Regional rainfall–runoff modeling is an old but still mostly outstanding problem in the hydrological sciences. The problem currently is that traditional hydrological models degrade significantly in performance when calibrated for multiple basins together instead of for a single basin alone. In this paper, we propose a novel, data-driven approach using Long Short-Term Memory networks (LSTMs) and demonstrate that under a “big data” paradigm, this is not necessarily the case. By training a single LSTM model on 531 basins from the CAMELS dataset using meteorological time series data and static catchment attributes, we were able to significantly improve performance compared to a set of several different hydrological benchmark models. Our proposed approach not only significantly outperforms hydrological models that were calibrated regionally, but also achieves better performance than hydrological models that were calibrated for each basin individually. Furthermore, we propose an adaption to the standard LSTM architecture, which we call an Entity-Aware-LSTM (EA-LSTM), that allows for learning catchment similarities as a feature layer in a deep learning model. We show that these learned catchment similarities correspond well to what we would expect from prior hydrological understanding.

Long short‐term memory (LSTM) networks offer unprecedented accuracy for prediction in ungauged basins. We trained and tested several LSTMs on 531 basins from the CAMELS data set using k‐fold validation, so that predictions were made in basins that supplied no training data. The training and test data set included ∼30 years of daily rainfall‐runoff data from catchments in the United States ranging in size from 4 to 2,000 km² with aridity index from 0.22 to 5.20, and including 12 of the 13 IGPB vegetated land cover classifications. This effectively “ungauged” model was benchmarked over a 15‐year validation period against the Sacramento Soil Moisture Accounting (SAC‐SMA) model and also against the NOAA National Water Model reanalysis. SAC‐SMA was calibrated separately for each basin using 15 years of daily data. The out‐of‐sample LSTM had higher median Nash‐Sutcliffe Efficiencies across the 531 basins (0.69) than either the calibrated SAC‐SMA (0.64) or the National Water Model (0.58). This indicates that there is (typically) sufficient information in available catchment attributes data about similarities and differences between catchment‐level rainfall‐runoff behaviors to provide out‐of‐sample simulations that are generally more accurate than current models under ideal (i.e., calibrated) conditions. We found evidence that adding physical constraints to the LSTM models might improve simulations, which we suggest motivates future research related to physics‐guided machine learning.

This paper is derived from a keynote talk given at the Google's 2020 Flood Forecasting Meets Machine Learning Workshop. Recent experiments applying deep learning to rainfall‐runoff simulation indicate that there is significantly more information in large‐scale hydrological data sets than hydrologists have been able to translate into theory or models. While there is a growing interest in machine learning in the hydrological sciences community, in many ways, our community still holds deeply subjective and nonevidence‐based preferences for models based on a certain type of “process understanding” that has historically not translated into accurate theory, models, or predictions. This commentary is a call to action for the hydrology community to focus on developing a quantitative understanding of where and when hydrological process understanding is valuable in a modeling discipline increasingly dominated by machine learning. We offer some potential perspectives and preliminary examples about how this might be accomplished.

Floods are one of the most common natural disasters, with a disproportionate impact in developing countries that often lack dense streamflow gauge networks¹. Accurate and timely warnings are critical for mitigating flood risks², but hydrological simulation models typically must be calibrated to long data records in each watershed. Here we show that artificial intelligence-based forecasting achieves reliability in predicting extreme riverine events in ungauged watersheds at up to a five-day lead time that is similar to or better than the reliability of nowcasts (zero-day lead time) from a current state-of-the-art global modelling system (the Copernicus Emergency Management Service Global Flood Awareness System). In addition, we achieve accuracies over five-year return period events that are similar to or better than current accuracies over one-year return period events. This means that artificial intelligence can provide flood warnings earlier and over larger and more impactful events in ungauged basins. The model developed here was incorporated into an operational early warning system that produces publicly available (free and open) forecasts in real time in over 80 countries. This work highlights a need for increasing the availability of hydrological data to continue to improve global access to reliable flood warnings.

Hydrological signatures are now used for a wide range of purposes, including catchment classification, process exploration, and hydrological model calibration. The recent boost in the popularity and number of signatures has however not been accompanied by the development of clear guidance on signature selection. Here we propose that exploring the predictability of signatures in space provides important insights into their drivers and their sensitivity to data uncertainties and is hence useful for signature selection. We use three complementary approaches to compare and rank 15 commonly used signatures, which we evaluate in 600+ U.S. catchments from the Catchment Attributes and MEteorology for Large‐sample Studies (CAMELS) data set. First, we employ machine learning (random forests) to explore how attributes characterizing the climatic conditions, topography, land cover, soil, and geology influence (or not) the signatures. Second, we use simulations of the Sacramento Soil Moisture Accounting model to benchmark the random forest predictions. Third, we take advantage of the large sample of CAMELS catchments to characterize the spatial autocorrelation (using Moran's I) of the signature field. These three approaches lead to remarkably similar rankings of the signatures. We show (i) that signatures with the noisiest spatial pattern tend to be poorly captured by hydrological simulations, (ii) that their relationship to catchments attributes are elusive (in particular they are not well explained by climatic indices), and (iii) that they are particularly sensitive to discharge uncertainties. We suggest that a better understanding of the drivers of hydrological signatures and a better characterization of their uncertainties would increase their value in hydrological studies.