Towards hybrid modeling of the global hydrological cycle

Abstract

Abstract. State-of-the-art global hydrological models (GHMs) exhibit large uncertainties in hydrological simulations due to the complexity, diversity, and heterogeneity of the land surface and subsurface processes, as well as the scale dependency of these processes and associated parameters. Recent progress in machine learning, fueled by relevant Earth observation data streams, may help overcome these challenges. But machine learning methods are not bound by physical laws, and their interpretability is limited by design. In this study, we exemplify a hybrid approach to global hydrological modeling that exploits the data adaptivity of neural networks for representing uncertain processes within a model structure based on physical principles (e.g., mass conservation) that form the basis of GHMs. This combination of machine learning and physical knowledge can potentially lead to data-driven, yet physically consistent and partially interpretable hybrid models. The hybrid hydrological model (H2M), extended from Kraft et al. (2020), simulates the dynamics of snow, soil moisture, and groundwater storage globally at 1∘ spatial resolution and daily time step. Water fluxes are simulated by an embedded recurrent neural network. We trained the model simultaneously against observational products of terrestrial water storage variations (TWS), grid cell runoff (Q), evapotranspiration (ET), and snow water equivalent (SWE) with a multi-task learning approach. We find that the H2M is capable of reproducing key patterns of global water cycle components, with model performances being at least on par with four state-of-the-art GHMs which provide a necessary benchmark for H2M. The neural-network-learned hydrological responses of evapotranspiration and grid cell runoff to antecedent soil moisture states are qualitatively consistent with our understanding and theory. The simulated contributions of groundwater, soil moisture, and snowpack variability to TWS variations are plausible and within the ranges of traditional GHMs. H2M identifies a somewhat stronger role of soil moisture for TWS variations in transitional and tropical regions compared to GHMs. With the findings and analysis, we conclude that H2M provides a new data-driven perspective on modeling the global hydrological cycle and physical responses with machine-learned parameters that is consistent with and complementary to existing global modeling frameworks. The hybrid modeling approaches have a large potential to better leverage ever-increasing Earth observation data streams to advance our understandings of the Earth system and capabilities to monitor and model it.

Widespread increasing vegetation sensitivity to soil moisture

“…We remove grid cells that have model performance worse than the mean of training data itself using cross-validation out-of-bag score (OOB R 2 > 0). We note that the rather low threshold (OOB R 2 > 0) is selected because of a typically significantly decreased model performance in predicting global vegetation productivity for anomalies compared to time series that include the mean seasonal cycles 5 , 70 , while it can still be efficiently used to study relationships between predictor variables and targets. Regions with R 2 < 0 are mostly associated with very low LAI variability or frequent human management (Supplementary Fig.…”

Geospatial Artificial Intelligence (GeoAI) in the Integrated Hydrological and Fluvial Systems Modeling: Review of Current Applications and Trends

“…In recent years, substantial attention has been paid to large-scale and global-scale hydrological modeling [252][253][254]. Although only experimental catchments have sufficient data to perform a reliable hydrological prediction, the global availability of climatological, hydrological, and remote sensing data allows for the parametrizing of the global-level hydrological model.…”

Diagnosing modeling errors of global terrestrial water storage interannual variability

“…Owing to the use of observational data, SINDBAD shows a similar or better performance in simulating TWS seasonality and IAV compared to an ensemble of global hydrological and land surface models (Trautmann et al, 2018). H2M has also been applied for different climatic regions over the globe to simulate TWS seasonality and interannual variability, where it is shown that H2M is capable to learn key patterns of the global water cycle components and has a comparable performance and better local adaptivity, compared to four state-of-the-art global hydrological models (Kraft et al, 2022). In summary, the two model simulations used here provide state-of-the-art where the model parameters and processes are partly learnt from GRACE observations, and are comparable to state-of-the-art hydrological models commonly used.…”

“…. Studies have reported that global hydrological models performed relatively worse in reproducing observed global TWS interannual variability (IAV) compared to other temporal scales (e.g., Zhang et al, 2017;Kraft et al, 2022). Jensen et al (2020) showed that the longer term variability of TWS is dominant in GRACE compared to seasonal variations in Earth system models; and suggested that the models are less reliable in reproducing the timing of peaks of TWS IAV across years compared to the magnitude and frequency.…”