Search

Precipitation is the most important driver of the hydrological cycle, but it is challenging to estimate it over large scales from satellites and models. Here, we assessed the performance of six global and quasi-global high-resolution precipitation datasets (European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis version 5 (ERA5), Climate Hazards group Infrared Precipitation with Stations version 2.0 (CHIRPS), Multi-Source Weighted-Ensemble Precipitation version 2.80 (MSWEP), TerraClimate (TERRA), Climate Prediction Centre Unified version 1.0 (CPCU), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Cloud Classification System-Climate Data Record (PERSIANN-CCS-CDR, hereafter PERCCDR) for hydrological modelling globally and quasi-globally. We forced the WBMsed global hydrological model with the precipitation datasets to simulate river discharge from 1983 to 2019 and evaluated the predicted discharge against 1825 hydrological stations worldwide, using a range of statistical methods. The results show large differences in the accuracy of discharge predictions when using different precipitation input datasets. Based on evaluation at annual, monthly, and daily timescales, MSWEP followed by ERA5 demonstrated a higher correlation (CC) and Kling–Gupta efficiency (KGE) than other datasets for more than 50 % of the stations, whilst ERA5 was the second-highest-performing dataset, and it showed the highest error and bias for about 20 % of the stations. PERCCDR is the least-well-performing dataset, with a bias of up to 99 % and a normalised root mean square error of up to 247 %. PERCCDR only show a higher KGE and CC than the other products for less than 10 % of the stations. Even though MSWEP provided the highest performance overall, our analysis reveals high spatial variability, meaning that it is important to consider other datasets in areas where MSWEP showed a lower performance. The results of this study provide guidance on the selection of precipitation datasets for modelling river discharge for a basin, region, or climatic zone as there is no single best precipitation dataset globally. Finally, the large discrepancy in the performance of the datasets in different parts of the world highlights the need to improve global precipitation data products.

Global warming is projected to result in changes in streamflow in West Africa with implications for frequent droughts and floods. This study investigates projected shifting in the timing, seasonality and magnitude of mean annual minimum (MAM) and annual maximum flows (AMF) in the Volta River basin (VRB) under climate change, using the method of circular statistics. River flow is simulated with the mesoscale hydrologic model (mHM), forced with bias-corrected climate projection datasets consisting of 43 regional and global climate model combinations under three representative concentration pathways (RCPs). Projected changes indicate that AMF increases between +1 % and +80 % across sub-basins, particularly in the near future (2021–2050), whereas MAM decreases between −19 % and −7 %, mainly from the late century (2071–2100), depending on RCPs. The date of occurrence of AMF is projected to change between −4 and +3 d, while MAM could shift between −4 and +14 d depending on scenarios over the 21st century. Annual high flows denote a strong seasonality with negligible future changes, whereas the seasonality of low flows has a higher variation, with a slight drop in the future.

Streamflow elasticity is the ratio of the expected percentage change in streamflow to a 1 % change in precipitation – a simple approximation of how responsive a river is to precipitation. Typically, streamflow elasticity is estimated for average annual streamflow; however, we propose a new concept in which streamflow elasticity is estimated for multiple percentiles across the full distribution of streamflow. This “elasticity curve” can then be used to develop a more complete depiction of how streamflow responds to climate. Representing elasticity as a curve which reflects the range of responses across the distribution of streamflow within a given time period, instead of as a single-point estimate, provides a novel lens through which we can interpret hydrological behaviour. As an example, we calculate elasticity curves for 805 catchments in the United States and then cluster them according to their shape. This results in three distinct elasticity curve types which characterize the streamflow–precipitation relationship at annual and seasonal timescales. Through this, we demonstrate that elasticity estimated from the central summary of streamflow, e.g. the annual median, does not provide a complete picture of streamflow sensitivity. Further, we show that elasticity curve shape, i.e. the response of different flow percentiles relative to one another in one catchment, can be interpreted separately from between-catchment variation in the average magnitude of streamflow change associated with a 1 % change in precipitation. Finally, we find that available water storage is likely the key control which determines curve shape.

Droughts are among the most complex and devastating natural hazards globally. High-resolution datasets of drought metrics are essential for monitoring and quantifying the severity, duration, frequency, and spatial extent of droughts at regional and particularly local scales. However, current global drought indices are available only at a coarser spatial resolution (>50 km). To fill this gap, we developed four high-resolution (5 km) gridded drought records based on the standardized precipitation evaporation index (SPEI) covering the period 1981–2022. These multi-scale (1–48 months) SPEI indices are computed based on monthly precipitation (P) from the Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS, version 2) and Multi-Source Weighted-Ensemble Precipitation (MSWEP, version 2.8), and potential evapotranspiration (PET) from the Global Land Evaporation Amsterdam Model (GLEAM, version 3.7a) and hourly Potential Evapotranspiration (hPET). We generated four SPEI records based on all possible combinations of P and PET datasets: CHIRPS_GLEAM, CHIRPS_hPET, MSWEP_GLEAM, and MSWEP_hPET. These drought records were evaluated globally and exhibited excellent agreement with observation-based estimates of SPEI, root zone soil moisture, and vegetation health indices. The newly developed high-resolution datasets provide more detailed local information and can be used to assess drought severity for particular periods and regions and to determine global, regional, and local trends, thereby supporting the development of site-specific adaptation measures. These datasets are publicly available at the Centre for Environmental Data Analysis (CEDA; https://doi.org/10.5285/ac43da11867243a1bb414e1637802dec) (Gebrechorkos et al., 2023).

The land surface, hydrological, and groundwater modelling communities all have expertise in simulating the hydrological processes at play in the terrestrial component of the Earth system. However, these communities, and the wider Earth system modelling community, have largely remained distinct with limited collaboration between disciplines, hindering progress in the representation of hydrological processes in the land component of Earth system models (ESMs). In order to address key societal questions regarding the future availability of water resources and the intensity of extreme events such as floods and droughts in a changing climate, these communities must come together and build on the strengths of one another to produce next-generation land system models that are able to adequately simulate the terrestrial water cycle under change. The development of a common modelling infrastructure can contribute to stimulating cross-fertilisation by structuring and standardising the interactions. This paper presents such an infrastructure, a land system framework, which targets an intermediate level of complexity and constrains interfaces between components (and communities) and, in doing so, aims to facilitate an easier pipeline between the development of (sub-)community models and their integration, both for standalone use and for use in ESMs. This paper first outlines the conceptual design and technical capabilities of the framework; thereafter, its usage and useful characteristics are demonstrated through case studies. The main innovations presented here are (1) the interfacing constraints themselves; (2) the implementation in Python (the Unified Framework for Hydrology, unifhy); and (3) the demonstration of standalone use cases using the framework. The existing framework does not yet meet all our goals, in particular, of directly supporting integration into larger ESMs, so we conclude with the remaining limitations of the current framework and necessary future developments.

Wetlands are the largest natural source of methane. The ability to model the emissions of methane from natural wetlands accurately is critical to our understanding of the global methane budget and how it may change under future climate scenarios. The simulation of wetland methane emissions involves a complicated system of meteorological drivers coupled to hydrological and biogeochemical processes. The Joint UK Land Environment Simulator (JULES) is a process-based land surface model that underpins the UK Earth System Model (UKESM) and is capable of generating estimates of wetland methane emissions. In this study, we use GOSAT satellite observations of atmospheric methane along with the TOMCAT global 3-D chemistry transport model to evaluate the performance of JULES in reproducing the seasonal cycle of methane over a wide range of tropical wetlands. By using an ensemble of JULES simulations with differing input data and process configurations, we investigate the relative importance of the meteorological driving data, the vegetation, the temperature dependency of wetland methane production and the wetland extent. We find that JULES typically performs well in replicating the observed methane seasonal cycle. We calculate correlation coefficients to the observed seasonal cycle of between 0.58 and 0.88 for most regions; however, the seasonal cycle amplitude is typically underestimated (by between 1.8 and 19.5 ppb). This level of performance is comparable to that typically provided by state-of-the-art data-driven wetland CH4 emission inventories. The meteorological driving data are found to be the most significant factor in determining the ensemble performance, with temperature dependency and vegetation having moderate effects. We find that neither wetland extent configuration outperforms the other, but this does lead to poor performance in some regions. We focus in detail on three African wetland regions (Sudd, Southern Africa and Congo) where we find the performance of JULES to be poor and explore the reasons for this in detail. We find that neither wetland extent configuration used is sufficient in representing the wetland distribution in these regions (underestimating the wetland seasonal cycle amplitude by 11.1, 19.5 and 10.1 ppb respectively, with correlation coefficients of 0.23, 0.01 and 0.31). We employ the Catchment-based Macro-scale Floodplain (CaMa-Flood) model to explicitly represent river and floodplain water dynamics and find that these JULES-CaMa-Flood simulations are capable of providing a wetland extent that is more consistent with observations in this regions, highlighting this as an important area for future model development.

Abstract Geomorphologists have long debated the relative importance of disturbance magnitude, duration, and frequency in shaping landscapes. For river‐channel adjustment by floods, some argue that the cumulative hydrograph, rather than magnitude or duration, matters most. However, studies of flood‐induced river‐channel change often draw upon small data sets. Here, we combine Sentinel‐2 imagery with flow data from laterally active rivers to address this question using a larger data set. We apply automated algorithms in Google Earth Engine to map rivers and detect their lateral shifting; we generate a large data set to quantify planform erosion during 175 floods at 34 selected sites. Erosion during these floods is best explained by their duration and then their cumulative hydrograph. We use a random forest regression model to predict flood‐induced erosion, with potential applications for hazard management. Ultimately, better global data on sediment supply and caliber would help us to understand flood‐driven change to river planforms.

Wetlands play a key role in hydrological and biogeochemical cycles and provide multiple ecosystem services to society. However, reliable data on the extent of global inundated areas and the magnitude of their contribution to local hydrological dynamics remain surprisingly uncertain. Global hydrological models and land surface models (LSMs) include only the most major inundation sources and mechanisms; therefore, quantifying the uncertainties in available data sources remains a challenge. We address these problems by taking a leading global data product on inundation extents (Global Inundation Extent from Multi-Satellites, GIEMS) and matching against predictions from a global hydrodynamic model (Catchment-based Macro-scale Floodplain – CaMa-Flood) driven by runoff data generated by a land surface model (Joint UK Land and Environment Simulator, JULES). The ability of the model to reproduce patterns and dynamics shown by the observational product is assessed in a number of case studies across the tropics, which show that it performs well in large wetland regions, with a good match between corresponding seasonal cycles. At a finer spatial scale, we found that water inputs (e.g. groundwater inflow to wetland) became underestimated in comparison to water outputs (e.g. infiltration and evaporation from wetland) in some wetlands (e.g. Sudd, Tonlé Sap), and the opposite occurred in others (e.g. Okavango) in our model predictions. We also found evidence for an underestimation of low levels of inundation in our satellite-based inundation data (approx. 10 % of total inundation may not be recorded). Additionally, some wetlands display a clear spatial displacement between observed and simulated inundation as a result of overestimation or underestimation of overbank flooding upstream. This study provides timely information on inherent biases in inundation prediction and observation that can contribute to our current ability to make critical predictions of inundation events at both regional and global levels.

Neural networks have been shown to be extremely effective rainfall-runoff models, where the river discharge is predicted from meteorological inputs. However, the question remains: what have these models learned? Is it possible to extract information about the learned relationships that map inputs to outputs, and do these mappings represent known hydrological concepts? Small-scale experiments have demonstrated that the internal states of long short-term memory networks (LSTMs), a particular neural network architecture predisposed to hydrological modelling, can be interpreted. By extracting the tensors which represent the learned translation from inputs (precipitation, temperature, and potential evapotranspiration) to outputs (discharge), this research seeks to understand what information the LSTM captures about the hydrological system. We assess the hypothesis that the LSTM replicates real-world processes and that we can extract information about these processes from the internal states of the LSTM. We examine the cell-state vector, which represents the memory of the LSTM, and explore the ways in which the LSTM learns to reproduce stores of water, such as soil moisture and snow cover. We use a simple regression approach to map the LSTM state vector to our target stores (soil moisture and snow). Good correlations (R2>0.8) between the probe outputs and the target variables of interest provide evidence that the LSTM contains information that reflects known hydrological processes comparable with the concept of variable-capacity soil moisture stores. The implications of this study are threefold: (1) LSTMs reproduce known hydrological processes. (2) While conceptual models have theoretical assumptions embedded in the model a priori, the LSTM derives these from the data. These learned representations are interpretable by scientists. (3) LSTMs can be used to gain an estimate of intermediate stores of water such as soil moisture. While machine learning interpretability is still a nascent field and our approach reflects a simple technique for exploring what the model has learned, the results are robust to different initial conditions and to a variety of benchmarking experiments. We therefore argue that deep learning approaches can be used to advance our scientific goals as well as our predictive goals.

A comprehensive evaluation of the impacts of climate change on water resources of the West Africa Volta River basin is conducted in this study, as the region is expected to be hardest hit by global warming. A large ensemble of 12 general circulation models (GCMs) from the fifth Coupled Model Intercomparison Project (CMIP5) that are dynamically downscaled by five regional climate models (RCMs) from the Coordinated Regional-climate Downscaling Experiment (CORDEX)-Africa is used. In total, 43 RCM–GCM combinations are considered under three representative concentration pathways (RCP2.6, RCP4.5, and RCP8.5). The reliability of each of the climate datasets is first evaluated with satellite and reanalysis reference datasets. Subsequently, the Rank Resampling for Distributions and Dependences (R2D2) multivariate bias correction method is applied to the climate datasets. The bias-corrected climate projections are then used as input to the mesoscale Hydrologic Model (mHM) for hydrological projections over the 21st century (1991–2100). Results reveal contrasting dynamics in the seasonality of rainfall, depending on the selected greenhouse gas emission scenarios and the future projection periods. Although air temperature and potential evaporation increase under all RCPs, an increase in the magnitude of all hydrological variables (actual evaporation, total runoff, groundwater recharge, soil moisture, and terrestrial water storage) is only projected under RCP8.5. High- and low-flow analysis suggests an increased flood risk under RCP8.5, particularly in the Black Volta, while hydrological droughts would be recurrent under RCP2.6 and RCP4.5, particularly in the White Volta. The evolutions of streamflow indicate a future delay in the date of occurrence of low flows up to 11 d under RCP8.5, while high flows could occur 6 d earlier (RCP2.6) or 5 d later (RCP8.5), as compared to the historical period. Disparities are observed in the spatial patterns of hydroclimatic variables across climatic zones, with higher warming in the Sahelian zone. Therefore, climate change would have severe implications for future water availability with concerns for rain-fed agriculture, thereby weakening the water–energy–food security nexus and amplifying the vulnerability of the local population. The variability between climate models highlights uncertainties in the projections and indicates a need to better represent complex climate features in regional models. These findings could serve as a guideline for both the scientific community to improve climate change projections and for decision-makers to elaborate adaptation and mitigation strategies to cope with the consequences of climate change and strengthen regional socioeconomic development.

Long short-term memory (LSTM) models are recurrent neural networks from the field of deep learning (DL) which have shown promise for time series modelling, especially in conditions when data are abundant. Previous studies have demonstrated the applicability of LSTM-based models for rainfall–runoff modelling; however, LSTMs have not been tested on catchments in Great Britain (GB). Moreover, opportunities exist to use spatial and seasonal patterns in model performances to improve our understanding of hydrological processes and to examine the advantages and disadvantages of LSTM-based models for hydrological simulation. By training two LSTM architectures across a large sample of 669 catchments in GB, we demonstrate that the LSTM and the Entity Aware LSTM (EA LSTM) models simulate discharge with median Nash–Sutcliffe efficiency (NSE) scores of 0.88 and 0.86 respectively. We find that the LSTM-based models outperform a suite of benchmark conceptual models, suggesting an opportunity to use additional data to refine conceptual models. In summary, the LSTM-based models show the largest performance improvements in the north-east of Scotland and in south-east of England. The south-east of England remained difficult to model, however, in part due to the inability of the LSTMs configured in this study to learn groundwater processes, human abstractions and complex percolation properties from the hydro-meteorological variables typically employed for hydrological modelling.

Soil moisture predictions from land surface models are important in hydrological, ecological, and meteorological applications. In recent years, the availability of wide-area soil moisture measurements has increased, but few studies have combined model-based soil moisture predictions with in situ observations beyond the point scale. Here we show that we can markedly improve soil moisture estimates from the Joint UK Land Environment Simulator (JULES) land surface model using field-scale observations and data assimilation techniques. Rather than directly updating soil moisture estimates towards observed values, we optimize constants in the underlying pedotransfer functions, which relate soil texture to JULES soil physics parameters. In this way, we generate a single set of newly calibrated pedotransfer functions based on observations from a number of UK sites with different soil textures. We demonstrate that calibrating a pedotransfer function in this way improves the soil moisture predictions of a land surface model at 16 UK sites, leading to the potential for better flood, drought, and climate projections.

Hydro-meteorological extremes such as drought and heavy precipitation can have large impacts on society and the economy. With potentially increasing risks associated with such events due to climate change, properly assessing the associated impacts and uncertainties is critical for adequate adaptation. However, the application of risk-based approaches often requires large sets of extreme events, which are not commonly available. Here, we present such a large set of hydro-meteorological time series for recent past and future conditions for the United Kingdom based on weather@home 2, a modelling framework consisting of a global climate model (GCM) driven by observed or projected sea surface temperature (SST) and sea ice which is downscaled to 25 km over the European domain by a regional climate model (RCM). Sets of 100 time series are generated for each of (i) a historical baseline (1900–2006), (ii) five near-future scenarios (2020–2049) and (iii) five far-future scenarios (2070–2099). The five scenarios in each future time slice all follow the Representative Concentration Pathway 8.5 (RCP8.5) and sample the range of sea surface temperature and sea ice changes from CMIP5 (Coupled Model Intercomparison Project Phase 5) models. Validation of the historical baseline highlights good performance for temperature and potential evaporation, but substantial seasonal biases in mean precipitation, which are corrected using a linear approach. For extremes in low precipitation over a long accumulation period ( > 3 months) and shorter-duration high precipitation (1–30 days), the time series generally represents past statistics well. Future projections show small precipitation increases in winter but large decreases in summer on average, leading to an overall drying, consistently with the most recent UK Climate Projections (UKCP09) but larger in magnitude than the latter. Both drought and high-precipitation events are projected to increase in frequency and intensity in most regions, highlighting the need for appropriate adaptation measures. Overall, the presented dataset is a useful tool for assessing the risk associated with drought and more generally with hydro-meteorological extremes in the UK.

It is hypothesized that more accurate prediction and warning of natural hazards, such as of the impacts of severe weather mediated through various components of the environment, require a more integrated Earth System approach to forecasting. This hypothesis can be explored using regional coupled prediction systems, in which the known interactions and feedbacks between different physical and biogeochemical components of the environment across sky, sea and land can be simulated. Such systems are becoming increasingly common research tools. This paper describes the development of the UKC2 regional coupled research system, which has been delivered under the UK Environmental Prediction Prototype project. This provides the first implementation of an atmosphere–land–ocean–wave modelling system focussed on the United Kingdom and surrounding seas at km-scale resolution. The UKC2 coupled system incorporates models of the atmosphere (Met Office Unified Model), land surface with river routing (JULES), shelf-sea ocean (NEMO) and ocean waves (WAVEWATCH III). These components are coupled, via OASIS3-MCT libraries, at unprecedentedly high resolution across the UK within a north-western European regional domain. A research framework has been established to explore the representation of feedback processes in coupled and uncoupled modes, providing a new research tool for UK environmental science. This paper documents the technical design and implementation of UKC2, along with the associated evaluation framework. An analysis of new results comparing the output of the coupled UKC2 system with relevant forced control simulations for six contrasting case studies of 5-day duration is presented. Results demonstrate that performance can be achieved with the UKC2 system that is at least comparable to its component control simulations. For some cases, improvements in air temperature, sea surface temperature, wind speed, significant wave height and mean wave period highlight the potential benefits of coupling between environmental model components. Results also illustrate that the coupling itself is not sufficient to address all known model issues. Priorities for future development of the UK Environmental Prediction framework and component systems are discussed.

Modelling land surface water flow is of critical importance for simulating land surface fluxes, predicting runoff and water table dynamics and for many other applications of Land Surface Models. Many approaches are based on the popular hydrology model TOPMODEL (TOPography-based hydrological MODEL), and the most important parameter of this model is the well-known topographic index. Here we present new, high-resolution parameter maps of the topographic index for all ice-free land pixels calculated from hydrologically conditioned HydroSHEDS (Hydrological data and maps based on SHuttle Elevation Derivatives at multiple Scales) data using the GA2 algorithm (GRIDATB 2). At 15 arcsec resolution, these layers are 4 times finer than the resolution of the previously best-available topographic index layers, the compound topographic index of HYDRO1k (CTI). For the largest river catchments occurring on each continent we found that, in comparison with CTI our revised values were up to 20% lower in, e.g. the Amazon. We found the highest catchment means were for the Murray–Darling and Nelson–Saskatchewan rather than for the Amazon and St. Lawrence as found from the CTI. For the majority of large catchments, however, the spread of our new GA2 index values is very similar to those of CTI, yet with more spatial variability apparent at fine scale. We believe these new index layers represent greatly improved global-scale topographic index values and hope that they will be widely used in land surface modelling applications in the future.