We confirm that energy dissipation weighting provides the most accurate approach to determining the effective hydraulic conductivity (Keff) of a binary K grid. Machine learning and deep learning algorithms of varying complexity (decision tree, vanilla CNN, UNET) can infer Keff with extremely high accuracy (R2 > 0.99), even given only the fraction of the grid occupied by the high K medium. Adding information derived from the energy dissipation distribution improved each algorithm. However, all methods failed to infer Keff accurately for outlier cases, all of which were inferred accurately using energy dissipation weighting directly. The UNET architecture could be trained to infer the energy dissipation weighting pattern from an image of the K distribution with high fidelity, although it was less accurate for cases with highly localized structures that controlled flow. Furthermore, the UNET architecture learned to infer the energy dissipation weighting even if it was not trained on this information. However, the weights were represented within the UNET in a way that was not immediately interpretable by a human user. This reiterates the idea that even if ML/DL algorithms are trained to make some hydrologic predictions accurately, they must be designed and trained to provide each user-required output if their results are to be used to improve our understanding of hydrologic systems most effectively.

Accurate estimation of the spatio-temporal distribution of snow water equivalent is essential given its global importance for understanding climate dynamics and climate change, and as a source of fresh water. Here, we explore the potential of using the Long Short-Term Memory (LSTM) network for continental and regional scale modeling of daily snow accumulation and melt dynamics at 4-km pixel resolution across the conterminous US (CONUS). To reduce training costs (data are available for ~0.31 million snowy pixels), we combine spatial sampling with stagewise model development, whereby the network is first pretrained across the entire CONUS and then subjected to regional fine-tuning. Accordingly, model evaluation is focused on out-of-sample predictive performance across space (analogous to the prediction in ungauged basins problem). We find that, given identical inputs (precipitation, temperature and elevation), a single CONUS-wide LSTM provides significantly better spatio-temporal generalization than a regionally calibrated version of the physical-conceptual temperature-index-based SNOW17 model. Adding more meteorological information (dew point temperature, vapor pressure deficit, longwave radiation and shortwave radiation) further improves model performance, while rendering redundant the local information provided by elevation. Overall, the LSTM exhibits better transferability than SNOW17 to locations that were not included in the training data set, reinforcing the advantages of structure learning over parameter learning. Our results suggest that an LSTM-based approach could be used to develop continental/global-scale systems for modeling snow dynamics.

A key step in model development is selection of an appropriate representational system, including both the representation of what is observed (the data), and the formal mathematical structure used to construct the input-state-output mapping. These choices are critical, because they completely determine the questions we can ask, the nature of the analyses and inferences we can perform, and the answers that we can obtain. Accordingly, a representation that is suitable for one kind of investigation might be limited in its ability to support some other kind. Arguably, how different representational approaches affect what we can learn from data is poorly understood. This paper explores three complementary representational strategies as vehicles for understanding how catchment-scale hydrological processes vary across hydro-geo-climatologically diverse Chile. Specifically, we test a lumped water-balance model (GR4J), a data-based dynamical systems model (LSTM), and a data-based regression-tree model (Random Forest). Insights were obtained regarding system memory encoded in data, spatial transferability by use of surrogate attributes, and informational deficiencies of the dataset that limit our ability to learn an adequate input-output relationship. As expected, each approach exhibits specific strengths, with LSTM providing the best characterization of dynamics, GR4J being the most robust under informationally deficient conditions, and RF being most supportive of interpretation. Overall, the complementary nature of the three approaches suggests the value of adopting a multi-representational framework in order to more fully extract information from the data. Our results show that a multi-representational approach better supports the goals of prediction, understanding, and scientific discovery in Hydrology.

Satellite-based precipitation products (SPPs) with short latencies provide a new opportunity for flood forecasting in ungauged basins. However, the larger uncertainties associated with such near-real-time SPPs can influence the accuracy of the resulting flood forecast. Here we propose a real-time updating method, referred to as “Constrained Runoff Correction (CRC-M)” that is based on the use of multi-source SPPs. The method is based on the hypothesis that the range over different near-real-time SPPs provides insight regarding the approximate range in which the true rainfall value lies, during the current period. Accordingly, the constrained runoff correction is performed in such a way as to be consistent with this range, and with the observed value of discharge at the basin outlet. Evaluation using real-data indicates that the new method performs well, with Nash–Sutcliffe (NS) values of 0.85 and 0.91 during calibration and evaluation, respectively. The necessity and value of imposing constraints is demonstrated by comparing CRC-M against a control, referred to as “Unconstrained Runoff Correction” (URC-S). Experiments indicate that the key factors resulting in good performance are 1) wider constraint ranges, and 2) relatively reliable SPPs. Further, inclusion of redundant information may only result in slight improvements to forecast performance, and can even cause the performance to deteriorate. Overall, the CRC-M method can result in accurate and stable flood forecasts for ungauged basins, without the need for increased model complexity (i.e., the numbers of model parameters).