Observed trends in timing and severity of streamflow droughts across global tropics

Drought is a recurrent climatic hazard impacting natural and built environmental systems, including human lives. Although several studies have assessed streamflow droughts and their multivariate characterization, very few studies have focused on understanding spatiotemporal changes in drought attributes, such as drought seasonality, severity and duration across global tropics. Further, the nonlinear response between onset time and severity of streamflow droughts at a large scale are unknown. Leveraging ground-based streamflow observations, this study for the first time investigate changes in streamflow drought characteristics across global tropics using two 30 year climate normal periods: 1961–1990 and 1991–2020. Our analyses of changes in probability distributions of onset time and severity (deficit volume) of streamflow droughts over the two time windows show significant shifts towards higher values for Northeast and South American Monsoon region, Western Africa, eastern South Africa, north and eastern Australia. Around 55% of the sites show an increase in drought frequency in recent times. We found that in the recent times, only 27% of sites depict an increase in deficit volume accompanied by delayed onset. Further, we identify a few regional hotspots, such as Northeast and South American monsoon region, and eastern coast of Australia show an increased frequency of droughts with an upward trend in deficit volume in recent years. As expected, the individual changes in drought attributes have translated into changes in joint occurrences of their interdependent attributes, assuming the correlation between onset time and deficit volume. Our analyses show robust dependence strengths between onset time and deficit volume, which strengthen further in the recent time window over 50% of catchments. The nonstationary changes identified here in individual drought attributes and their joint dependence can alter the hazard potential of extreme droughts, which has consequences in risk management, climate adaptation and water resources planning.


Introduction
The recent past decades have witnessed a notable shift towards a warming climate, contributing to increased droughts (Dai 2013, Chiang et al 2021).Drought is one of the most common and devastating natural hazards (Misnawati et al 2021, UNCCD 2022) that predominantly affects ecological, agricultural, and economic sectors (Wilhite et al 2007, Sternberg 2011, Vicente-Serrano et al 2013, Aguirre-Gutiérrez et al 2020).Droughts in 2022 have impacted 50% more of the global population than the exposed population during 2002-2021(CRED 2023)).While deficit in precipitation leads to a meteorological drought, a prolonged meteorological drought episode further escalates the depletion of available water in streams and reservoirs, leading to hydrological droughts.The streamflow drought is identified as the occurrence of low flows in the rivers for an extended period of time.The global tropics, which encompass the majority of the global population, are indisputably more susceptible to the impacts of climate change (Mishra et al 2023).
The earlier studies (Gudmundsson et al 2019, Teutschbein et al 2022) have assessed streamflow trends over the globe and regional scales, explicitly emphasizing mean changes in different streamflow indices, such as mean and extreme flows statistics.Brunner et al (2023) have detected changes in hydrological drought generating signatures over the Central Alps in recent periods  compared to the past decades  and found shifts in drought seasonality and severity in high-elevation catchments.Although previous regional-scale assessments have identified increasing trends in deficit volume and the spatial coverage of droughts across South Asia Further, it is essential to understand nonlinear response between multivariate streamflow drought attributes, such as time to onset (i.e., seasonality) and severity (typically described by deficit volume) and their joint occurrences over time for environmental planning and monitoring (Blöschl et al 2017, Funk and Shukla 2020, Slater et al 2021).This is especially required for areas where seasonality predominates stream flows since delay in surface irrigation may impact cropwater scheduling operation leading to loss of agricultural production (Qutbudin et al 2019) and hydropower generation (Qiu et al 2023).Moreover, a large body of assessments (Fleig et al 2006, Sheffield et al 2009, Rajsekhar et al 2015, Ganguli et al 2022b) have performed both univariate and multivariate frequency analyses of streamflow droughts, considering interdependent attributes such as severity (or intensity) versus duration and severity versus spatial extents.While there is little research on the analysis of onset patterns and its temporal evolution, very few regional scale studies have assessed shift in onset and deficit volume and their joint trends.However, the derived insights from regional scale assessments may not be generalized to vast areas of global tropics with varied climate regimes.
Our research underpins the need for comprehensive knowledge of shifts in streamflow drought characteristics and their joint occurrences, such as earlier (delayed) onset accompanied by increased (lower) deficit volume, especially in the tropics, where seasonality dominates the streamflow time series (Henry 2005).Furthermore, over the tropics, the recovery time of drought is claimed to be the longest (Schwalm et al 2017) compared to droughts in other climate types, underscoring the importance of assessing streamflow drought evolutions and the nonlinear responses between interdependent characteristics across medium-to-large-sized river basins in this region.

Datasets
We considered the daily streamflow observations throughout the tropics (23.5 • N-23.5 • S), with a buffer area extending from 30 • N to 30 • S, to accommodate even those river basins for which the headwater lies outside the tropics.This allows the inclusion of over 800 catchments that are archived across the Global Runoff Data Centre (GRDC 2015) and the India-Water Resource Information System (WRIS; https://indiawris.gov.in/wris/#/)repositories.The gauging sites are selected based on the following criteria: (A) the catchment size of at least 1000 km 2 or more.(B) A minimum of 70% of daily discharge time series must be available.(C) The catchment outlet location should not fall within a 10 km radius of a 'Mega Dam' (i.e.Dam height >15 m, Best 2019) (see supplementary information (SI); SI 1.1 for details).Following earlier assessments (Konapala and Mishra 2020, Kam 2021), we have identified catchments with near-natural flow and the discharge time series, which is least impacted by human influence.To this end, we have eliminated stations within a 10 km radius of a mega-dam (Kam 2021).Based on these criteria, 472 stream gauges were selected initially.Further, considering quality of streamflow records and availability of streamflow drought events, we scrutinize 135 sites over 19 IPCC reference climate regions across the tropic (figure S1, tables S1 and S2).Consistent with the literature (Fritsch andCarlson 1980, Gnauck 2004), we estimated the missing records in the streamflow series by time series interpolation using a shapepreserving piecewise cubic polynomial function at a daily time step (see SI 1.2 for details).On comparing observed versus infilled records, we find a good agreement between the two time series (see table S3 and figure S2).

Evaluation of shifts in drought attributes and selection of metrics for correlation
We compared streamflow drought responses over two non-overlapping climate normal periods (CNPs): the baseline climate period is referred to as CNP1 , whereas the recent climate period is denoted CNP2 .These CNPs represent a 30 year average of global climate as described by the World Meteorological Organization (WMO 2020).We use the daily variable threshold approach for the identification of streamflow droughts.Considering a relatively low threshold, for instance, a threshold value of 70% would result in several drought events, which may not qualify for extremes.In contrast, considering a relatively higher threshold, for instance, a 90% threshold value would yield a very few drought events.Therefore, similar to earlier assessments and being the most common choice (Heudorfer and Stahl 2017, Ahmadi and Moradkhani 2019, Ganguli et al 2022a, Brunner et al 2023), we have adopted a threshold of an 80% exceedance probability value for streamflow drought identification.Following previous studies (Brunner et al 2021(Brunner et al , 2023) ) and being the most common choice, we implemented a centered moving average of 31 days (d) to smooth the variable threshold time series.Further, a smoothing window excludes several short-term deficits resulting from variable thresholds, which cannot be considered droughts (Ahmadi and Moradkhani 2019).We further investigate the sensitivity of streamflow drought characteristics at different smoothing windows, such as 15 (±7 d), 21 (±10 d), 31 (±15 d), and 51 (±25 d) across a few selected representative catchments (figure S3).We find that the onset day and the drought frequency (table S4) remained relatively constant across catchments irrespective of the window lengths.We identify streamflow drought, if the discharge value is below the daily variable threshold persistently for 30 d or more (Brunner et al 2023).Next, we identify primary drought attributes, such as time to onset, duration and deficit volume (figure 1).Since streamflow timing follows a cyclical continuum, i.e., the ordinal day of an event at the end of a year can be close to an event occurring at the start of the following year, circular statistics are applied to investigate streamflow drought timing.We performed at-site trend analysis for individual drought properties using Sen slope (see SI 1.3 for details), while trend significance is determined using standard bootstrap-based procedure (Yue and Pilon 2004).
We further explored, the nonlinear changes in joint occurrences of deficit volume and onset pattern over the two CNPs considering nonparametric kernel density estimators (Van Loon et al 2014).For modeling drought deficit volume, we evaluate the applicability of several probability distribution families, including Gamma, GEV, log-normal and Loglogistic distributions.However, for modeling drought onset, we use the von Mises distribution (see SI 1.4).Following earlier assessments (Dhakal et al 2015, Rutkowska et al 2018, Dhakal and Palmer 2020), von Mises distribution is used for fitting drought onset due to its (i) analytical tractability over other complex circular distributions, and (ii) closer resemblance to a simple linear Gaussian distribution, but with the unique added value that its coordinates exist on a circular plane, which is the case for random variables describing the timing of an event.The circular distribution is focused on the perimeter of a circle with unit radius (Jammalamadaka and SenGupta 2001).To assess the uniformity of the drought onset data within the circular framework, we performed Rayleigh's test, which is a powerful statistical tool for measuring the degree of concentration or dispersion in circular data (Villarini 2016, Wasko et al 2020a).Table S5 shows for almost all gauges onset records qualify for circularity, which further confirms the applicability of von Mises distribution for modeling drought onset.For drought deficit volume, we select the best probability distribution across CNPs (table S6) following the minimum value of the Akaike Information Criteria with the small sample corrections (Chakrabarti and Ghosh 2011).Table S7 shows the fidelity of different distribution in fitting drought deficit volume for a few selected representative catchments.Finally, we evaluate the goodness-of-fit of selected distribution using (i) Kolmogorov-Smirnov test for drought deficit volume (table S8) (ii) Watson test for drought onset time (table S9).The goodness-of-fit was assessed at 10% significance level with p-value > 0.10.
Next, we compared the probabilistic shift of individual drought properties, i.e., onset time and deficit volume during recent versus the retrospective CNPs across each catchment using the entropy-based Kullback-Leibler (KL) divergence metric (Santhosh andSrinivas 2013, Ji et al 2022).Moreover, we assess the nonlinear changes in the dependence strengths using the non-parametric form of total linear-circular dependency (see SI 1.5; Mardia and Jupp 1999) and an extremal dependence metric, Capéraá-Fougères-Genest (CFG; Frahm et al 2005).The total dependency metric (TDM; linear-circular dependency) can better represent the pairwise clustered behavior of two variables towards the center of distribution.In contrast, the Upper tail dependency metric (UTDM) focus on the likelihood of two random variables simultaneously reaching extreme values (Hao and Singh 2016).

Changes in drought onset time, deficit volume and duration
The comparative assessment of trend in spatial pattern of onset, deficit volume and durations for two non-overlapping time windows (1961-1990 versus 1991-2020) are presented in figure 2. In the recent time window, 43% (58 out of 135) of catchments show increasing trends in deficit volume, and 68% (92 out of 135) of sites show a delayed onset.Further, we found only 27% (37 out of 135) sites depict an increase in deficit volume accompanied by delayed onset.The mean onset time of 60% of sites are clustered between January and March, with the interquartile range (75th-25th percentile) of median deficit volume remains 11.5 mm in CNP1 (figures S4(a) and (c)).In contrast, in CNP2, the mean onset time for around 65% of sites spread between January and April, while the number of sites showing onset during April in CNP2 is three times that of CNP1 (figure S4(b)).Further, in CNP2, we find a decrease in the median deficit volume across most regions, with an interquartile range of 8.5 mm, except in the South American Monsoon zone, where there has been an increase in deficit volume in recent period  (figure S4(d)).In South America region, >50% of catchments showed a delayed onset (figure 2 S5).Furthermore, the increase in relative frequency (number of events/year) of droughts are more prominent across the Northeast and South American monsoon region and the extratropical region of eastern coast of Australia (figure 3).However, there is a decrease in relative frequency of droughts in Western Africa, eastern coast of South Asia and a few catchments in Northern Australia (figure 3).

Spatial patterns of correlations between drought onset time and deficit volume
Figure 6(a) compares the TDM using linear (deficit volume)-circular (onset timing) dependence metrics of CNP2 versus CNP1.The difference in the frequency of TDM (figure 6(a); right) shows a weak asymmetric pattern with a negative skewness of close to zero.This suggests a little shift in TDM metrics over the two non-overlapping CNPs.However, at atsite level, >50% of sites across Asia and Africa, and >45% of sites in South America show notable positive changes in TDM dependence strengths, indicating stronger correlations in the recent period.In contrast, around 66% of sites in North America and 57% of sites across Australia display negative changes, indicating a decrease in dependence strengths in recent periods.Overall, 23% more sites show the dependence strengths of >0.5 in CNP2 compared to CNP1.In particular, Southern Asia, North-East South America, and North Australia regions showed stronger TDM in the CNP2 compared to CNP1 (figure S6(a)).
The spatial variation in the UTDM shows a bimodal nature of distribution for CNP2  (figure 6(b)).The leftward tail of CNP2 shows a slight more increase in frequency (i.e., counts) compared to CNP1, suggesting a leftward shift of overall distribution.The decrease in strength of dependence in CNP2 shows a decrease in bivariate hazard over South-East Asia, North-West South America and South East South America regions.Taken together, considering the UTDM >50% sites show a decrease in dependence strengths in CNP2.Although, this overall shift in CNP2 relative to CNP1 is not statistically significant.Nevertheless, a few sites show an increase in UTDM, explaining an increase in bivariate hazard.Such sites are clustered around Eastern-South Africa, Western-South Africa, East-North America and North-Central America regions (figures 7 and S6(b)).More sites show higher uncertainty bounds in UTDM than that of TDM, as depicted by the interquartile range.This could be due to the limited sample size at the upper tail quadrants of bivariate pair onset timedeficit volume.In contrast, in TDM all samples are considered for determining dependence strength across the central part of the bivariate distribution.Considering UTDM, notable uncertainty is observed in Western Africa, followed by the South-West South America regions (figure 7).On the other hand, considering TDM, North-Central America shows the highest uncertainty, followed by West-Southern Africa.

Discussion
Despite observational constraints, this study provides the first comprehensive assessment to quantify changes in drought attributes, such as onset time, deficit volume, and duration, in two non-overlapping time windows.Our results show that over the tropics, both onset time and deficit volume show a significant variability over two CNPs with a large spatial extent, showing a delayed arrival of drought onset but with longer duration and increased deficit volume (figure 2).These changes are consistent with the previous findings of global scale assessments of hydrological droughts (Gudmundsson et al 2021, Vicente-Serrano et al 2022).Further, we identify a few regional hotspots, such as north-eastern Brazil, West and South Africa, a few catchments in Westnorth America, and North-east Australia that show an increased deficit volume and increased frequency of droughts, which corroborates well with a few regional    The shift in precipitation (Loo et al 2015) leading to delayed arrival of southwest monsoon could have resulted in earlier drought onset in south-eastern India and increasing deficit volume of streamflow droughts in southeast Asia in CNP2 (figure 2(a)).Likewise, over extra-tropical southeast Australia (EAU in figure S1), a delayed shift in extreme precipitation timing have contributed to earlier arrival of streamflow droughts with increasing deficit volume in the latter time window (Wasko et al 2020a).Conversely, a decreasing trend in the seasonality strength of drought onset time in the tropical region of Australia (NAU; see figure S1) is associated with gradual wetter condition in this region (Wasko et al 2020b).The similar shift in drought timing and severity in central and northeastern South America is due to warming temperatures and southward shifts in large-scale convergence, major driver for South American precipitation (Chagas et al 2022).Likewise, our finding of a delayed occurrence of droughts with no significant trend in deficit volume, however with a few localized increasing trend in drought duration over Western Africa in CNP2 is due to the overall increase in precipitation since 1980s accompanied by frequent shorter dry spells (Bichet and Diedhiou 2018).We present a detailed analysis of changes in the timing of lower-than-average precipitation and the shift in the timing of streamflow droughts across a few representative stations, which confirms the hypothesis that a shift in rainfall may drive the changes in the timing of droughts (figure S7 and table S10).Precipitation patterns in the tropics are strongly influenced by shifts in sea surface temperature changes (Trenberth 2011, Larbi et al 2021, Bochow and Boers 2023), whereas streamflow responses are often controlled by changes in evapotranspiration, antecedent soil water content, and land atmospheric feedback (Farrick and Branfireun 2014, Wendling et al 2019, Setti et al 2020).While a detailed investigation of these drivers in controlling streamflow drought onset and associated changes is beyond the scope of the present study, understanding such changes requires an extensive examination across different IPCC reference regions.The increased variability in drought deficit volume can result from climatic shifts, atmospheric circulation changes, and changes in the landsurface feedback, as shown in earlier assessments (Laguë et al 2021, Alizadeh 2023).
Next, this study for the first time provides quantitative assessments of changes in dependence strengths between onset time and deficit volume across IPCC reference regions.A previous global scale study have shown interdependence of two linear drought properties (Ganguli et al 2022b), deficit volume and duration across different climate types and found a power-law scaling relationship between dependence metric and bivariate drought hazard, confirming interdependence between drought attributes substantially impact the likelihood of extreme drought and subsequently its hazard potential.The identified individual trends in drought attributes and the change in the dependence strengths across two climate-normal periods point toward apparent nonstationary signals in low flow series, which can serve as a basis for understanding multivariate drought risks in climate change projections.Evaluating change in the power-law distribution of interdependent drought attributes over time and its impact in mediating drought frequency could open up further research avenues to analyze hydrologic responses to climate change across global tropics.Our assessment shows that considering total and upper tail dependence measures, the West-Central Asia, Eastern-Southern Africa, North-East South America, Eastern North America and Northern Central America show substantially large positive changes, strengthening onset time-deficit volume relation in recent periods.These changes are due to nonstationarities in individual drought attributes, such as changes in drought seasonality and deficit volume, as demonstrated by the KL divergence metrics.Since risk is a function of hazard and consequences, the strength of dependency between drought attributes can significantly affects risk quantification (Zscheischler and Seneviratne 2017, Zscheischler et al 2018).The insights presented here have implications for resilient planning and management of water resources where low flow magnitude and its timing are critical.While this study statistically demonstrate a linkage between the timing of streamflow drought and deficit volume in observation, an assessment of the multivariate propagation of meteorological to hydrologic droughts considering interdependencies between onset pattern and deficit volume across large river basins can be considered as a potential future direction.
A few potential caveats for the analysis are worth highlighting.The observational constraints due to the uneven distribution of streamflow records at spatial and temporal scales across several tropical regions, such as the Caribbean, Africa, and Southeast Asia, limit precisely characterizing streamflow droughts across these regions.Our study does not consider the possible effects of human alterations, such as population expansion and socioeconomic developments, on streamflow droughts.However, the impact of the Anthropocene on streamflow drought onset pattern and its severity requires detailed analyses, such as population dynamics, socioeconomic considerations, and land-use alterations, which can significantly influence streamflow droughts.

Conclusions
Based on our analysis, the following key insights emerge: 1. Using ground-based daily streamflow records, we find around 43% spatial coverage show an increasing trends in streamflow drought deficit volume, while 68% show a delayed arrival of droughts in the recent time window (1991-2020) compared to the distant time slice .Only 27% of sites show an increase in deficit volume accompanied by delayed onset.2. Around 55% of the sites show an increase in drought frequency in recent times, with a few regional hotspots concentrated around Northeast and South American monsoon region and eastern coast of Australia.3. We show the distribution of onset time and deficit volume have substantially shifted towards larger values in the latter period compared to the earlier period in all regions with varying degrees, particularly in areas such as the Northeastern and South American Monsoon, Western Africa, East Southern-Africa, South Asia and North Australia.4. The onset time-deficit volume dependence strengthens in recent time window in more than half of the sites, whereas the strength of upper tail dependence decreases over majority of sites.The changes in dependence strengths of onset timedeficit volume may play a crucial role in bivariate risk quantification in areas with strong seasonality in streamflow.
(Ullah et al 2022, Ganguli et al 2022a), Southeast Asia (Zhang et al 2021, Ha et al 2022), sub-Saharan Africa (Faustin Katchele et al 2017, Gebrechorkos et al 2022), and the Caribbean (Herrera et al 2018, Moraes et al 2022), a comprehensive assessment of change in streamflow drought onset, its deficit volume and spatiotemporal investigation of their combined trends have not yet investigated across the global tropics.

Figure 1 .
Figure 1.Schematic representation of the detailed workflow.

Figure 2 .
Figure 2.The trends in drought attributes during two non-overlapping climate normals(1961-1990 versus 1991-2020).(a) Trend in time to drought onset; the shades of the circles in red and yellow indicate an earlier onset, whereas the shades of blue indicate delayed onset with a trend significance evaluated at 10% significance level.The significant trends in (b) deficit volume and (c) duration are shown with upward (increasing) and downward (decreasing) triangles.The circles in white show sites with insignificant trends.The donut chart shows the percentage of sites that show a significant earlier onset/decreasing trend and delayed onset/ increasing trend during the CNPs.

Figure 3 .
Figure 3.Comparison of drought frequency (number of droughts).(a) Average frequency of droughts per year during the two time periods.(b) The percentage changes in drought frequencies during the recent (CNP2) versus the distant time windows (CNP1).

Figure 4 .
Figure 4. Joint occurrences in drought onset versus deficit volume trends.Classification of trends in streamflow drought attributes into earlier (or delayed) onset together with increasing (or decreasing) drought deficit volume.Each quadrant represents onset versus deficit volume changes.The at-site significant trend in the scatter plot in CNP1 is shown using a square symbol, whereas a circle denotes the same for CNP2.The shades of the markers showing the IPCC reference regions (shown in the color bar) portray at-site significant trends for both onset time and deficit volume.If only the trend in onset time is significant, then it is shown using a symbol with no shade, whereas if only the deficit volume is significant, it is marked using a symbol in gray shade.The dashed and solid black lines indicate the slope changes during the CNP1 and CNP2 respectively.The marginal PDFs in (dashed) solid blacks and solid red represent the past and recent time windows, respectively.
scale assessments (McGree et al 2016, Alves da Silva et al 2020, Tripathy et al 2023).While increase in deficit volume in streamflow drought is often linked to high headwater demand and land cover changes, such as in north-east Brazil and southern Australia (Vicente-Serrano et al 2022), the trends in onset timing is consistent with large-scale shifts in precipitation (Ashfaq et al 2009, Loo et al 2015).

Figure 5 .
Figure 5. Probabilistic shifts in drought characteristics, (a) onset and (b) deficit volume in the recent versus distant time windows.The right panels portray Hovmöller diagrams showing the KL divergence, depicting the extent of shifts (for 1991-2020 versus 1961-1990) on the X-axis, whereas the latitude is shown on the Y-axis.

Figure 6 .
Figure 6.Spatial maps depicting changes in (a) total (TDM) and (b) upper tail dependence (UTDM) in the recent versus distant time windows.The circles with shades of blue to green show a decrease in dependence strengths, whereas those with brown shades in progression show sites with increasing dependence strengths respectively.The right panel in (a) presents the probability density function (PDFs) depicting the difference (distant-recent) in TDM during the two CNPs.The right panel in (b) compares the PDFs of the UTDM values of CNP2 versus CNP1.The PDFs with red shade portray the recent time window, whereas in blue shows the retrospective era.

Figure 7 .
Figure 7.Changes in dependence strengths over climate normal periods across the IPCC reference climate regions.The boxplot showing the differences(1991-2020 versus 1961-1990) in (a) total versus (b) upper tail dependencies.Shades in the boxplots show the same IPCC reference climate regions as in figure4.The vertical black lines in the boxplots represent the median value, whereas the spread shows the interquartile range.The '+' symbols in red show the outliers.The vertical lines in red perpendicular to the zero mark indicate normal conditions with no changes, with positive changes, suggesting an overall increase in dependence strengths.In contrast, the negative changes indicate a decrease in dependence strengths.
(a)).Interestingly, overall >60% of catchments illustrate significant (p-value < 0.1) decreasing trends in deficit volume in CNP1 (figure 2(b)).On the other hand, the percentage share of sites showing increasing trends in deficit volume with decreasing trends in durations are nearly equal (figure2(c)).Compared to the CNP1, we find in CNP2, more sites show a delayed arrival of streamflow droughts accompanied by an increase in deficit volume with a longer drought duration (figure2; right panel).A persistently prolong drought with an increase in deficit volume is apparent in 14% (10 out of 73) of catchments in South America and in a few catchments in northern Australia, such as Gregory Downs.In contrast, droughts over Southern Africa tend to become lesser severe with decrease in duration (figures 2(b) and (c)).Over the eastern coast of South Asia, an earlier onset of streamflow droughts is generally accompanied by a shift towards longer drought duration during CNP2 (figures 2(a) and (b): right).The South American continent shows the maximum changes in drought attributes (figure et al 2017 Global patterns of drought recovery Nature 548 202-5 Setti S, Maheswaran R, Radha D, Sridhar V, Barik K K and Narasimham M L 2020 Attribution of hydrologic changes in a tropical river basin to rainfall variability and land-use change: case study from India J. Hydrol.Eng.25 05020015 Sheffield J, Andreadis K M, Wood E F and Lettenmaier D P 2009Global and continental drought in the second half of the twentieth century: severity-area-duration analysis and temporal variability of large-scale events J. Clim.22 1962-81 Slater L J et al 2021 Nonstationary weather and water extremes: a review of methods for their detection, attribution, and management Hydrol.Earth Syst.Sci. 25 3897-935 Sternberg T 2011 Regional drought has a global impact Nature 472 169 Teutschbein C, Quesada Montano B, Todorović A and Grabs T 2022 Streamflow droughts in Sweden: spatiotemporal patterns emerging from six decades of observations J. Hydrol.42 101171 Trenberth K 2011 Changes in precipitation with climate change Clim.Res.47 123-38 Tripathy K P, Mukherjee S, Mishra A K, Mann M E and Williams A P 2023 Climate change will accelerate the high-end risk of compound drought and heatwave events Proc.Natl Acad.Sci. 120 e2219825120 Ullah I, Ma X, Ren G, Yin J, Iyakaremye V, Syed S, Lu K, Xing Y and Singh V P 2022 Recent changes in drought events over South Asia and their possible linkages with climatic and dynamic factors Remote Sens. 14 3219 UNCCD 2022 Desertification and drought day 2022 UNCCD (available at: www.unccd.int/events/desertification-droughtday/2022)Van Loon A F, Tijdeman E, Wanders N, Van Lanen H J, Teuling A J and Uijlenhoet R 2014 How climate seasonality modifies drought duration and deficit J. Geophys.Res.Atmos.119 4640-56 Vicente-Serrano S M et al 2013 Response of vegetation to drought time-scales across global land biomes Proc.Natl Acad.Sci.110 52-57 Vicente-Serrano S M, Peña-Angulo D, Beguería S, Domínguez-Castro F, Tomás-Burguera M, Noguera I, Gimeno-Sotelo L and El Kenawy A 2022 Global drought trends and future projections Phil.Trans.R. Soc.A 380 20210285 Villarini G 2016 On the seasonality of flooding across the continental United States Adv.Water Resour.87 80-91 Wasko C, Nathan R and Peel M C 2020a Changes in antecedent soil moisture modulate flood seasonality in a changing climate Water Resour.Res.56 e2019WR026300 Wasko C, Nathan R and Peel M C 2020b Trends in global flood and streamflow timing based on local water year Water Resour.Res.56 e2020WR027233 Wendling V, Peugeot C, Mayor A G, Hiernaux P, Mougin E, Grippa M, Kergoat L, Walcker R, Galle S and Lebel T 2019 Drought-induced regime shift and resilience of a Sahelian ecohydrosystem Environ.Res.Lett.14 105005 Wilhite D A, Svoboda M D and Hayes M J 2007 Understanding the complex impacts of drought: a key to enhancing drought mitigation and preparedness Water Resour.Manage.21 763-74 WMO 2020 WMO climatological normals (World Meteorological Organization) (available at: https://community.wmo.int/en/wmo-climatological-normals) Yue S and Pilon P 2004 A comparison of the power of the t test, Mann-Kendall and bootstrap tests for trend detection/Une comparaison de la puissance des tests t de Student, de Mann-Kendall et du bootstrap pour la détection de tendance Hydrol.Sci.J. 49 21-37 Zhang L, Chen Z and Zhou T 2021 Human influence on the increasing drought risk over Southeast Asian monsoon region Geophys.Res.Lett.48 e2021GL093777 Zscheischler J et al 2018 Future climate risk from compound events Nat.Clim.Change 8 469-77 Zscheischler J and Seneviratne S I 2017 Dependence of drivers affects risks associated with compound events Sci.Adv. 3 e1700263