 200cm was significantly high for each season. This difference could be attributed to maize\u0026rsquo;s root depth. Its fibrous roots reach a depth of up to 100 cm and draws most of its soil moisture from that depth (Leenaars et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). VCI and SPEI showed similar results, with very low values in JJA. SPEI was quite high in SON, however, whereas VCI was quite low. This can be explained by VCI only considering a greenness index, whereas SPEI also accounts for climate factors (rainfall and PET). While precipitation and PET have cumulative effects on vegetation conditions, these is a time lag (Gebrehiwot et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; ZHANG et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;3 shows Lake Victoria Basin climate variable trends from 2003 to 2016. The blue and red colors represent decreasing and increasing trends, respectively. The dark points mark significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.1) decreased or increased areas. Rainfall decreased from 2003 to 2006 in the west basin along with a significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.1) decrease trend in the mid-west area. A rainfall increase was found in the east with additional pockets showing significant increases in the east and southeast. PET increased significantly in the eastern part of the basin while the far west and southwestern basin area have pockets indicating decreasing PET. GPP in the northern portions of the basin increased while it decreased in the southern area. Soil moisture from 0\u0026ndash;10 cm, 10\u0026ndash;40 cm and 40\u0026ndash;100 cm decreased, with a significant incremental decrease in the east. There is an overall significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.1) soil moisture increase, however, from 100\u0026ndash;200 cm throughout the basin. Similar to GPP, VCI\u0026rsquo;s trends are evenly distributed throughout the basin with scattered significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.1) increases in the western and northeastern areas. SPEI decreases over almost the basin\u0026rsquo;s entirety with a significant trend decrease (p\u0026thinsp;\u0026lt;\u0026thinsp;0.1) in the northwest and western portions. Overall, decreasing rainfall and increasing PET in the western basin led to decreasing SPEI. Although rainfall in the eastern basin increased from 2003 to 2006, the increasing PET and decreasing SM from each depth effects vegetation condition and results in a decreasing GPP and VCI. This is especially true in the basin\u0026rsquo;s southeast area.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Drought characterization\u003c/h2\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e4.2.1 Annual drought conditions\u003c/h2\u003e \u003cp\u003eTo assess drought conditions in the study area, we estimated SPEI and VCI annual spatial distribution (Fig.\u0026nbsp;4). Generally, mean SPEI and VCI values indicate that near-normal conditions imply a potentially stable canopy cover and greenness. As shown in Fig.\u0026nbsp;4a, SPEI mean annual spatial distribution showed near-normal conditions with very moderate drought tendencies. For example, in 2007, the entire basin was almost normal. But drought prevalence has progressed since then, with conditions peaking in 2016 when the majority of the basin experienced moderate to severe drought conditions.\u003c/p\u003e \u003cp\u003eVCI mean annual spatial distribution shows almost no drought, with light to moderate drought occurring in the southeastern part of the basin (Fig.\u0026nbsp;4b). 2004, 2005, 2006, and 2016 had the largest spatial coverage (almost the entirety of the basin) for drought. In other years, however, the western side was not affected. 2005, 2009, 2011, and 2016 have the lowest VCI values, indicating moderate to severe drought conditions on the eastern outskirts of the study area. 2016 was the driest year, with both the lowest VCI value and largest drought spatial coverage. The highest VCI value and largest non-drought spatial coverage was in 2007. This is consistent with the observed SPEI drought conditions, which indicates that the driest and wettest years were 2016 and 2007, respectively.\u003c/p\u003e \u003cp\u003eSPEI and VCI differed spatially in 2006 and 2015, presumably because of a potential temporal lag between VCI and SPEI. For example, 2006 SPEI has high values even though VCI in the same year is low. This is because VCI is an indicator of greenness, whereas SPEI considers PET and rainfall climate factors. That fundamental difference in approach implies a time lag in the results (Fig.\u0026nbsp;4); the low 2005 SPEI value led to the low 2006 VCI value. The same phenomenon can be found in 2015, when the high VCI value can be explained by the high SPEI value in 2014, not the SPEI value in 2015.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e4.2.2 Monthly drought conditions\u003c/h2\u003e \u003cp\u003eWe calculated monthly SPEI and VCI spatial distributions to assess detailed monthly drought variation with a scale denoting drought severity from \u0026ldquo;extreme wet\u0026rdquo; to \u0026ldquo;extreme drought\u0026rdquo; (Fig.\u0026nbsp;5). The results shown in Fig.\u0026nbsp;2, Fig.\u0026nbsp;4, and Fig.\u0026nbsp;5 are consistent with the evidence from several reports referenced in Table\u0026nbsp;4. From 2003 to 2016, July has consistently experienced the most severe to extreme drought conditions. Generally, JJA has the most significant drought conditions, which are consistent with the results shown in Fig.\u0026nbsp;2, showing the most drought severity occurring in JJA. DJF also show significant drought conditions, especially in the northern parts of the basin. MAM and SON show near normal conditions with occasional oddly scattered spatial instances of drought. 2016 and 2009 had the highest drought severity conditions, which is consistent with average SPEI 2016 values and average 2009 VCI values, respectively. This result could be due to the different times in which the basin receives rainfall, as a rainfall deficit tends to correlate with drought conditions and vice versa.\u003c/p\u003e \u003cp\u003eTo evaluate drought condition results, we analyzed recorded drought years for different countries in the study area, as shown in Table\u0026nbsp;4. Burundi, located in the southwestern part of the area, has almost 4\u0026nbsp;million people in the basin, and 95% of the country\u0026rsquo;s population is rural. Thus, drought impacts are quite severe (Gebremeskel et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Yao et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). According to drought event records (Table\u0026nbsp;4), the drought progressed from 2003 to 2005 and again from 2008 to 2010, mostly affecting the northeastern parts of Burundi (East African Community (EAC), \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). This finding matches the data shown in both Fig.\u0026nbsp;4 and Fig.\u0026nbsp;5, demonstrating moderate to extreme drought conditions in the southwestern part of the study area. For Western Kenya, located in the eastern and northeastern parts of the study area, drought events occurred in 2004 and affected 3.3\u0026nbsp;million Kenyans. Furthermore, as shown in both Fig.\u0026nbsp;4 and Fig.\u0026nbsp;5, Western Kenya suffered repeated drought events in 2005, 2008, 2010, 2012, 2014, and 2016 (Agutu et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Ayugi et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; East African Community (EAC), \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Gebremeskel et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Nyaoro et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Opiyo et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Schmidt et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). As a result, three major provinces, Nyanza, Western, and Rift Valley, bore the brunt of the drought impact (Awange et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). The 2003 drought event in Rwanda in the western basin was characterized by below-average rainfall. It disproportionately affected rural areas in Rwanda\u0026rsquo;s semi-arid east, where people are much more dependent on rainfall for both crops and animal husbandry (East African Community (EAC), \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). This is consistent with the data in Fig.\u0026nbsp;4 and Fig.\u0026nbsp;5, which show moderate to extreme drought events in the western basin. The drought record in Tanzania, the southern part of the basin, shows drought first occurred in 2003, affecting the regions of Kagera, Mwanza, and Mara. In addition, as shown in Fig.\u0026nbsp;4 and Fig.\u0026nbsp;5, drought conditions in this region recurred in 2004, 2006, and 2011 but not in 2007. The failure of short-season rains caused severe drought in late 2005 and early 2006 (Bhaga et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In early 2007, this region experienced heavy rains, (Fig.\u0026nbsp;5), with moderate wet conditions in January (Bhaga et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Uganda lies in the northern portion of the study area, with some parts in the western, central, and eastern areas. More than 600,000 people were affected by drought as of 2005 in the Masaka and Mpigi districts (Hakuza and Waita, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Droughts recurred in 2008 and 2010 and affected similar areas. For this reason, farming communities in central, eastern and southwestern Uganda suffered economic and financial damage (Hakuza and Waita, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). The drought variation\u0026rsquo;s spatial extent is shown in Fig.\u0026nbsp;5. It is notable that Fig.\u0026nbsp;5 shows a one-month delay in \u0026lsquo;dry\u0026rsquo; conditions predominant from June to August using the SPEI index and from July to September, even extending into October, with the VCI index. This is because SPEI relies on meteorological data whereas VCI relies on vegetation conditions, implying drought is first observed meteorologically before it\u0026rsquo;s observed in vegetation.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Drought duration time (DDT) and drought recovery time\u003c/h2\u003e \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e \u003ch2\u003e4.3.1 Drought event identification based on drought severity\u003c/h2\u003e \u003cp\u003eTo understand drought event history based on SPEI and VCI results, we calculated the percentage of different drought events based on severity (Table\u0026nbsp;3). Spatial distribution is shown in Fig.\u0026nbsp;6. The most moderate SPEI drought events (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(-1.5cript\u003e\\)\u003c/span\u003e\u003c/span\u003e) were in the southern part of the study area (Fig.\u0026nbsp;6). Most of the study area has up to 70% possibility of experiencing moderate drought and a 10 to 20% possibility of experiencing a severe drought (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(-2cript\u003e\\)\u003c/span\u003e\u003c/span\u003e). Extreme drought events (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\text{S}\\text{P}\\text{E}\\text{I}\\le -2)\\)\u003c/span\u003e\u003c/span\u003e mostly occurred in the eastern and northeastern tip of the study area (Fig.\u0026nbsp;6). Overall trends indicate that the basin has a higher spatial area percentage prone to moderate drought. Using VCI data, 50 to 80% of the study area may experience a moderate drought (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(20cript\u003e\\)\u003c/span\u003e\u003c/span\u003e) except for regions in the southeast and southern tip, where the percentage is only 10 to 30%. VCI-predicted severe drought (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(10cript\u003e\\)\u003c/span\u003e\u003c/span\u003eis similar to SPEI. The majority of the study area has a 10 to 20% chance of experiencing extreme drought (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(0cript\u003e\\)\u003c/span\u003e\u003c/span\u003e) except for the southeast region, which shows a 40 to 70% likelihood (Fig.\u0026nbsp;6). Both VCI- and SPEI-based percentage coverage trends for moderate (M), severe (S), and extreme (E) drought are similar where the spatial drought coverage percentage decreases from moderate to severe. There are some differences, however, in the southeast region for SPEI_M, VCI_M, and SPEI_E, VCI_E which show an inverse trend. This is because VCI indicates drought based on actual vegetation conditions whereas SPEI indicates drought based on meteorological conditions.\u003c/p\u003e \u003cp\u003eOverall, more moderate and severe drought events occurred in the southern part of the study area; the eastern and western parts of the basin experienced extreme drought events. As previously discussed, the western area is mostly farming communities in Uganda, Tanzania, Rwanda, and Burundi, including large food crop (maize) belts in central and eastern Uganda. The areas prone to moderate, severe and extreme drought events (Fig.\u0026nbsp;6) should be considered high-risk drought areas and likely future bottlenecks for water balance. Drought mitigation strategies are imperative in these areas.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section3\"\u003e \u003ch2\u003e4.3.2 Drought duration and recovery time based on SPEI and VCI\u003c/h2\u003e \u003cp\u003eUsing SPEI and VCI drought indices, we calculated and compared mean DDT and DRT spatial distributions in the Lake Victoria Basin from 2003 to 2016 (Fig.\u0026nbsp;7). The mean SPEI-DDT ranges from 2 to 4 months with a 2.45-day average value. The northeastern and southwestern study area were most likely to suffer from meteorological drought (i.e., SPEI) for at least 3 months, compared to 2 months in other areas. The VCI-DDT spatial distribution (2.97 months) was similar to SPEI-DDT (2.45 months) except for the southeast study area, where VCI-DDT is 2 months longer than SPEI-DDT. This means the southeastern part of the basin is more likely to suffer a long agricultural drought. The SPEI-DRT mean value is 2.02 months, though there are a few instances in the northwest basin area where SPEI-DRT took more than 8 months (black). The VCI-DRT mean value is 1.63 months, which is 0.39 months shorter than the SPEI-DRT mean value.\u003c/p\u003e \u003cp\u003eThe study area\u0026rsquo;s mean DDT is higher than the mean DRT, though the mean DDT and mean DRT are almost the same in the southern area, particularly in the southwest. The mean DRT significantly exceeded the mean DDT in the western and northwestern parts of the basin (in black), which show a mean DRT of 8 months compared to three months for DDT. Our results are consistent with Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), who showed that most drought events in the study area required less than 6 months to recover, though parts of the western basin needed 7\u0026ndash;12 months. This long mean DRT puts the western and northwestern basin regions at risk of excessively long drought periods. These include the semi-arid eastern parts of Rwanda, northeastern regions of Burundi, communities in the Kagera region of Tanzania, and central, and southwestern parts of Uganda. It is imperative to pay attention to areas with long DRTs because the northeastern (i.e., Nyanza, Western, and Rift Valley provinces) and southwestern (i.e., Uganda, Tanzania, Rwanda, and Burundi) study areas are mostly cropland (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These regions take 4 to 5 months to recover from drought events. Several sub-basins in the western and northern study area create microclimates that potentially affect land\u0026ndash;atmosphere interactions. For example, the catchment\u0026rsquo;s western portion is in the purview of the Kagera sub-basin, which encompasses Uganda, Tanzania, Kenya, Rwanda, and Burundi. Countries such as Uganda, Rwanda, and Burundi lie entirely in the Nile basin, which includes the Victoria basin. The various microclimates created by the sub-basins influence drought characteristics. The Great Rift Valley in the western part of the basin has several lakes and rivers, such as Lake Albert, Lake Edward, and Lake Kivu. These are connected by river systems west of the Victoria basin and affect the climate. The Great Rift Valley also passes close to the eastern side of the basin, but has no significant water bodies.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Factors affecting DRT\u003c/h2\u003e \u003cp\u003eWe ranked the importance of DDT, SM depths, PET, GPP, rainfall, and drought indices on DRT (Fig.\u0026nbsp;8). We selected and ranked these factors using a \u0026ldquo;z score\u0026rdquo; for SPEI-DRT and VCI-DRT, where a higher value means the factor is more important. The shadow Min, shadow Mean and shadow Max (blue box plots) add randomness to the data to allow for more precise parameter ranking. The green box plots show that each features\u0026rsquo; z score is higher than the shadow value, which means they are all significantly important in determining SPEI-DRT and VCI-DRT.\u003c/p\u003e \u003cp\u003eDDT is the most important parameter for determining DRT for SPEI-DRT followed by GPP, PET, SPEI, and rainfall. Soil moisture, though still relevant, is the least important parameter for DRT determination. DDT is also the most crucial parameter for determining DRT for VCI-DRT followed by VCI and rainfall. Soil moisture is again classified as the least important parameter, though still important. Based on these results, DDT is the most important factor for determining DRT, regardless of drought indices. We examined DRT spatial distribution after 1, 2-, 3-, 4-, and 5-month drought events to understand the DDT and DRT relationship (Fig.\u0026nbsp;9). After a one-month drought, SPEI-DRT (SPEI_DRT_1) shows that the entire basin needs to recover from the drought event, with DRTs ranging from 1 to 4 months with some scattered pockets requiring up to 6 months. For a 2-month drought, SPEI_DRT_2, less areas need to recover from drought than for the SPEI_DRT_1 scenario, though there are some isolated pockets in the far east study area that require more than 8 months. SPEI_DRT_3 indicates similar DRTs to SPEI_DRT_2, but the area requiring longer time was larger than for SPEI_DRT_1 or SPEI_DRT_2. As larger basin percentages do not experience 4- or 5-month drought events, SPEI_DRT_4 shows only the southwest and northeast portions of the study area requiring drought recovery, and SPEI_DRT_5 indicates that almost all of the study area does not require any DRT other than sparce pockets in the south. VCI-DRT results are similar to SPEI-DRT except for the DRTs after 4- and 5-month droughts. Results show a significant number of pockets in the northwest part of the basin that need to recover from drought. These were not present in SPEI_DRT_4. VCI_DRT_5 shows the southeastern parts of the study area require time to recover from drought whereas none is needed for SPEI_DRT_5.\u003c/p\u003e \u003cp\u003eIn the western portion of the study area, SPEI_DRT_1 to SPEI_DRT_4 trends indicate that as the number of drought months increase, drought recovery time also increases. Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) found similar results. Long droughts affect the water balance, hydrometeorological cycles, and natural land\u0026ndash;atmosphere relationships. Thus, it makes sense that long droughts have long recovery times. High PET correlates with a low DRT. This can be attributed to Lake Victoria\u0026rsquo;s natural hydrological cycle, where a high PET yields high precipitation levels and wards off drought; low PET will allow little to no precipitation to reach the basin (Avanzi et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). As noted in Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), low SPEI values correlate with a high DRT. Low SPEI values indicate severe drought conditions from which it often takes a long time to recover. Low GPP implies long DRTs because it indicates an unhealthy plant eco-system requiring a long recovery time. Rainfall has a negative relationship with DRT because rainfall encourages plant growth; thus, more rainfall results in a shorter DRT. He et al. (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) also found that more rainfall shortened the ecosystem\u0026rsquo;s DRT. Overall, increasing DDT is the most important factor influencing DRT. Tropical environments take a long time to recover after droughts in dry conditions. Our results are consistent with Wright et al. (\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2002\u003c/span\u003e), who found that drought recovery is more likely to occur in wet conditions. Furthermore, wet conditions with high PET and high levels of rainfall shortened DRT, whereas dry conditions, with a large SPEI values and long DDTs, led to low GPP and lengthened DRT. This also agrees with Schwalm et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eAccess to accurate drought duration and recovery time information is vitally important in drought-prone areas used for agricultural purposes. Using grid-based data with high resolution, we investigated seasonal patterns in drought-related variables, identified drought events, and examined both SPEI- and VCI-based DRTs in the Lake Victoria Basin from 2003 to 2016. This study is the first to determine meteorological and agricultural DRTs for the Lake Victoria Basin on a 0.05\u0026deg; monthly scale. Of the four seasons, JJA was the driest, with lower values of rainfall, GPP, SM, VCI, and SPEI. Decreasing GPP and VCI is caused by increasing PET and decreasing SM from each depth in the basin\u0026rsquo;s eastern area. Decreasing SPEI is due to decreasing rainfall and increasing PET in the basin\u0026rsquo;s western area. Drought indices SPEI and VCI showed that 2016 and 2007 were Lake Victoria Basin\u0026rsquo;s driest and wettest years in the study range. Meteorological drought calculations showed that moderate droughts occurred at higher frequency in the southeastern part of the basin, whereas the northeastern and mid-western areas were more likely to suffer from extreme drought events. Agricultural drought measurements showed that extreme drought events occurred at higher frequency in the basin\u0026rsquo;s southern areas. On average, SPEI-based DRT (2.02 months) was longer than VCI-based DRT (1.63 months). SPEI-DRT and VCI-DRT showed similar spatial distribution though SPEI-based DRT (2.02 months) was longer than VCI-based DRT (1.63 months) on average. DDT is the most important parameter for determining DRT, though regions with higher PET, SPEI, GPP, and precipitation values are also associated with shorter recovery times. These results improve understanding of drought on an ecosystem level. Nevertheless, a global DRT product with high accuracy and good spatial and temporal resolution remains challenging, and requires additional investigation.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgments\u003c/h2\u003e \u003cp\u003eWe acknowledge the satellite and reanalysis data freely provided by the Land Cover (LC) project of the European Space Agency (ESA) Climate Change Initiative (CCI), the NASA Goddard Earth Sciences Data and Information Services Center (GES DISC), the Oak Ridge National Laboratory Distributed Active Archive Center (ORNL DAAC), and the Land Processes Distributed Active Archive Center (LP DAAC) from NASA Earth Observing System Data and Information System (EOSDIS). This research was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (NRF-2019R1A2B5B01070196).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAadhar S, Mishra V (2017) Data Descriptor: High-resolution near real-time drought monitoring in South Asia. Sci Data 4:1\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/sdata.2017.145\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbramowitz M, Stegun IA (1965) Handbook of Mathematical Functions, with Formulas, Graphs, and Mathematical Tables. Dover Publications\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAghaKouchak A (2015) A multivariate approach for persistence-based drought prediction: Application to the 2010\u0026ndash;2011 East Africa drought. J Hydrol 526:127\u0026ndash;135. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jhydrol.2014.09.063\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgutu NO, Awange JL, Zerihun A, Ndehedehe CE, Kuhn M, Fukuda Y (2017) Assessing multi-satellite remote sensing, reanalysis, and land surface models\u0026rsquo; products in characterizing agricultural drought in East Africa. Remote Sens Environ 194:287\u0026ndash;302. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2017.03.041\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmadi B, Ahmadalipour A, Moradkhani H (2019) Hydrological drought persistence and recovery over the CONUS: A multi-stage framework considering water quantity and quality. Water Res 150:97\u0026ndash;110. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.watres.2018.11.052\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmadi B, Moradkhani H (2019) Revisiting hydrological drought propagation and recovery considering water quantity and quality. Hydrol Process 33:1492\u0026ndash;1505. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/hyp.13417\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderegg WRL, Konings AG, Trugman AT, Yu K, Bowling DR, Gabbitas R, Karp DS, Pacala S, Sperry JS, Sulman BN, Zenes N (2018) Hydraulic diversity of forests regulates ecosystem resilience during drought. Nature 561:538\u0026ndash;541. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41586-018-0539-7\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAvanzi F, Rungee J, Maurer T, Bales R, Ma Q, Glaser S, Conklin M (2019) Evapotranspiration feedbacks shift annual precipitation-runoff relationships during multi-year droughts in a Mediterranean mixed rain-snow climate. Hydrol Earth Syst Sci Discuss 1\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/hess-2019-377\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAwange JL, Anyah R, Agola N, Forootan E, Omondi P (2013) Potential impacts of climate and environmental change on the stored water of Lake Victoria Basin and economic implications. Water Resour Res 49:8160\u0026ndash;8173. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1002/2013WR014350\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAwange JL, O.O (2006) Lake Victoria: Ecology, Resources, Environment. Springer-Verlag, Heidelberg. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-1-4020-9726-3_12\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAyugi B, Tan G, Rouyun N, Zeyao D, Ojara M, Mumo L, Babaousmail H, Ongoma V (2020) Evaluation of meteorological drought and flood scenarios over Kenya, East Africa. Atmosphere (Basel). 11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/atmos11030307\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanerjee O, Bark R, Connor J, Crossman ND (2013) An ecosystem services approach to estimating economic losses associated with drought. Ecol Econ 91:19\u0026ndash;27. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ecolecon.2013.03.022\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhaga TD, Dube T, Shekede MD, Shoko C (2020) Impacts of climate variability and drought on surface water resources in sub-saharan africa using remote sensing: A review. Remote Sens 12:1\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs12244184\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEast African Community (EAC) (2010) Lake Victoria Water Supply and Sanitation Program Phase II_Appraisal\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFord TW, Labosier CF (2017) Meteorological conditions associated with the onset of flash drought in the Eastern United States. Agric For Meteorol 247:414\u0026ndash;423. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.agrformet.2017.08.031\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFunk C, Hoell A, Shukla S, Blad\u0026eacute; I, Liebmann B, Roberts JB, Robertson FR, Husak G (2014) Predicting East African spring droughts using Pacific and Indian Ocean sea surface temperature indices. Hydrol Earth Syst Sci 18:4965\u0026ndash;4978. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/hess-18-4965-2014\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGebrehiwot T, van der Veen A, Maathuis B (2011) Spatial and temporal assessment of drought in the Northern highlands of Ethiopia. Int J Appl Earth Obs Geoinf 13:309\u0026ndash;321. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.jag.2010.12.002\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGebremeskel G, Tang Q, Sun S, Huang Z, Zhang X, Liu X (2019) Droughts in East Africa: Causes, impacts and resilience. Earth-Science Rev 193:146\u0026ndash;161. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.earscirev.2019.04.015\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGunston H, Batchelor CH (1983) A comparison of the Priestley-Taylor and Penman methods for estimating reference crop evapotranspiration in tropical countries. Agric Water Manag 6:65\u0026ndash;77. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0378-3774(83)90026-4\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHakuza A, Waita J (2008) REVIEW AND ANALYSIS OF EXISTING DROUGHT RISK REDUCTION POLICIES AND PROGRAMMES IN UGANDA\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao Z, AghaKouchak A, Nakhjiri N, Farahmand A (2014) Global integrated drought monitoring and prediction system. Sci data 1:140001. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/sdata.2014.1\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao Z, Singh VP, Xia Y (2018) Seasonal Drought Prediction: Advances, Challenges, and Future Prospects. Rev Geophys 56:108\u0026ndash;141. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/2016RG000549\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe B, Liu J, Guo L, Wu X, Xie X, Zhang Y, Chen C, Zhong Z, Chen Z (2018) Recovery of Ecosystem Carbon and Energy Fluxes From the 2003 Drought in Europe and the 2012 Drought in the United States. Geophys Res Lett 45:4879\u0026ndash;4888. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2018GL077518\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang L, Zhou P, Cheng L, Liu Z (2021) Dynamic drought recovery patterns over the Yangtze River Basin. Catena 201. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.catena.2021.105194\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuffman GJ, Adler RF, Bolvin DT, Gu G, Nelkin EJ, Bowman KP, Hong Y, Stocker EF, Wolff DB (2007) Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J Hydrometeorol 8:38\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JHM560.1\u003c/span\u003e\u003c/span\u003e The TRMM Multisatellite Precipitation Analysis (TMPA)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKizza M, Rodhe A, Xu CY, Ntale HK, Halldin S (2009) Temporal rainfall variability in the Lake Victoria Basin in East Africa during the twentieth century. Theor Appl Climatol 98:119\u0026ndash;135. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00704-008-0093-6\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKogan FN (1995) Application of vegetation index and brightness temperature for drought detection. Adv Sp Res 15:91\u0026ndash;100. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/0273-1177(95)00079-T\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKOGAN FN (1990) Remote sensing of weather impacts on vegetation in non-homogeneous areas. Int J Remote Sens 11:1405\u0026ndash;1419. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431169008955102\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKursa MB, Rudnicki WR (2010) Feature selection with the boruta package. J Stat Softw 36:1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.18637/jss.v036.i11\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeenaars JGB, Claessens L, Heuvelink GBM, Hengl T, Ruiperez Gonz\u0026aacute;lez M, van Bussel LGJ, Guilpart N, Yang H, Cassman KG (2018) Mapping rootable depth and root zone plant-available water holding capacity of the soil of sub-Saharan Africa. Geoderma 324:18\u0026ndash;36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.geoderma.2018.02.046\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu L, Gudmundsson L, Hauser M, Qin D, Li S, Seneviratne SI (2019) Revisiting assessments of ecosystem drought recovery. Environ Res Lett 14:114028. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1088/1748-9326/ab4c61\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyon B, Dewitt DG (2012) A recent and abrupt decline in the East African long rains. Geophys Res Lett 39. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2011GL050337\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMadani N, Parazoo NC (2020) Global Monthly GPP from an Improved Light Use Efficiency Model, 1982\u0026ndash;2016. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3334/ORNLDAAC/1789\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMailu A (2001) Preliminary assessment of the social, economic and environmental impacts of water hyacinth in Lake Victoria Basin and status of control\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcKee TB, Doesken NJ, Kleist J (1993) The relationship of drought frequency and duration to time scales, in: Proceedings of the 8th Conference on Applied Climatology. Boston, pp.\u0026nbsp;179\u0026ndash;183\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMishra AK, Singh VP (2010) A review of drought concepts. J Hydrol 391:202\u0026ndash;216. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jhydrol.2010.07.012\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMu Q, Zhao M, Kimball JS, McDowell NG, Running SW (2013) A Remotely Sensed Global Terrestrial Drought Severity Index. Bull Am Meteorol Soc 94:83\u0026ndash;98. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/BAMS-D-11-00213.1\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNyaoro D, Schade J, Schmidt K (2016) Assessing the Evidence: Migration, Environment and Climate Change in Kenya\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpiyo F, Wasonga O, Nyangito M, Schilling J, Munang R (2015) Drought Adaptation and Coping Strategies Among the Turkana Pastoralists of Northern Kenya. Int J Disaster Risk Sci 6:295\u0026ndash;309. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13753-015-0063-4\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeel MC, Finlayson BL, McMahon TA (2007) Updated world map of the K\u0026ouml;ppen-Geiger climate classification. Hydrol Earth Syst Sci 11:1633\u0026ndash;1644. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/hess-11-1633-2007\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrasad R, Deo RC, Li Y, Maraseni T (2019) Weekly soil moisture forecasting with multivariate sequential, ensemble empirical mode decomposition and Boruta-random forest hybridizer algorithm approach. Catena 177:149\u0026ndash;166. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.catena.2019.02.012\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePriestley, C.H.B., Taylor, R.J., 1972. On the Assessment of Surface Heat Flux and Evaporation Using Large-Scale Parameters. Mon. Weather Rev. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/1520-0493(1972)100\u0026lt;0081:OTAOSH\u0026gt;2.3.CO;2\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodell M, Houser PR, Jambor U, Gottschalck J, Mitchell K, Meng CJ, Arsenault K, Cosgrove B, Radakovich J, Bosilovich M, Entin JK, Walker JP, Lohmann D, Toll D (2004) The Global Land Data Assimilation System. Bull Am Meteorol Soc 85, 381\u0026ndash;394. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/BAMS-85-3-381\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchmidt W, Peter Uhe A, Kimutai J, Otto F, Cullen H (2017) The Drought in Kenya, 2016\u0026ndash;2017, The Drought in Kenya, 2016\u0026ndash;2017. Climate and Development Knowledge Network and World Weather Attribution Initiative\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwalm CR, Anderegg WRL, Michalak AM, Fisher JB, Biondi F, Koch G, Litvak M, Ogle K, Shaw JD, Wolf A, Huntzinger DN, Schaefer K, Cook R, Wei Y, Fang Y, Hayes D, Huang M, Jain A, Tian H (2017) Global patterns of drought recovery. Nature 548:202\u0026ndash;205. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nature23021\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeneviratne SI, Ciais P (2017) Trends in ecosystem recovery from drought. Nature 548:164\u0026ndash;165. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/548164a\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh VP, Guo H, Yu FX (1993) Parameter estimation for 3-parameter log-logistic distribution (LLD3) by Pome. Stoch Hydrol Hydraul 7:163\u0026ndash;177. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/BF01585596\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSternberg T (2011) Regional drought has a global impact. Nature 472:169\u0026ndash;169. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/472169d\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrenberth KE, Dai A, Van Der Schrier G, Jones PD, Barichivich J, Briffa KR, Sheffield J (2014) Global warming and changes in drought. Nat Clim Chang 4:17\u0026ndash;22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nclimate2067\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan der Molen MK, Dolman AJ, Ciais P, Eglin T, Gobron N, Law BE, Meir P, Peters W, Phillips OL, Reichstein M, Chen T, Dekker SC, Doubkov\u0026aacute; M, Friedl MA, Jung M, van den Hurk BJJM, de Jeu RAM, Kruijt B, Ohta T, Rebel KT, Plummer S, Seneviratne SI, Sitch S, Teuling AJ, van der Werf GR, Wang G (2011) Drought and ecosystem carbon cycling. Agric For Meteorol 151:765\u0026ndash;773. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.agrformet.2011.01.018\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Dijk AIJM, Beck HE, Crosbie RS, De Jeu RAM, Liu YY, Podger GM, Timbal B, Viney NR (2013) The Millennium Drought in southeast Australia (2001\u0026ndash;2009): Natural and human causes and implications for water resources, ecosystems, economy, and society. Water Resour Res 49:1040\u0026ndash;1057. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/wrcr.20123\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVicente-Serrano SM, Beguer\u0026iacute;a S, L\u0026oacute;pez-Moreno JI (2010) A Multiscalar Drought Index Sensitive to Global Warming: The Standardized Precipitation Evapotranspiration Index. J Clim 23:1696\u0026ndash;1718. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/2009JCLI2909.1\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWest H, Quinn N, Horswell M (2019) Remote sensing for drought monitoring \u0026amp; impact assessment: Progress, past challenges and future opportunities. Remote Sens Environ 232:111291. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2019.111291\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWright JF, Gunn RJM, Winder JM, Wiggers R, Vowles K, Clarke RT, Harris I (2002) A comparison of the macrophyte cover and macroinvertebrate fauna at three sites on the River Kennet in the mid 1970s and late 1990s. Sci Total Environ 282\u0026ndash;283, 121\u0026ndash;142. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/S0048-9697(01)00948-2\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao Y, Liang S, Li X, Hong Y, Fisher JB, Zhang N, Chen J, Cheng J, Zhao S, Zhang X, Jiang B, Sun L, Jia K, Wang K, Chen Y, Mu Q, Feng F (2014) Bayesian multimodel estimation of global terrestrial latent heat flux from eddy covariance, meteorological, and satellite observations. J Geophys Res Atmos 119:4521\u0026ndash;4545. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/2013JD020864\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu Z, Wang J, Liu S, Rentch JS, Sun P, Lu C (2017) Global gross primary productivity and water use efficiency changes under drought stress. Environ Res Lett 12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1088/1748-9326/aa5258\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang D, Liu X, Bai P (2019) Assessment of hydrological drought and its recovery time for eight tributaries of the Yangtze River (China) based on downscaled GRACE data. J Hydrol 568:592\u0026ndash;603. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jhydrol.2018.11.030\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZHANG F, ZHANG L, WANG X, HUNG J (2013) Detecting Agro-Droughts in Southwest of China Using MODIS Satellite Data. J Integr Agric 12:159\u0026ndash;168. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/S2095-3119(13)60216-6\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou M, Brandt P, Pelster D, Rufino MC, Robinson T, Butterbach-Bahl K (2014) Regional nitrogen budget of the Lake Victoria Basin, East Africa: Syntheses, uncertainties and perspectives. Environ Res Lett 9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1088/1748-9326/9/10/105009\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eDue to technical limitations, table 1 to 4 is only available as a download in the Supplemental Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"stochastic-environmental-research-and-risk-assessment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"serr","sideBox":"Learn more about [Stochastic Environmental Research and Risk Assessment](https://www.springer.com/journal/477)","snPcode":"477","submissionUrl":"https://submission.nature.com/new-submission/477/3","title":"Stochastic Environmental Research and Risk Assessment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Drought, Drought recovery time, SPEI, VCI, GPP, Lake Victoria Basin","lastPublishedDoi":"10.21203/rs.3.rs-774555/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-774555/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDrought imposes severe, long-term effects on global environments and ecosystems. A better understanding of how long it takes a region to recover to pre-drought conditions after drought is essential for addressing future ecology risks. In this study, drought-related variables were obtained using remote sensing and reanalysis products for 2003 to 2016. The meteorological drought index (standardized precipitation evapotranspiration index [SPEI]) and agricultural drought index (vegetation condition index [VCI]) were employed to estimate drought duration time (DDT) and drought recovery time (DRT). To the basin\u0026rsquo;s west, decreasing rainfall and increasing potential evapotranspiration led to decreasing SPEI. On the east side, decreasing soil moisture from each depth effects vegetation condition, which results in a decreasing gross primary productivity and VCI. Extreme meteorological drought events are likely to occur in the basin\u0026rsquo;s northeastern and middle western areas, while the southern basin is more likely to suffer from extreme agricultural drought events. The mean SPEI-based DDT (2.45 months) was smaller than the VCI-based DDT (2.97 months); the average SPEI-based DRT (2.02 months) was larger than the VCI-based DRT (1.63 months). Most of the area needs 1 or 2 months to recover from drought except for the basin\u0026rsquo;s northwestern area, where the DRT is more than 8 months. DDT is the most important parameter in determining DRT. These results provide useful information about regional drought recovery that will help local governments looking to mitigate potential environmental risks and formulate appropriate agricultural policies in Lake Victoria Basin.\u003c/p\u003e","manuscriptTitle":"Detecting Agricultural and Meteorological Drought With Gross Primary Production Recovery Including Spatiotemporal Statistical Analysis in East Africa's Lake Victoria Basin","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2021-08-05 16:37:25","doi":"10.21203/rs.3.rs-774555/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2021-08-05T07:36:47+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2021-08-02T12:57:49+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"Stochastic Environmental Research and Risk Assessment","date":"2021-08-02T12:50:19+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2021-08-02T07:02:42+00:00","index":"","fulltext":""},{"type":"submitted","content":"Stochastic Environmental Research and Risk Assessment","date":"2021-08-01T23:44:23+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"stochastic-environmental-research-and-risk-assessment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"serr","sideBox":"Learn more about [Stochastic Environmental Research and Risk Assessment](https://www.springer.com/journal/477)","snPcode":"477","submissionUrl":"https://submission.nature.com/new-submission/477/3","title":"Stochastic Environmental Research and Risk Assessment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"cfbd336b-d44f-40d2-b2f9-2dfd9e6c7df8","owner":[],"postedDate":"August 5th, 2021","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":6242802,"name":"Environmental Policy"}],"tags":[],"updatedAt":"2021-11-09T22:14:06+00:00","versionOfRecord":[],"versionCreatedAt":"2021-08-05 16:37:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-774555","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-774555","identity":"rs-774555","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}