Abstract Precipitation is important for hydrological monitoring and modeling. The accuracy of mean areal precipitation (MAP) estimation relies largely on the configuration of the precipitation network. This study proposes a spatial pattern–based framework for optimizing rain gauge networks that uses the climatological precipitation field from high-quality reanalysis data as an external benchmark to evaluate MAP estimation. In contrast to entropy- and variance-based methods, the framework requires no long in situ records, explicitly matches both the magnitude and the full spatial distribution of the precipitation field, and can assess as-yet-uninstrumented locations. Using the Qingyi River basin as a case study, we employ the China Meteorological Administration Multisource Precipitation Analysis System (CMPAS) fusion data to characterize precipitation spatial patterns. The framework introduces two metrics, the mean areal precipitation bias (MB) and the Kullback–Leibler (KL) divergence, to quantify differences between gauge-derived and CMPAS-derived precipitation fields. Evaluation of the current network reveals significant MAP estimation discrepancies in subbasins with high precipitation variability. Through a significance index analysis of candidate gauges and a hierarchical optimization, strategic gauge placement reduced MAP bias from up to 16% to below 1% in the optimized subbasins. An independent hydrological validation with the Xin’anjiang model showed that the optimized network improved streamflow simulation, with the basin-average Nash–Sutcliffe efficiency increasing from 0.63 to 0.76, confirming that better MAP estimation propagates to improved hydrological prediction. The framework is computationally efficient and applicable to both network expansion and rationalization.
Abstract Anthropogenic warming is projected to enhance future drought risk in the northeastern United States, yet the region has experienced recent precipitation increases. Here, we investigate how seasonal Northeast drought has changed over the last century in response to hydrologic intensification. To assess drought from 1901 to 2022, we use the Palmer Drought Severity Index calculated with Global Historical Climatology Network station data. We find that Northeast drought across all months is ∼59% less frequent and ∼19% less intense over the last 40 years relative to 1901–1983. Warm season drought frequency and intensity decreased by ∼56% and ∼16%, respectively. These results are spatially consistent throughout the Northeast. The variance in monthly dry and wet events has not significantly changed, suggesting that the Northeast hydroclimate is not more volatile than in the past. We explore the potential drivers of evolving drought by rerunning our analysis with changes in temperature and precipitation removed, finding that reduced drought is attributable to precipitation increases, which to date have outpaced warming-driven evapotranspiration increases. Collectively, our results show a marked reduction in seasonal drought since 1984 in contrast with the enhanced regional drought risk projected across generations of climate model ensembles under anthropogenic forcing. However, these findings do not preclude the possibility of historical increases in drought at sub-seasonal timescales or enhanced future drought. Our results inform current and future Northeast water management decisions, as well as efforts to prioritize climate change hazards.
Abstract Flash droughts, defined by their rapid intensification, are typically treated as distinct from conventional, slow-evolving droughts. This study challenges this separation by investigating flash and conventional droughts as a unified continuum. Leveraging a cloud-tracking methodology, we develop an object-based drought inventory containing 868 drought events that occurred over the conterminous U.S. during the 1979–2023 growing seasons (April–October). These “drought objects” had an average duration of 6 weeks and an average area of 480,000 km 2 . We categorized them as conventional drought-dominated (394), mixed (401), or flash drought-dominated (73) based on the fractional area experiencing rapid intensification. A Lagrangian tracking analysis demonstrates that while more than half of these objects remain geographically anchored, the south-central U.S. emerges as a primary corridor for active drought genesis and migratory propagation, exhibiting a strong tendency for northeastward movement. Our analysis reveals a critical finding: flash-dominated drought objects are associated with more extensive areas and longer durations than their conventional or mixed counterparts. Both the area-averaged intensification rate and the fractional area of flash drought significantly correlate with the object's duration, affected area, and severity. This suggests that rapid intensification during flash drought is a key mechanism driving drought expansion and persistence.
Abstract Reliable assessment of Mediterranean precipitation remains challenging because gridded datasets differ in their construction, observational constraints, and representation of strong spatial variability. Although station-based studies show that precipitation characteristics and trends vary across the region, widely used gridded products have not been consistently evaluated against extensive station-derived references. Here, we assess seven products for 1981–2020 against two complementary station-derived benchmarks using mean annual precipitation, interannual variability, long-term slope, and annual maximum 1-day precipitation. The benchmarks combine published statistics from more than 23 000 stations in 27 countries with extreme precipitation records from HYADES (Archive of Yearly Maxima of Daily Precipitation Records), a global archive of annual maximum daily precipitation (RX1DAY) derived from Global Historical Climatology Network-Daily observations. After aggregation, the reference network represents 2096 grid cells and covers 52% of nondesert land cells in the study domain. Multi-Source Weighted-Ensemble Precipitation (MSWEP), Multi-Source Weather (MSWX), and the Global Precipitation Climatology Centre (GPCC) product show the most consistent performance for mean precipitation, variability, and slope. Across matched grid cells, typical product-level root-mean-square errors against the station-derived references are 60–80 mm for mean annual precipitation, 25 mm for interannual variability, and 1.5 mm yr −1 for slope. For RX1DAY, the European daily high-resolution gridded dataset (E-OBS) performs best ( R 2 = 0.57), whereas the other products reach 0.18–0.30 and generally underestimate the strongest extremes. A separate analysis for 2001–20 shows stronger product agreement for precipitation changes associated with wet-day frequency than with wet-day intensity. These findings support more informed dataset selection for Mediterranean drought, flood, and water resource assessments. Significance Statement Global datasets are widely used to study changes in Mediterranean precipitation, but they do not describe precipitation equally well. We find that Multi-Source Weighted-Ensemble Precipitation (MSWEP), Multi-Source Weather (MSWX), and Global Precipitation Climatology Centre (GPCC) best represent average conditions and year-to-year variability across the region. European daily high-resolution gridded dataset (E-OBS) captures annual daily extremes more closely than the other products, although all datasets fail to capture the strongest events. Agreement is much weaker for long-term slopes, indicating greater interproduct uncertainty in estimated temporal changes than in mean-state precipitation characteristics. These differences mean that conclusions about drought risk, flood potential, and water availability can vary depending on the dataset used. Our results support more careful dataset selection in climate assessments and practical planning for water resources and related hazards across the Mediterranean region.
Abstract Elucidation of variation and environmental drivers of evapotranspiration (ET) are crucial for improving our understanding of climate change, water cycle, and ecosystem stability in the Yellow River basin (YRB). Based on ERA5-Land reanalysis data, this study analyzed the variations in summer ET and its environmental drivers in the YRB. Results showed that ET exhibited a notable downward trend during 1980–2021, with the fastest rate of −1.52 mm yr −1 in the downstream region of the YRB (DYRB), a rate of −1.35 mm yr −1 in the Hetao region of the YRB (HYRB), and a moderate level in the upper region of the YRB (UYRB). In the UYRB, HYRB, and DYRB, EOF1 primarily characterized the ET climatology, showing a notable decline around 2000, and EOF2 exhibited a dipolar spatial pattern with contrasting trends in ET distribution. The results suggested that the effect of climate warming on ET is stronger than vegetation, indicating that climate change is the dominant factor influencing ET variations in the YRB. In the UYRB, reduced net radiation (Rn) contributed substantially (−12.9%) to the decline in ET, presenting an “energy-limited” pattern. Owing to the limited water supply in the HYRB, reduction in precipitation P resulted in drier soil, combined with an increase in runoff, leading to a reduction in ET, presenting a “soil moisture–limited” pattern, as evidenced by the interannual temporal correlation coefficients of ET with P (0.49), soil moisture (SoilM; 0.54), root zone soil moisture (RZSM; 0.61), and runoff (−0.56). In the DYRB, the contributions of Rn, SoilM, and RZSM to ET variation are −9.2%, −11.7%, and −8.6%, respectively. The decrease in Rn, SoilM, and RZSM, combined with the increase in runoff, leads to the decrease in ET in the DYRB, presenting an “energy- and soil moisture–limited” pattern in the DYRB. Significance Statement The primary objective of this study is to explore the spatiotemporal patterns and potential driving factors of evapotranspiration (ET) in the Yellow River basin using the ERA5-Land dataset. This is important because ET is a key link of energy and water cycles and also influences the carbon cycle. ET is currently undergoing modifications and is expected to become more pronounced as the climate continues to warm. Our findings showed that the variations of summer ET were characterized by its climatic factors and quantified the dominance of energy and soil moisture limitations across its diverse subbasins. Given that the drivers are numerous and continuously change under climate change, it is necessary to pay attention to the variation and attribution of ET.
Abstract Seasonal forecasting in operational centers has focused predominantly on prediction of temperature and precipitation. Here, we use the skill of model forecasts against observations (actual prediction skill) and against each model ensemble member (perfect-model skill) to assess the predictability of these two variables and five additional hydrological variables across two distinct hydrologic regions (the Missouri basin and California) of the United States. The forecasts from two operational coupled models [National Centers for Environmental Prediction Climate Forecast System version 2 (NCEP CFSv2) and European Centre for Medium-Range Weather Forecasts Seasonal Forecasting System version 5 (ECMWF SEAS5)] are used. Results show variables with high memory, such as soil moisture, total water storage, and snow water equivalent, have both high perfect-model and actual prediction skill. Runoff and evapotranspiration, which are highly dependent on the amount of water storage, generally have intermediate skill. Subbasins within these two hydrologic regions show similar results. The high memory variables also maintain high skill for longer prediction lead and exhibit less seasonal variability, particularly over the Missouri basin region which experiences a lower amplitude in the annual cycle. However, over California, skill drops off more quickly with increasing lead and has a strong seasonal cycle with the highest skill during late spring and early summer. Results also demonstrate that actual prediction skill is positively related to perfect-model skill for all variables across different regions and seasons which give some promise to using perfect-model skill in ungauged areas of the world as a proxy to their real-world skill. Significance Statement Operational seasonal forecasting models are commonly evaluated based on their predictions of temperature and precipitation. Here, we assess the forecasts for these two variables and five additional variables associated with the hydrologic cycle across two distinct hydrologic regions of the United States. We find that total water storage, soil moisture, and snowpack are more dependent on past values (i.e., with high memory) and hence have the highest predictability. Evapotranspiration and runoff, which are highly dependent on water storage, have higher predictability than temperature and precipitation. Also, the variables and regions with the highest memory maintain higher skill at longer lead times and have consistent skill throughout the year. Basins and seasons that have higher intrinsic model predictability also have higher skill predicting real-world values for all variables.
Abstract The Volta River basin in West Africa, covering a substantial area of Ghana, faces significant water management challenges due to highly variable rainfall and the limited availability of high-temporal-resolution rain gauges. This study, developed under the 2024 Global Precipitation Measurement (GPM) mission mentorship program, evaluates precipitation in the Volta River basin using high-resolution satellite-derived data from the Integrated Multi-satellitE Retrievals for GPM (IMERG). In the absence of subdaily rainfall data, comparisons were made between daily precipitation from local rain gauges and IMERG version 06 and version 07 (Early, Late, and Final) runs. Results indicated a very weak correlation between daily precipitation estimates from IMERG and gauge data but a significantly improved correlation for monthly estimates. In addition, version 07 performed better than version 06 in many of the metrics used including, higher r -squared, lower root-mean-square error, and improved bias. Extreme precipitation days and no-rain days were also compared to determine how well IMERG estimated occurrences. While IMERG failed to accurately estimate extreme precipitation, version 07 estimates performed better than version 06 estimates. IMERG also consistently estimated light precipitation when rain gauges observed no precipitation. While this study shows that IMERG did not perform well in the region, it also highlights the improvements from version 06 to 07. This study contributes to our understanding of precipitation estimates from satellite remote sensing in the Volta River basin which can enhance disaster risk management and climate resilience in Ghana.
With global warming, the intensity and frequency of floods have markedly increased, resulting in substantial losses of life and property. The Pearl River basin (PRB) in South China, with its complex topography, remains highly susceptible to flooding. To enhance the precision of flood simulation and forecast in the PRB, an overland flow scheme was first integrated into the community Noah land surface model with multiparameterization options (Noah-MP) and subsequently coupled with the Weather Research and Forecasting (WRF) Model. These models were applied to a record precipitation event occurring over the PRB in April 2024 to validate their improvements. Results reveal that the modified Noah-MP can effectively simulate hydrological processes. The cumulative surface runoff is strongly affected by topography and has a higher magnitude in low-lying areas. The accumulated water depth generally aligns with the satellite-observed inundation, and the error in soil moisture between the model and the observation is reduced. Further, the modified WRF Model has successfully reproduced the inundation area in most regions, contrasting with the original scheme's inability to simulate flooding. In addition, the improvement in hydrological processes in the modified WRF also enhances the ability to simulate precipitation through land-atmosphere interactions. A comparison with the WRF-Hydro simulations further demonstrates that our scheme achieves a certain degree of improvement in simulating inundation. This study presents a promising approach for improving flood simulations in complex topography, which is instrumental in mitigating the loss of life and property caused by flood disasters in the PRB.
This study investigates spatiotemporal changes in socioeconomic exposure to tropical cyclone precipitation (TCP) extremes across China using high-resolution rainfall and gridded population datasets. Results show that extreme TCP exhibits a significant increasing trend over central-southern China, while decreasing trends are observed in parts of the Yangtze River delta. These contrasting patterns are linked to changes in TC characteristics, including intensified precipitation rates and reduced translation speeds over coastal regions, which enhance rainfall persistence and accumulation and thereby contribute to increasing extreme TCP events inland. Regions with high socioeconomic exposure are mainly concentrated in southeastern and central-eastern China. Based on a modified exposure index that integrates physical hazard and socioeconomic factors, we find that although exposure is jointly determined by these components, its recent increase is primarily driven by intensified TCP intensity and duration, with population growth and economic development further amplifying the impacts. Population and gross domestic product exposure to extreme TCP events (>200 mm) increase by approximately 27% and 17%, respectively, with pronounced spatial heterogeneity. Rapid urbanization and economic expansion in recent decades have further elevated vulnerability to TC-induced precipitation extremes. This multiscale framework provides actionable insights for prioritizing mitigation and adaptation strategies, enhancing climate resilience in high-risk zones.
Abstract This research develops a deep learning (DL)–based algorithm to fill gaps in satellite data by leveraging multisource input combinations, assessing temporal information impact, and implementing a hybrid loss function that combines Kullback–Leibler (KL) divergence with mean-square error (MSE) computed in logarithmic space and mean absolute error (MAE) to improve the model’s ability to capture heavy rainfall events. The model uses a convolutional neural network U-Net (CNN-UNet) architecture to fuse infrared brightness temperature (IR Tb) from National Oceanic and Atmospheric Administration (NOAA)’s geostationary satellites with merged microwave precipitation (MWprecipitation) from National Aeronautics and Space Administration (NASA)’s Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM)–Early Run (IMERG-E), delivering a temporally and spatially complete trajectory of rainfall. The model is trained on June–August (2014–17) data and tested on June–August (2018) data. The NASA Ground Validation Multi-Radar Multi-Sensor (GV-MRMS) system serves as the reference dataset, while the NASA uncalibrated IMERG-E (precipitationUncal) precipitation is used as the baseline. Different CNN-UNet configurations are explored using varying input combinations and loss functions. Results show that CNN-UNet effectively reconstructs missing MWprecipitation data by exploiting spatial and temporal patterns in IR Tb. Models incorporating both data sources, temporal context, and a balanced hybrid loss function outperform simpler configurations. The best-performing CNN-UNet configuration shows good agreement with MRMS observations, better captures rainfall frequency distributions, and shows improvements in continuous and categorical metrics for summertime precipitation relative to near-real-time IMERG-E over the study area. Furthermore, comparing CNN-UNet and IMERG-E with MRMS across several heavy rainfall events shows promising results. These findings highlight the potential of CNN-UNet for near-real-time satellite rainfall estimation.
Abstract Hydroclimatic extremes are intensifying across the eastern Himalaya, yet watershed-scale evidence for adaptation planning remains limited. This study evaluates historical (1980–2020) and projected (2021–95) temperature and precipitation extremes in the Dudhkoshi watershed, Nepal, using daily data from nine precipitation and two temperature stations. Eleven precipitation and twelve temperature indices were computed using ClimPACT2 at annual and seasonal scales. Future changes were evaluated for near- (2021–45), mid- (2046–70), and far-future (2071–95) periods under shared socioeconomic pathway (SSP) 2-4.5 and SSP5-8.5 scenarios using a selected multi–global climate model (GCM) ensemble. GCM outputs were bias-corrected using a Bernoulli–Weibull distribution transformation for precipitation and nonparametric quantile mapping with robust empirical quantiles (tricub) for temperature. Observations indicate significant warming in the Middle Mountain (MM) and High Mountain regions, characterized by increasing hot days, declining cold days, and a rise in consecutive dry days (+0.68 days yr −1 ), alongside reductions in wet-day totals and heavy-rainfall indices at multiple stations. Projections indicate continuous warming across all periods, strongest under SSP5-8.5. By the late century, annual maximum and minimum temperatures in the MM are projected to increase by up to 3.34° and 4.58°C, respectively, alongside an overall increase in precipitation. Monsoon and postmonsoon heavy extremes will most likely intensify mostly in MM (Rx1day + 0.44 mm yr −1 ; Rx5day + 1.69 mm yr −1 ), with premonsoon drying under SSP2-4.5 and amplified winter precipitation in the High Himalaya under SSP5-8.5. Ultimately, the MM region emerges as the most climate-sensitive zone due to compounding warming and hydroclimatic shifts, highlighting the need for physiography- and season-specific adaptation strategies to enhance resilience in Himalayan watersheds.
Abstract This study quantified the modulating effects of ice-phase microphysical processes on surface precipitation intensity in typhoon spiral rainbands using convection-permitting numerical simulations (3-km grid spacing) of Supertyphoon Lekima (2019). The Weather Research and Forecasting Model with the Morrison double-moment microphysics scheme was employed to diagnose ice process rates and their contributions to surface precipitation. The simulated 48-h accumulated precipitation was validated on a common 0.1° grid against Global Precipitation Measurement (GPM) Integrated Multi-satellitE Retrievals for GPM (IMERG) (V07B) retrievals, and the simulated peak accumulations along the Zhejiang coast were further compared with regional automatic weather station gauges, supporting the model’s representation of the rainband structure. Results revealed that ice-phase processes dominated precipitation formation, contributing 68%–92% of surface rainfall through melting of ice particles below the freezing level. Graupel melting emerged as the primary mechanism, accounting for 52% of surface precipitation in convective cores, followed by snow melting (31%) and cloud-ice-origin particles (17%). Systematic differences between inner and outer rainbands were identified, with inner rainbands exhibiting 40% higher ice water content and enhanced graupel production due to stronger updrafts. The storm-centered environmental shear remained within a moderate regime of roughly 11–14 m s −1 throughout the analysis period, and against this large-scale background, a gridpoint sensitivity analysis showed that vertical wind shear and midlevel humidity strongly modulated ice-phase precipitation efficiency, with moderate shear reducing efficiency from 0.78 to 0.52. Pronounced diurnal variations and marked transformations during landfall highlighted complex interactions between thermodynamic, dynamic, and microphysical processes. These findings provide insights for improving typhoon precipitation forecasts and underscore the importance of accurately representing ice-phase processes in operational weather prediction models for typhoon-affected regions. Significance Statement When typhoons bring devastating rainfall to coastal communities, the rain actually begins as ice crystals high in storm clouds that melt as they fall to Earth. This study reveals that ice formation processes control 68%–92% of typhoon rainfall reaching the ground, with soft hail particles called graupel being the dominant contributor. Current weather prediction models often poorly represent these ice processes, leading to inaccurate rainfall forecasts during typhoons. By analyzing Supertyphoon Lekima using advanced computer simulations and satellite observations, this research provides the scientific foundation for improving typhoon rainfall predictions. A better understanding of these ice processes will help meteorologists develop more accurate weather models, ultimately leading to improved warnings that could save lives and reduce economic losses for the millions of people living in typhoon-prone regions worldwide.
Abstract This study evaluates the ability of CONUS404, a 4-km resolution historical climate reconstruction that uses a coupled atmosphere–land surface model driven by a global reanalysis, to accurately simulate the precipitation, temperature, and snow water equivalent (SWE) across the mountainous western United States by comparing it against a variety of datasets including observations, statistical interpolations, and data assimilation products. The analysis was performed across four subdomains with distinct hydroclimates over the period from 1985 to 2021. CONUS404 cold-season precipitation generally agrees well with gauge-based gridded estimates, with some exceptions. CONUS404 matches observationally based SWE estimates during the early accumulation period but consistently exhibits earlier-peaking, shallower snowpacks with slower late-season ablation rates. CONUS404’s good agreement in terms of seasonal precipitation and SWE during the accumulation period (early in the cold season) suggests that snowfall is not the primary driver of CONUS404’s negative SWE bias that begins to emerge in January. While the average temperatures are a fairly good representation, CONUS404 cannot adequately simulate the full extent of nighttime cold temperatures (minima are warm biased) or daytime warm temperatures (maxima are cold biased), resulting in a significantly reduced daily temperature range. This study cannot fully attribute the cause of the negative SWE bias in CONUS404 in late winter, and further investigation is required. Significance Statement Streamflow in the interior western United States and other midlatitude semiarid mountainous regions strongly depends on cold-season orographic precipitation and seasonal snowpack. Therefore, an accurate, highly resolved description of these fields, particularly over snow-dominated mountains, is essential for high-fidelity, physics-based watershed hydrologic predictions. Here, the accuracy of a 4-km resolution historical climate reconstruction that uses a coupled atmosphere–land surface model is examined. While precipitation is captured rather well, the peak seasonal snowpack in late winter/early spring is considerably less than observations suggest. This discrepancy calls for a better understanding of the surface energy balance and snow ablation rate at snowpack measurement sites.
Abstract The persistent warming of the North China Plain (NCP) poses a severe threat to local livelihoods and production. However, attributing this warming trend remains challenging due to the complex influences of internal climate variability and underlying surface changes. Utilizing ERA5 reanalysis data, this study decomposes 2-m temperature into two components: the atmospheric baseline ( T base ) and the surface-induced effects ( T sfc ). It quantifies the contributions of both atmospheric and surface effects to the warming trend in the NCP, emphasizing the dominant role of atmospheric changes in the spatiotemporal variation of this warming. The results show that after 1996, summer temperatures in the NCP began to rise at an accelerated rate of 0.35°C decade −1 , with atmospheric changes serving as the primary driver of this acceleration, contributing 62.9% to the subsequent warming trend. Meanwhile, T sfc sustained a consistent warming trend across 1968–2023. Furthermore, this study applies a dynamic adjustment algorithm to decompose the atmospheric warming trend into two components: circulation and thermodynamic effects, revealing that the spatiotemporal variation in the trend of T base is primarily driven by circulation changes triggered by internal climate variability. Significance Statement Rapid warming in the North China Plain (NCP) severely threatens agricultural production and local livelihoods. However, identifying the exact causes remains challenging because atmospheric weather patterns and surface changes are deeply intertwined. Using comprehensive climate datasets, this study decomposes 2-m temperatures into an atmospheric baseline ( T base ) and surface-induced effects ( T sfc ). After 1996, summer warming accelerated significantly to 0.35°C decade −1 . Atmospheric changes—driven by natural climate cycles—dominated this trend, accounting for 62.9% of the warming and dictating its spatial distribution. Meanwhile, the ground surface provided a sustained, consistent background warming effect throughout 1968–2023. Accurately attributing these drivers provides a crucial scientific basis for implementing targeted measures to safeguard regional food security and mitigate climate-related disasters.
Abstract This study investigates soil moisture decorrelation time scales, commonly referred to as soil moisture memory, by examining their sensitivity to meteorological forcing, specifically the autocorrelation structure in precipitation data and land–atmosphere coupling using the Community Land Model, version 5 (CLM5). We conduct CLM5 experiments using two widely used meteorological datasets: the Climate Forecast System Reanalysis (CFSR) and the Global Soil Wetness Project phase 3 (GSWP3), along with randomized meteorological forcing to isolate the role of climate variability and persistence on soil moisture decorrelation time scales. Results show that the CFSR-forced CLM5 simulation yields decorrelation time scales that are, on average, twice as high as those from the GSWP3-forced simulation, particularly in tropical and subtropical regions, due to significant precipitation autocorrelation and enhanced soil moisture–precipitation feedback in the CFSR case. Randomized meteorological forcing significantly reduces decorrelation time scales in CFSR-forced CLM5 simulations (by 50%–70%) but only marginally in GSWP3-forced simulations (by 10%–20%). Additional analysis using the drydown time-scale metric reveals minimal differences between CFSR- and GSWP3-forced simulations, highlighting the dependency of findings on the memory metrics and the role of land surface hydrologic processes. Comparisons with the Community Earth System Model version 2 (CESM2) Large Ensemble (CESM2-LE) reveal that a fully coupled model underestimates the decorrelation time scales relative to the CFSR-forced CLM5 simulation and aligns more closely with the GSWP3-forced CLM5 simulation. Soil moisture reemergence, identified as a secondary autocorrelation peak, disappears under randomized forcing, indicating its dependence on climate variability rather than solely on land surface processes. Significance Statement Soil moisture memory enables the land to “remember” past wet or dry periods, influencing future weather, particularly water availability and droughts. In this study, we utilized a climate model to investigate how various weather datasets impact the soil moisture memory. One dataset [Climate Forecast System Reanalysis (CFSR)] had more persistent weather patterns than the other [Global Soil Wetness Project phase 3 (GSWP3)]. We found that, on average, the soil moisture decorrelation time scale was twice as high when using CFSR, particularly in tropical and subtropical regions. We also ran a test where we scrambled the weather patterns to remove their persistence. As a result, decorrelation time scales were reduced sharply in the CFSR case (by 50%–70%) but only slightly in GSWP3 (by 10%–20%). This indicates that long-term memory of soil moisture depends not only on the land but also on the coupling and persistence from the atmosphere. We also found that when weather patterns are random, a process called “soil moisture reemergence,” in which buried moisture signals resurface, disappears. Our results suggest that improved weather data and feedback models are necessary to enhance long-range weather and drought forecasts.
Abstract Sound observations are essential for understanding lake–atmosphere energy exchanges, which govern lake water levels and energy budgets and are critical for weather forecasting, water quality assessment, and navigation. While advances have been made with observation programs on large lakes, such as North America’s Laurentian Great Lakes, current fixed-location platforms measure relatively little of the spatial variability in turbulent fluxes across these large systems. Here, we apply ship-borne eddy-covariance-derived turbulent flux measurements that reveal the presence of evaporation zones, defined as contiguous areas of upward latent heat fluxes, separated from areas of downward latent heat fluxes by turbulent flux boundaries. Boundaries continue to migrate with time, with atmospheric conditions and surface water temperatures, but collapse as the lake warms through summer and the entire lake area begins to evaporate. These results have implications for how evaporation should be estimated, which would be improved with distributed approaches. Coupled lake–atmosphere representation in predictive platforms is likely necessary to capture these spatial dynamics, which would improve water level prediction, lake thermal and ecosystem characterization, and weather forecasting. Last, the date of turbulent flux boundary collapse could be used as a new metric of lake function and state that augments existing metrics such as date of stratification.
Abstract Climate change has intensified variations in the terrestrial water cycle, increasing the occurrence of extreme events such as droughts, floods, and compound events. Using long-/short-cycle drought–flood abrupt alternation (DFAA) indices, combined with the random forest regression and the Shapley additive explanations method, this study aims to explore the spatiotemporal variation and driving mechanisms of DFAA indices, i.e., drought-to-flood (DTF) and flood-to-drought (FTD) events, in the 218 tertiary river basins of China from 1951 to 2020. Results show that regions with significant decreases in the long-cycle DFAA index are mainly in northern China, where FTD events tend to occur, with their gravity center shifting northward by approximately 328 km. Conversely, regions with an increasing DFAA index are mostly in southern China, where DTF events are prone to happen, with their gravity center shifting southward by approximately 325 km. A similar migration pattern is observed for short-cycle DFAA events: FTD hotspots move northward in May–June, and DTF hotspots move southward in June–July. Relative humidity and downward shortwave solar radiation are the main drivers of DFAA index variation: Higher relative humidity or weaker downward shortwave solar radiation contributes more positively to the DFAA index, favoring the occurrence of DTF events. In addition, the influence of teleconnection indices on the DFAA index is generally weaker than that of local variables, which mainly affect local anomalies by modulating large-scale atmospheric circulation. These findings help to understand the spatiotemporal characteristics of DFAA events in China and their underlying causes, providing valuable insights for decision-makers to formulate response policies. Significance Statement Based on a 70-yr analysis across China’s river basins, this study reveals that regions prone to shifts from drought to flood have moved southward, while flood-to-drought hotspots have migrated northward. Using a machine learning framework, we demonstrate that these compound extreme events are primarily driven by meteorological factors, particularly relative humidity, and are further intensified by large-scale climate patterns. These spatiotemporal shifts pose severe threats to water security, agriculture, and ecosystems. Our findings provide a crucial scientific basis for developing targeted early warning systems and adaptive watershed management strategies, thereby helping society better anticipate and mitigate the cascading impacts of these complex hydrological extremes in a changing climate.
Abstract The basin runoff consists of three components: surface water runoff, baseflow runoff, and groundwater runoff. Accurately estimating the components of basin runoff can deepen our understanding of hydrological processes and enhance the reliability of water balance analysis. However, estimating runoff composition in large-scale basins with sparse ground observations remains a challenge, especially for groundwater runoff. This study proposed a framework using water balance equations and streamflow routing models to estimate the runoff composition. The hydrometeorological data used include fifth generation European Centre for Medium-Range Weather Forecasts (ECMWF) atmospheric reanalysis (ERA5) precipitation data, Global Land Evaporation Amsterdam Model (GLEAM) data, Gravity Recovery and Climate Experiment (GRACE) data, and Global Land Data Assimilation System (GLDAS) surface runoff data. Moreover, Markov chain Monte Carlo was applied to calibrate the parameter of streamflow routing models. Taking the headstream area of Tarim River basin (TRB) as the study area, the results demonstrated that the baseflow series during the year can be either positive or negative. During 2003–17, the average duration of surface water recharging groundwater is 4.8 month yr −1 , while groundwater recharging surface water is 7.2 months. Nevertheless, the average amount of surface water recharging groundwater is 16.6 × 10 9 m 3 yr −1 , while groundwater recharging surface water is 9.3 × 10 9 m 3 yr −1 . Groundwater runoff is the primary component (88.9%) of the total runoff of study area, with surface runoff (7.9%) and base flow (3.2%) as the second and third components, respectively. Thus, this study provides a feasible approach for estimating runoff composition in large-scale areas with sparse ground observations. Significance Statement Accurately quantifying the components of basin runoff is a prerequisite for the scientific management of water resources in a basin. However, due to the limited availability of ground-based monitoring data, it is challenging to precisely estimate the three components of basin runoff, particularly groundwater runoff. This study proposed a novel method for estimating the components of basin runoff based on remote sensing data, which integrates water balance equations, streamflow routing models, and Bayesian uncertainty analysis. The proposed method makes it feasible to accurately quantify the surface water–groundwater interaction and groundwater runoff in large-scale data-scarce basins.
Satellite and reanalysis rainfall datasets are crucial for meteorological research, hydrological applications, and validating numerical weather prediction models. However, their reliability must be carefully assessed before operational use. The Mumbai mesoscale rain gauge network (MESONET), a high-density network of automatic rain gauges providing minute-level temporal resolution, offers a unique and robust ground-based reference for such validation efforts. This study uses MESONET observations to evaluate the accuracy and consistency of various satellite-based precipitation estimates [multisatellite precipitation estimates (MPEs)] and reanalysis products across daily and subdaily time scales. Our findings indicate that the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement Final (IMERG-F) and fifth generation ECMWF atmospheric reanalysis (ERA5) datasets have the best performance on daily scales, successfully capturing the overall variability of rainfall, albeit with some discrepancies in magnitude. While satellite and reanalysis products can capture diurnal rainfall cycles, they often differ in their representation of rainfall peaks, particularly during heavy precipitation events. Bias-corrected products, such as Global Satellite Mapping of Precipitation (GSMaP)moving vector with Kalman filter (MVK) and IMERG-F, demonstrate improved accuracy but continue to underestimate extreme rainfall events. Although satellite-based products show limitations in detecting light rainfall at shorter time scales, they perform reasonably well for moderate-to-heavy rainfall events. Statistical indices show that the performance of the target datasets considered for evaluation improves significantly after a 12-h accumulation period. Based on the normalized composite score derived from all evaluation metrics, IMERG-Early (IMERG-E) emerges as the most suitable product for real-time applications, such as flood forecasting and hydrological modeling. Indian Monsoon Data Assimilation and Analysis (IMDAA) outperforms Indian National Satellite System (INSAT)-based products [i.e., INSAT-Hydro-Estimator (HE) and INSAT Multispectral Rainfall Algorithm (IMSRA)] in terms of overall accuracy over the study region.
