Land Surface Hydrological Models Research Articles

Learning Distributed Parameters of Land Surface Hydrologic Models Using a Generative Adversarial Network

AbstractLand surface hydrologic models adeptly capture crucial terrestrial processes with a high level of spatial detail. Typically, these models incorporate numerous uncertain, spatially varying parameters, the specification of which can profoundly impact the simulation capabilities. There is a longstanding tradition wherein parameter calibration has served as the conventional procedure to enhance model performance. However, calibrating distributed land surface hydrologic models presents a great challenge, often resulting in uneven spatial performance due to the compression of information inherent in model outputs and observations into a single‐value objective function. To address this problem, we propose a novel Generative Adversarial Network‐based Parameter Optimization (GAN‐PO) method. By leveraging a deep neural network to discern model spatial biases, we train a generative network to produce spatially coherent parameter fields, minimizing distinctions between simulations and observations. By leveraging neural network‐based surrogate models to make the physical model differentiable, we employ GAN‐PO to calibrate the Variable Infiltration Capacity (VIC) model against evapotranspiration (ET) over China's Huaihe basin. The results show that GAN‐PO can diminish errors in simulated ET derived from default parameters across nearly all grid cells within the study region, surpassing the conventional calibration approach based on the parameter regionalization technique. Ablation analysis indicates that relying solely on the traditional loss could lead to deteriorated model performance, underscoring the crucial role of the discriminator. Notably, due to the discriminator's explicit identification of model spatial biases, GAN‐PO excels in maintaining spatial consistency, outperforming the state‐of‐the‐art differentiable parameter learning (dPL) method in terms of model spatial performance.

Unraveling phenological and stomatal responses to flash drought and implications for water and carbon budgets

Abstract. In recent years, extreme droughts in the United States have increased in frequency and severity, underlining a need to improve our understanding of vegetation resilience and adaptation. Flash droughts are extreme events marked by the rapid dry down of soils due to lack of precipitation, high temperatures, and dry air. These events are also associated with reduced preparation, response, and management time windows before and during drought, exacerbating their detrimental impacts on people and food systems. Improvements in actionable information for flash drought management are informed by atmospheric and land surface processes, including responses and feedbacks from vegetation. Phenologic state, or growth stage, is an important metric for modeling how vegetation modulates land–atmosphere interactions. Reduced stomatal conductance during drought leads to cascading effects on carbon and water fluxes. We investigate how uncertainty in vegetation phenology and stomatal regulation propagates through vegetation responses during drought and non-drought periods by coupling a land surface hydrology model to a predictive phenology model. We assess the role of vegetation in the partitioning of carbon, water, and energy fluxes during flash drought and carry out a comparison against drought and non-drought periods. We selected study sites in Kansas, USA, that were impacted by the flash drought of 2012 and that have AmeriFlux eddy covariance towers which provide ground observations to compare against model estimates. Results show that the compounding effects of reduced precipitation and high vapor pressure deficit (VPD) on vegetation distinguish flash drought from other drought and non-drought periods. High VPD during flash drought shuts down modeled stomatal conductance, resulting in rates of evapotranspiration (ET), gross primary productivity (GPP), and water use efficiency (WUE) that fall below those of average drought conditions. Model estimates of GPP and ET during flash drought decrease to rates similar to what is observed during the winter, indicating that plant function during drought periods is similar to that of dormant months. These results have implications for improving predictions of drought impacts on vegetation.

Exploring the Drivers for Changes in Lake Area in a Typical Arid Region during Past Decades

Lakes are important surface water bodies, and ongoing climate change is a growing threat to the hydrological cycle and water resource availability of lakes in arid regions. Accurately estimating different drivers’ contributions to lake water volume can enhance our understanding of lake variations in arid regions. In this study, we combined the land surface model and hydrological model, as well as statistical methods, to analyze the spatiotemporal heterogeneity of lake area changes and the factors affecting these changes during the past decades in Bosten Lake, Ulungur Lake, Ebinur Lake, and Sayram Lake, which are located in a typical dry region in China. The study revealed that the average amounts of river inflow, TWVF, lake ice sublimation, lake surface precipitation, and river outflow in the four lakes were 17.41 × 108 m3 yr−1, 6.60 × 108 m3 yr−1, 0.41 × 108 m3 yr−1, 0.98 × 108 m3 yr−1, and 9.12 × 108 m3 yr−1, respectively. We found that river inflow is the dominant factor affecting changes in open lake areas, while lake surface precipitation is the main factor affecting changes in closed lake areas. Our findings suggest that the main factors dominating the variability of lake water volume differ in different phases and lake types.

Cycles‐L: A Coupled, 3‐D, Land Surface, Hydrologic, and Agroecosystem Landscape Model

AbstractManaging landscapes to increase agricultural productivity and environmental stewardship can be informed by spatially‐distributed models that operate at spatial and temporal scales that are intervention‐relevant. This paper presents Cycles‐L, a landscape‐scale agroecosystem and hydrologic modeling system, using as a test case a watershed in Pennsylvania. Cycles‐L emerges from melding the landscape and hydrology structure of Flux‐PIHM, a 3‐D land surface hydrologic model, with the agroecosystem processes in the Cycles model. Consequently, Cycles‐L can simulate processes affected by topography, soil heterogeneity, and management practices, owing to its physically‐based hydrology that can simulate horizontal and vertical transport of solutes with water. The model was tested at a 730‐ha experimental watershed within the Mahantango Creek watershed. Cycles‐L simulated well stream water and mineral nitrogen discharge (Nash‐Sutcliffe coefficient 0.55 and 0.60, respectively) and grain yield (root mean square error 1.2 Mg ha−1). Cycles‐L outputs are as good or better than those obtained with the uncoupled Flux‐PIHM (water discharge) and Cycles (grain yield) models. Modeled spatial patterns of nitrogen fluxes like denitrification illustrate the combined control of crop management and topography. For example, denitrification is almost twice as high when simulated with Cycles‐L than when simulated with Cycles 1‐D. Due to its spatial and temporal resolution, Cycles‐L fills a gap in the availability of models that operate at a scale relevant to evaluate interventions in the landscape. Cycles‐L can become a central component in tools for climate change scenario analysis, precision agriculture, precision conservation, and artificial intelligence‐based decision support systems.

The Evaluation of Snow Depth Simulated by Different Land Surface Models in China Based on Station Observations

Snow plays an important role in catastrophic weather, climate change, and water recycling. In order to analyze the ability of different land surface models to simulate snow depth in China, we used atmospheric forcing data from the China Meteorological Administration (CMA) Land Data Assimilation System (CLDAS) to drive the CLM3.5 (the Community Land Model version 3.5), Noah (NCEP, OSU, Air Force and Office of Hydrology Land Surface Model), and Noah-MP (the community Noah land surface model with multi-parameterization options) land surface models. We also used 2380 daily snow-depth site observations of CMA to analyze the simulation effects of different models on the snow depth in China and different regions during the periods of snow accumulation and snowmelt from 2015 to 2019. The results show that CLM3.5, Noah, and Noah-MP can simulate the spatial distribution of the snow depth in China, but there are some differences between the models. In particular, the snow depth and snow cover simulated by CLM3.5 are lower than those simulated by Noah and Noah-MP in Northwest China and the Tibetan Plateau. From the overall quantitative assessment results for China, the snow depth simulated by CLM3.5 is underestimated, while that simulated by Noah is overestimated. Noah-MP has the best overall performance; for example, the biases of the three models during the snow-accumulation periods are −0.22 cm, 0.27 cm, and 0.15 cm, respectively. Furthermore, the three models perform differently in the three snowpack regions of Northeast China, Northwest China, and the Tibetan Plateau; Noah-MP has the best snow-depth performance in Northeast China, while CLM3.5 has the best snow-depth performance in the Tibetan Plateau region. Noah-MP performs best in the snow-accumulation period, and Noah performs best in the snowmelt period for Northwest China. In conclusion, no single model can perform optimally for snow simulations in different regions of China and at different times of the year, and the multi-model integration of snow may be an effective way to obtain high-quality snow simulation results. So this study provides some scientific references for the spatiotemporal evolution of snow in the context of climate change, monitoring and analysis of snow, the study of land surface models for snow, and the sustainable development and utilization of snow resources in China and other regions.

A surrogate modeling method for distributed land surface hydrological models based on deep learning

Surrogate modelling methods have been used to estimate spatially distributed parameters of land surface hydrological models due to their computational efficiency. However, traditional surrogate techniques have trouble in modelling high-dimensional input–output maps. In addition, their applications to a distributed model over large spatial and temporal domains are very unlikely to produce consistent spatial pattern and dynamic relationship of model input and output, leading to unbalanced model performance across different spatiotemporal domains. To address these issues, we employ a deep autoregressive neural network to construct an accurate and reliable surrogate system of distributed dynamic model outputs. The network uses a deep convolutional encoder-decoder architecture to take full advantage of the deep convolutional layers in high-dimensional image processing and spatial feature extraction. Moreover, the autoregressive strategy makes the network good at handling dynamic relationship between the time-varying model input and output. We apply this deep learning (DL) network to building a surrogate model of the Variable Infiltration Capacity (VIC) model simulated soil moisture over the Huaihe river basin of China. The results show that the adopted deep autoregressive neural network can provide an accurate surrogate model of the high-dimensional dynamic relationship between 15 input fields and 1 output field, each covering 408 grids, over multiple years. This surrogate model significantly outperforms the popular long short-term memory (LSTM) model, achieving an average mean squared error (MSE) of 0.26 and an average R2 of 0.76 using the test datasets. The surrogate's MSE and R2 performance represents an average improvement of 43% and 33%, respectively, compared to the LSTM model. In addition, it shows better spatial performance than the LSTM model, demonstrating its superior performance in approximating spatial patterns of VIC simulated soil moisture. The accuracy of this surrogate method and its ability to handle high-dimensional data is promising for advancing parameter uncertainty quantification of distributed land surface hydrological models, which usually have high dimensionality due to the large number of variables involved, including distributed parameters, input forcing variables, and output variables.

A Transformer-Based Framework for Parameter Learning of a Land Surface Hydrological Process Model

The effective representation of land surface hydrological models strongly relies on spatially varying parameters that require calibration. Well-calibrated physical models can effectively propagate observed information to unobserved variables, but traditional calibration methods often result in nonunique solutions. In this paper, we propose a hydrological parameter calibration training framework consisting of a transformer-based parameter learning model (ParaFormer) and a surrogate model based on LSTM. On the one hand, ParaFormer utilizes self-attention mechanisms to learn a global mapping from observed data to the parameters to be calibrated, which captures spatial correlations. On the other hand, the surrogate model takes the calibrated parameters as inputs and simulates the observable variables, such as soil moisture, overcoming the challenges of directly combining complex hydrological models with a deep learning (DL) platform in a hybrid training scheme. Using the variable infiltration capacity model as the reference, we test the performance of ParaFormer on datasets of different resolutions. The results demonstrate that, in predicting soil moisture and transferring calibrated parameters in the task of evapotranspiration prediction, ParaFormer learns more effective and robust parameter mapping patterns compared to traditional and state-of-the-art DL-based parameter calibration methods.

The Impact of Inter-Basin Water Transfer Schemes on Hydropower Generation in the Upper Reaches of the Yangtze River during Extreme Drought Years

The Yangtze River Basin experiences frequent extreme heatwaves and prolonged droughts, resulting in a tight supply demand balance of electricity and negatively impacting socioeconomic production. Meanwhile, ongoing inter-basin water diversion projects are planned that will divert approximately 25.263 billion cubic meters of water from the Yangtze River Basin annually, which may further affect the power supply in the region. In this study, the CLHMS-LSTM model, a land-surface hydrological model coupled with a long short-term memory (LSTM)-based reservoir operation simulation model, is used to investigate the impact of water diversions on the power generation of the Yangtze River mainstream reservoirs under extreme drought conditions. Two different water diversion schemes are adopted in this study, namely the minimum water deficit scheme (Scheme 1) and minimum construction cost scheme (Scheme 2). The results show that the land surface–hydrological model was able to well characterize the hydrological characteristics of the Yangtze River mainstem, with a daily scale determination coefficient greater than 0.85. The LSTM reservoir operation simulation model was able to simulate the reservoir releases well, with the determination coefficient greater than 0.93. The operation of the water diversion projects will result in a reduction in the power generation of the Yangtze River mainstem by 14.97 billion kilowatt-hours. As compared to the minimum construction cost scheme (Scheme 2), the minimum water deficit scheme (Scheme 1) reduces the loss of power generation by 1.38 billion kilowatt-hours. The research results provide new ideas for the decision-making process for the inter-basin water diversion project and the formulation of water diversion plans, which has implications for ensuring the security of the power supply in the water diversion area.

Hydrological response to climate change and human activities in the Three-River Source Region

Abstract. The Three-River Source Region (TRSR), which is known as “China's Water Tower” and affects the water resources security of 700 million people living downstream, has experienced significant hydrological changes in the past few decades. In this work, we used an extended variable infiltration capacity (VIC) land surface hydrologic model (VIC-Glacier) coupled with the degree-day factor algorithm to simulate the runoff change in the TRSR during 1984–2018. VIC-Glacier performed well in the TRSR, with Nash–Sutcliffe efficiency (NSE) above 0.68, but it was sensitive to the quality of the limited ground-based precipitation. This was especially marked in the source region of the Yangtze River: when we used Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks – Climate Data Record (PERSIANN-CDR), which has better spatial details, instead of ground-based precipitation, the NSE of Tuotuohe station increased from 0.31 to 0.86. Using the well-established VIC-Glacier model, we studied the contribution of each runoff component (rainfall, snowmelt, and glacier runoff) to the total runoff and the causes of changes in runoff. The results indicate that rainfall runoff contributed over 80 % of the total runoff, while snowmelt runoff and glacier runoff both contributed less than 10 % in 1984–2018. Climate change was the main reason for the increase in runoff in the TRSR after 2004, accounting for 75 %–89 %, except in the catchment monitored by Xialaxiu station. Among climate change factors, precipitation had the greatest impact on runoff. Finally, through a series of hypothetical climate change scenario experiments, we found that a future simultaneous increase in precipitation and temperature would increase the total runoff, rainfall runoff, and glacier runoff. The snowmelt runoff might remain unchanged because the increased precipitation, even with seasonal fluctuations, was basically completely compensated for by the decreased solid-to-liquid precipitation ratio. These findings improve our understanding of hydrological processes and provide insights for policy-makers on how to optimally allocate water resources and manage the TRSR in response to global climate change.

Assessing the Future Water and Energy Security of a Regulated River Basin with a Coupled Land Surface and Hydrologic Model

To address the water-related issues faced by humans, the planning and construction of dams, water diversion projects, and other water infrastructures have been continuously adopted by decision makers worldwide. This is especially the case for the Yalong River Basin (YRB) in China, which is expected to be one of the most regulated rivers due to reservoir construction and the planned South-to-North Water Diversion project. To understand the potential impact of these water infrastructures on the water resources and hydropower production of the basin and downstream areas, we employ a land surface–hydrologic model with explicit representations of dam operation and water diversions in order to quantify the impact of reservoir operation and water diversion on the future water and energy security of the YRB. In particular, a conceptual reservoir operation scheme and a hydropower-optimized reservoir operation scheme are employed to predict the future release, storage and hydropower generation of the YRB, respectively. Results indicate that reservoirs can have noticeable, cumulative effects in enhancing the water security by reducing the wet season streamflow by 19% and increasing the dry season streamflow by 66%. The water diversion can result in an overall decrease in the streamflow, while the downstream reservoirs are expected to fully mitigate the decline in the dry season streamflow. The hydropower production is likely to decrease by 16% and 10% with conventional and optimized operation schemes, respectively, which suggests that the adaptation of operation rules alone cannot reverse the decline in the electricity production. Our findings can provide implications for sustainable water resource management.

Prediction of water resources change trend in the Three Gorges Reservoir Area under future climate change

In this study, the water intake module and the confluence module were added to the Community Land Model version 4.5 (CLM 4.5), forming the Community Land Model-Dualistic Water Cycle (CLM-DWC) model (land surface hydrological model). It solved the problem that the influence of human activities was not considered in the existing hydrological coupling simulation process, which so as to the inaccuracy of the simulation process. The CLM-DWC model was driven by the average results of five Global Climate Models (GCMs) ensemble to form the “atmosphere-land-hydrology” fully coupled model to predict the temporal and spatial variation trends of water resources in the Three Gorges Reservoir Area (TGRA) from 2021 to 2050 under Representative Concentration Pathway (RCP) 2.6, RCP4.5 and RCP8.5. In the future scenarios, in terms of spatial distribution, the runoff, surface water and groundwater all indicated the high at the belly of the reservoir and low at the head and tail of the reservoir, which were in good consistency with the baseline period. In terms of intra-annual distribution and inter-annual variability, runoff depth increases by 20.8 mm at RCP4.5 compared with the baseline period, and it tended to decrease under RCP2.6 and RCP8.5. At the same time, under the three scenarios, the average increase of surface water was 3.44 %, groundwater decreases by an average of 11.72 %. Accurately predicting the change trend of water resources in the TGRA under climate change has great significance for future water resources management and planning in the reservoir area.

Corrected ERA5 Precipitation by Machine Learning Significantly Improved Flow Simulations for the Third Pole Basins

Abstract Precipitation is one of the most important atmospheric inputs to hydrological models. However, existing precipitation datasets for the Third Pole (TP) basins show large discrepancies in precipitation magnitudes and spatiotemporal patterns, which poses a great challenge to hydrological simulations in the TP basins. In this study, a gridded (10 km × 10 km) daily precipitation dataset is constructed through a random-forest-based machine learning algorithm (RF algorithm) correction of the ERA5 precipitation estimates based on 940 gauges in 11 upper basins of TP for 1951–2020. The dataset is evaluated by gauge observations at point scale and is inversely evaluated by the Variable Infiltration Capacity (VIC) hydrological model linked with a glacier melt algorithm (VIC-Glacier). The corrected ERA5 (ERA5_cor) agrees well with gauge observations after eliminating the severe overestimation in the original ERA5 precipitation. The corrections greatly reduce the original ERA5 precipitation estimates by 10%–50% in 11 basins of the TP and present more details on precipitation spatial variability. The inverse hydrological model evaluation demonstrates the accuracy and rationality, and we provide an updated estimate of runoff components contribution to total runoff in seven upper basins in the TP based on the VIC-Glacier model simulations with the ERA5_cor precipitation. This study provides good precipitation estimates with high spatiotemporal resolution for 11 upper basins in the TP, which are expected to facilitate the hydrological modeling and prediction studies in this high mountainous region. Significance Statement The Third Pole (TP) is the source of water to the people living in the areas downstream. Precipitation is the key driver of the terrestrial hydrological cycle and the most important atmospheric input to land surface hydrological models. However, none of the current precipitation data are equally good for all the TP basins because of high variabilities in their magnitudes and spatiotemporal patterns, posing a great challenge to the hydrological simulation. Therefore, in this study, a gridded daily precipitation dataset (10 km × 10 km) is reconstructed through a random-forest-based machine learning algorithm correction of ERA5 precipitation estimates based on 940 gauges in 11 TP basins for 1951–2020. The data eliminate the severe overestimation of original ERA5 precipitation estimates and present more reasonable spatial variability, and also exhibit a high potential for hydrological application in the TP basins. This study provides long-term precipitation data for climate and hydrological studies and a reference for deriving precipitation in high mountainous regions with complex terrain and limited observations.

Microwave radiometry experiment for snow in Altay, China: time series of in situ data for electromagnetic and physical features of snowpack

Abstract. In this paper, we present a comprehensive experiment, namely, an Integrated Microwave Radiometry Campaign for snow (IMCS), in Xinjiang, China, during the snow season of 2015–2016. The campaign hosted a dual-polarized microwave radiometer operating at L, K, and Ka bands to provide minutely passive microwave observations of snow cover at a fixed site, along with daily manual snow pit observations of snow physical parameters, automatic observations of 10 min four-component radiation and layered snow temperatures, and meteorological observations of hourly weather data and soil data. To the best of our knowledge, our dataset is unique in providing continuous daily snow pit data and coincident microwave brightness temperatures, radiation and meteorological data, at a fixed site over a full season, which can be straightforwardly used for evaluation and development of microwave radiative transfer models and snow process models, along with land surface process and hydrology models. The consolidated data are available at (https://doi.org/10.11888/Snow.tpdc.270886) (Dai, 2020).

Ensemble streamflow forecasting over a cascade reservoir catchment with integrated hydrometeorological modeling and machine learning

Abstract. A popular way to forecast streamflow is to use bias-corrected meteorological forecasts to drive a calibrated hydrological model, but these hydrometeorological approaches suffer from deficiencies over small catchments due to uncertainty in meteorological forecasts and errors from hydrological models, especially over catchments that are regulated by dams and reservoirs. For a cascade reservoir catchment, the discharge from the upstream reservoir contributes to an important part of the streamflow over the downstream areas, which makes it tremendously hard to explore the added value of meteorological forecasts. Here, we integrate meteorological forecasts, land surface hydrological model simulations and machine learning to forecast hourly streamflow over the Yantan catchment, where the streamflow is influenced by both the upstream reservoir water release and the rainfall–runoff processes within the catchment. Evaluation of the hourly streamflow hindcasts during the rainy seasons of 2013–2017 shows that the hydrometeorological ensemble forecast approach reduces probabilistic and deterministic forecast errors by 6 % compared with the traditional ensemble streamflow prediction (ESP) approach during the first 7 d. The deterministic forecast error can be further reduced by 6 % in the first 72 h when combining the hydrometeorological forecasts with the long short-term memory (LSTM) deep learning method. However, the forecast skill for LSTM using only historical observations drops sharply after the first 24 h. This study implies the potential of improving flood forecasts over a cascade reservoir catchment by integrating meteorological forecasts, hydrological modeling and machine learning.

Loess Plateau evapotranspiration intensified by land surface radiative forcing associated with ecological restoration

Evapotranspiration (ET) is a key variable coupled with water and energy availability and vegetation coverage. Since the year around 2000, several ecological restoration programs have been enacted within the Loess Plateau, China. The restoration programs have promoted vegetation greening and air cleaning, which has changed land surface albedo and solar radiation and consequently altered the land surface radiative forcing. Evidence has showed that ET is impacted by climate change and land use/cover change, but little is known about the effect of the radiative forcing on ET. This study reports the contribution of the surface radiative forcing during the growing season associated with ecological restoration to changes in potential ET (PET) and actual ET (AET) in the Loess Plateau. A differential attribution method based on the Penman-Monteith equation is proposed to isolate the contribution of radiative forcing. Accurate estimates of PET and AET are obtained from a sophisticated land surface hydrological model driven by long-term dynamic albedo, solar radiation, vegetation index, and meteorological forcings. We find that during 2003-2018, the surface albedo significantly decreased with increased solar radiation, which increased radiative forcing by 3.68 W m−2. These changes in radiative forcing increased PET and AET by 1.02 mm yr−1 and 0.12 mm yr−1, respectively. Despite spatial variabilities in radiative forcing, PET and AET intensified in approximately 50% (3.03 × 105 km2) and 60% (3.63 × 105 km2) of the area of the Loess Plateau, respectively. However, the PET and AET intensified by radiative forcing accounted for 8% and 5% of the total change in PET and AET, respectively. Although the average contribution of radiative forcing to the Loess Plateau is relatively small, the accumulated impact is non-negligible; it may mediate hydrological cycle and result in substantial water resource stress, particularly in areas implementing intensive ecological restoration.

Experimental Coupling of TOPMODEL with the National Water Model: Effects of Coupling Interface Complexity on Model Performance

AbstractThe study had two objectives; (1) Substitute National Water Model’s (NWM) runoff calculation with a conceptual hydrologic model (TOPography‐based hydrological MODEL [TOPMODEL]) to simplify the model structure and resolve potential drawbacks of applying NWM in headwater catchments. (2) Investigate how varying the coupling interface (location of coupling, type of fluxes used, modification of sub‐models) affects model behavior of when one‐way coupling the NWM’s land surface model (LSM; Noah‐Multi Parameterization) with TOPMODEL using six different scenarios. The one‐way coupled model outperformed NWM and noncoupled TOPMODEL. The coupling option limiting reliance on LSM’s surface and subsurface water fluxes by constraining them within the TOPMODEL structure was the most successful. Performance declined when coupling configurations relied more on LSM calculated fluxes to override TOPMODEL internal processes. Varying the coupling interface brought unexpected changes in TOPMODEL’s parameter sensitivity and water budget even while the statistical score remained similar. The coupling interface represents a source of structural uncertainty that could be identified through conventional evaluation of performance, uncertainty, and sensitivity due to the simple structure of our one‐way coupling design. The study shows that the benefits of combining the strengths of land surface and conceptual hydrological models, while recognizing that structural uncertainty from coupling design needs to be acknowledged.

Hydrological evaluation of high-resolution precipitation estimates from the WRF model in the Third Pole river basins

AbstractIn this study, two sets of precipitation estimates based on the regional Weather Research and Forecasting model (WRF) –the high Asia refined analysis (HAR) and outputs with a 9 km resolution from WRF (WRF-9km) are evaluated at both basin and point scales, and their potential hydrological utilities are investigated by driving the Variable Infiltration Capacity (VIC) large-scale land surface hydrological model in seven Third Pole (TP) basins. The regional climate model (RCM) tends to overestimate the gauge-based estimates by 20–95% in annual means among the selected basins. Relative to the gauge observations, the RCM precipitation estimates can accurately detect daily precipitation events of varying intensities (with absolute bias < 3 mm). The WRF-9km exhibits a high potential for hydrological application in the monsoon-dominated basins in the southeastern TP (with NSE of 0.7–0.9 and bias of -11% to 3%), while the HAR performs well in the upper Indus (UI) and upper Brahmaputra (UB) basins (with NSE of 0.6 and bias of -15% to -9%). Both the RCM precipitation estimates can accurately capture the magnitudes of low and moderate daily streamflow, but show limited capabilities in flood prediction in most of the TP basins. This study provides a comprehensive evaluation of the strength and limitation of RCMs precipitation in hydrological modeling in the TP with complex terrains and sparse gauge observations.

Understanding the Asymmetry of Annual Streamflow Responses to Seasonal Warming in the Western United States

AbstractThe response of annual runoff volume to sub‐annual climate warming is highly uncertain, and the governing mechanisms remain poorly understood, challenging adaptive water management. A typical exemplar is the Western United States, where climate models project substantially stronger warming in the warm season (April to September) than in the cool season (October to March). We investigate the asymmetrical responses of annual and seasonal streamflow changes to warm season and cool season warming for four regionally important basins in the Western United States using an ensemble of four land surface (hydrological) models. Our results show that (i) the general features of seasonal and annual streamflow responses to asymmetrical warming are consistent across models, although the magnitudes vary. The multi‐model mean shows annual runoff declining from 2.0% up to 7.5% under 3°C warm season warming, and from 2.2% up to 4.7% under 3°C cool season warming across the four basins. (ii) The asymmetry of the seasonal evapotranspiration sensitivity to temperature constrains the asymmetry of annual streamflow responses to seasonal warming; and (iii) basins with characteristics such as high ratios of warm to cool season gross incoming water, cooler summers, and colder winters have the strongest relative annual streamflow decreases for warm season warming relative to cool season warming. The pattern in (iii) is explained by the variation of evapotranspiration‐temperature sensitivity in response to a compound set of processes, including the enhanced rate of water holding capacity increase with warmer temperatures, temperature‐related snowmelt‐albedo feedback, and enhanced surface resistance with warmer temperatures.

Data assimilation of satellite-based terrestrial water storage changes into a hydrology land-surface model

Accurate estimation of snow mass or snow water equivalent (SWE) over space and time is required for global and regional predictions of the effects of climate change. This work investigates whether integration of remotely sensed terrestrial water storage (TWS) information, which is derived from the Gravity Recovery and Climate Experiment (GRACE), can improve SWE and streamflow simulations within a semi-distributed hydrology land surface model. A data assimilation (DA) framework was developed to combine TWS observations with the MESH (Modélisation Environnementale Communautaire – Surface Hydrology) model using an ensemble Kalman smoother (EnKS). The snow-dominated Liard Basin was selected as a case study. The proposed assimilation methodology reduced bias of monthly SWE simulations at the basin scale by 17.5% and improved unbiased root-mean-square difference (ubRMSD) by 23%. At the grid scale, the DA method improved ubRMSD values and correlation coefficients for 85% and 97% of the grid cells, respectively. Effects of GRACE DA on streamflow simulations were evaluated against observations from three river gauges, where it effectively improved the simulation of high flows during snowmelt season from April to June. The influence of GRACE DA on the total flow volume and low flows was found to be variable. In general, the use of GRACE observations in the assimilation framework not only improved the simulation of SWE, but also effectively influenced streamflow simulations.

Mapping groundwater abstractions from irrigated agriculture: big data, inverse modeling, and a satellite–model fusion approach

Abstract. The agricultural sector in Saudi Arabia has witnessed rapid growth in both production and area under cultivation over the last few decades. This has prompted some concern over the state and future availability of fossil groundwater resources, which have been used to drive this expansion. Large-scale studies using satellite gravimetric data show a declining trend over this region. However, water management agencies require much more detailed information on both the spatial distribution of agricultural fields and their varying levels of water exploitation through time than coarse gravimetric data can provide. Relying on self-reporting from farm operators or sporadic data collection campaigns to obtain needed information are not feasible options, nor do they allow for retrospective assessments. In this work, a water accounting framework that combines satellite data, meteorological output from weather prediction models, and a modified land surface hydrology model was developed to provide information on both irrigated crop water use and groundwater abstraction rates. Results from the local scale, comprising several thousand individual center-pivot fields, were then used to quantify the regional-scale response. To do this, a semi-automated approach for the delineation of center-pivot fields using a multi-temporal statistical analysis of Landsat 8 data was developed. Next, actual crop evaporation rates were estimated using a two-source energy balance (TSEB) model driven by leaf area index, land surface temperature, and albedo, all of which were derived from Landsat 8. The Community Atmosphere Biosphere Land Exchange (CABLE) model was then adapted to use satellite-based vegetation and related surface variables and forced with a 3 km reanalysis dataset from the Weather Research and Forecasting (WRF) model. Groundwater abstraction rates were then inferred by estimating the irrigation supplied to each individual center pivot, which was determined via an optimization approach that considered CABLE-based estimates of evaporation and TSEB-based satellite estimates. The framework was applied over two study regions in Saudi Arabia: a small-scale experimental facility of around 40 center pivots in Al Kharj that was used for an initial evaluation and a much larger agricultural region in Al Jawf province comprising more than 5000 individual fields across an area exceeding 2500 km2. Total groundwater abstraction for the year 2015 in Al Jawf was estimated at approximately 5.5 billion cubic meters, far exceeding any recharge to the groundwater system and further highlighting the need for a comprehensive water management strategy. Overall, this novel data–model fusion approach facilitates the compilation of national-scale groundwater abstractions while also detailing field-scale information that allows both farmers and water management agencies to make informed water accounting decisions across multiple spatial and temporal scales.