A document by Jonathan Frame. Click on the document to view its contents.

All rainfall-runoff models tend to degrade in performance when used in arid basins, as compared to their performance in basins with plenty of precipitation with respect to actual evapotranspiration. Neural networks are not exempt. Even though deep learning models such as LSTM provide superior predictive performance of streamflow in arid regions, as compared to their conceptual counterparts, performance degradation is still apparent as the aridity index increases (models perform worse when used in more arid basins). Physically, runoff generation in arid regions requires a critical mass of precipitation to overcome many hydrologic processes to eventually trigger overland flow. Conceptually this occurs when suitable antecedent soil moisture conditions match with suitable atmospheric conditions and land surface energy flux conditions. The alignment of these conditions causes a spontaneous shift in the hydrologic phase from initial abstraction to runoff. Runoff then persists until another spontaneous alignment of conditions shifts the hydrologic phase from runoff back to abstraction. We present evidence that the reason poor model performance under these scenarios is not model structure, but the inherent sensitivity to spontaneous synchronization of soil, atmospheric and land surface energy conditions. Both conceptual and deep learning models demonstrate these non-reciprocal phase transitions dynamically [1], but fail to calibrate correctly to these conditions due to their infrequent recurrence in the hydrography relative to their spontaneity. Deep learning models in particular contain sufficient dynamic complexity to represent this behavior well [2], but perhaps a rethinking or model training for representing these conditions is necessary. Finally will test the sensitivity of model training/calibration under hydrologic phase shifts with respect to data disinformation in these regions [3].Fruchart, M., Hanai, R., Littlewood, P. B., Vitelli, V., & Information, S. (2021). Non-reciprocal phase transitions. Nature, 592(April 2020), 363–369. https://doi.org/https://doi.org/10.1038/s41586-021-03375-9Gauthier, D. J. (2021). Next generation reservoir computing. Nature Communications, 2021, 1–8. https://doi.org/10.1038/s41467-021-25801-2Beven, K. (2023). Benchmarking hydrological models for an uncertain future. In Hydrological Processes (Vol. 37, Issue 5). John Wiley and Sons Ltd. https://doi.org/10.1002/hyp.14882

Rapid and accurate maps of floods across large domains, with high temporal resolution capturing event peaks, have applications for flood forecasting and resilience, damage assessment, and parametric insurance. Satellite imagery produces incomplete observations spatially and temporally, and hydrodynamic models require tradeoffs between computational efficiency and accuracy. We address these challenges with a novel flood model which predicts surface water area from the U.S. National Water Model using a convolutional neural network (NWM-CNN). We trained NWM-CNN on 780 flood events, at a 250m resolution with an RMSE of 4.58% on held out validation geographies. We demonstrate NWM-CNN across California during the 2023 atmospheric rivers, comparing predictions against Sentinel-1 mapped flood observations. Historically, we compared the data from 1979-2023 to flood damage reports in Sacramento County, California. Results show that NWM-CNN captures inundation extent better than the Height Above Nearest Drainage (HAND) approach (25% to 36% RMSE, respectively).

It has been proposed that conservation laws might not be beneficial for accurate hydrological modeling due to errors in input (precipitation) and target (streamflow) data (particularly at the event time scale), and this might explain why deep learning models (which are not based on enforcing closure) can out-perform catchment-scale conceptual and process-based models at predicting streamflow. We test this hypothesis at the event and multi-year time scale using physics-informed (mass conserving) machine learning and find that: (1) enforcing closure in the rainfall-runoff mass balance does appear to harm the overall skill of hydrological models, (2) deep learning models learn to account for spatiotemporally variable biases in data (3) however this “closure” effect accounts for only a small fraction of the difference in predictive skill between deep learning and conceptual models.

Recent advancement of computational linguistics, machine learning, including a variety of toolboxes for Natural Language Processing (NLP), help facilitate analysis of vast electronic corpuses for a multitude of objectives. Research papers published as electronic text files in different journals offer windows into trending topics and developments, and NLP allows us to extract information and insight about these trends. This project applies Latent Dirichlet Allocation (LDA) Topic Modeling for bibliometric analyses of all abstracts in selected high-impact (Impact Factor > 0.9) journals in hydrology. Topic modeling uses statistical algorithms to extract semantic information from a collection of texts and has become an emerging quantitative method to assess substantial textual data. The resulting generated topics are interpretable based on our prior knowledge of hydrology and related sub-disciplines. Comparative topic trend, term, and document level cluster analyses based on different time periods was performed. These analyses revealed topics such as climate change research gaining popularity in Hydrology over the last decade. An inter-topic correlation analysis also revealed the nature of information exchange and absorption between various communities within the hydrology domain. The primary objective of this work is to allow researchers to explore new branches and connections in the Hydrology literature, and to facilitate comprehensive and inclusive literature reviews. We aim to use these results combined with probability distribution between topics, journals and authors to create an ontology that is useful for scientists and environmental consultants for exploring relevant literature based on topics and topic relationships.