Accumulation of snowfall during winter and snowmelt in the subsequent spring or earlier summer provides a dominant water source in alpine regions. Most land surface and hydrological models use near‐surface air temperature (T_a) thresholds to partition precipitation into snow and rain, underestimating snowfall over drier regions. We developed a snow‐rain partitioning scheme using the wet‐bulb temperature (T_w), which is closer to the surface temperature of a falling hydrometeor than T_a. T_w becomes more depressed in drier environments as derived from T_w depression equation using T_a and surface air humidity, resulting in a greater fraction of snowfall. We implemented this new T_w scheme in the Noah‐MP land surface model and evaluated the model against a high‐quality ground‐based snow product over the contiguous United States. The results suggest that the new T_w scheme substantially improves the model skill in simulating snow depth and snow water equivalent over most snow‐covered grids, especially the higher and drier continental mountain ranges in the Western United States, while it retains the modeling accuracy over the more humid Eastern United States.

Although decades of effort have been devoted to building Physical‐Conceptual (PC) models for predicting the time‐series evolution of geoscientific systems, recent work shows that Machine Learning (ML) based Gated Recurrent Neural Network technology can be used to develop models that are much more accurate. However, the difficulty of extracting physical understanding from ML‐based models complicates their utility for enhancing scientific knowledge regarding system structure and function. Here, we propose a physically interpretable Mass‐Conserving‐Perceptron (MCP) as a way to bridge the gap between PC‐based and ML‐based modeling approaches. The MCP exploits the inherent isomorphism between the directed graph structures underlying both PC models and GRNNs to explicitly represent the mass‐conserving nature of physical processes while enabling the functional nature of such processes to be directly learned (in an interpretable manner) from available data using off‐the‐shelf ML technology. As a proof of concept, we investigate the functional expressivity (capacity) of the MCP, explore its ability to parsimoniously represent the rainfall‐runoff (RR) dynamics of the Leaf River Basin , and demonstrate its utility for scientific hypothesis testing. To conclude, we discuss extensions of the concept to enable ML‐based physical‐conceptual representation of the coupled nature of mass‐energy‐information flows through geoscientific systems.

We investigate the applicability of machine learning technologies to the development of parsimonious, interpretable, catchment‐scale hydrologic models using directed‐graph architectures based on the mass‐conserving perceptron (MCP) as the fundamental computational unit. Here, we focus on architectural complexity (depth) at a single location, rather than universal applicability (breadth) across large samples of catchments. The goal is to discover a minimal representation (numbers of cell‐states and flow paths) that represents the dominant processes that can explain the input‐state‐output behaviors of a given catchment, with particular emphasis given to simulating the full range (high, medium, and low) of flow dynamics. We find that a “HyMod Like” architecture with three cell‐states and two major flow pathways achieves such a representation at our study location, but that the additional incorporation of an input‐bypass mechanism significantly improves the timing and shape of the hydrograph, while the inclusion of bi‐directional groundwater mass exchanges significantly enhances the simulation of baseflow. Overall, our results demonstrate the importance of using multiple diagnostic metrics for model evaluation, while highlighting the need for properly selecting and designing the training metrics based on information‐theoretic foundations that are better suited to extracting information across the full range of flow dynamics. This study sets the stage for interpretable regional‐scale MCP‐based hydrological modeling (using large sample data) by using neural architecture search to determine appropriate minimal representations for catchments in different hydroclimatic regimes.

Due largely to challenges associated with physical interpretability of machine learning (ML) methods, and because model interpretability is key to credibility in management applications, many scientists and practitioners are hesitant to discard traditional physical‐conceptual modeling approaches despite their poorer predictive performance. Here, we examine how to develop parsimonious minimally‐optimal representations that can facilitate better insight regarding system functioning. The term “minimally‐optimal” indicates that the desired outcome can be achieved with the smallest possible effort and resources, while “parsimony” is widely held to support understanding. Accordingly, we suggest that ML‐based modeling should use computational units that are inherently physically‐interpretable, and explore how generic network architectures comprised of Mass‐Conserving‐Perceptron can be used to model dynamical systems in a physically‐interpretable manner. In the context of spatially‐lumped catchment‐scale modeling, we find that both physical interpretability and good predictive performance can be achieved using a “distributed‐state” network with context‐dependent gating and “information‐sharing” across nodes. The distributed‐state mechanism ensures a sufficient number of temporally‐evolving properties of system storage while information‐sharing ensures proper synchronization of such properties. The results indicate that MCP‐based ML models with only a few layers (up to two) and relativity few physical flow pathways (up to three) can play a significant role in ML‐based streamflow modeling.

This study proposed a two-phase risk analysis scheme for flood management considering flood inundation losses, including: (1) simplified qualitative-based risk analysis incorporating the principles of failure mode and effect analysis (FMEA) to identify all potential failure modes associated with candidate flood control measures, to formulate a remedial action plan aiming for mitigating the inundation risk within an engineering system; and (2) detailed quantitative-based risk analysis to employ numerical models to specify the consequences including flood extent and resulting losses. Conventional qualitative-based risk analysis methods have shown to be time-efficient but without quantitative information for decision making. However, quantitative-based risk analysis methods have shown to be time- and cost- consuming for a full spectrum investigation. The proposed scheme takes the advantages of both qualitative-based and quantitative-based approaches of time-efficient, cost-saving, objective and quantitative features for better flood management in term of expected loss. The proposed scheme was applied to evaluate the Chiang-Yuan Drainage system located on Lin-Bien River in southern Taiwan, as a case study. The remedial action plan given by the proposed scheme has shown to greatly reduce the inundation area in both highlands and lowlands. These measures was investigated to reduce the water volume in the inundation area by 0.2 million cubic meters, even in the scenario that the flood recurrence interval exceeded the normal (10-year) design standard. Our results showed that the high downstream water stage in the downstream boundary may increase the inundation area both in downstream and upstream and along the original drainage channel in the vicinity of the diversion. The selected measures given by the proposed scheme have shown to substantially reduce the flood risk and resulting loss, taking account of various scenarios: short duration precipitation, decreased channel conveyance, pump station failure and so forth.