Post processing the U.S. National Water Model with a Long Short-Term Memory network

Hydrol. Earth Syst. Sci.

Hochreiter

et al. 2021

Self Cite

Abstract. A deep learning rainfall–runoff model can take multiple meteorological forcing products as input and learn to combine them in spatially and temporally dynamic ways. This is demonstrated with Long Short-Term Memory networks (LSTMs) trained over basins in the continental US, using the Catchment Attributes and Meteorological data set for Large Sample Studies (CAMELS). Using meteorological input from different data products (North American Land Data Assimilation System, NLDAS, Maurer, and Daymet) in a single LSTM significantly improved simulation accuracy relative to using only individual meteorological products. A sensitivity analysis showed that the LSTM combines precipitation products in different ways, depending on location, and also in different ways for the simulation of different parts of the hydrograph.

Section: Discussionmentioning

confidence: 99%

A note on leveraging synergy in multiple meteorological data sets with deep learning for rainfall–runoff modeling

Hydrol. Earth Syst. Sci.

Hochreiter

et al. 2021

Self Cite

“…This finding (differences between pure ML and physics-informed ML) is worth discussing. The project of adding physical constraints to ML is an active area of research across most fields of science and engineering (Karniadakis et al, 2021), including hydrology (e.g., Zhao et al, 2019;Jiang et al, 2020;Frame et al, 2020). It is important to understand that there is only one type of situation in which adding any type of constraint (physically-based or otherwise) to a data-driven model can add value: if constraints help optimization.…”

Section: Conclusion and Discussionmentioning

confidence: 99%

“…These metrics depend heavily on the observed flow characteristics during a particular test period and, because they are less stable, are somewhat less useful in terms of drawing general conclusions. We report them here primarily for continuity with previous studies (Kratzert et al, 2019b(Kratzert et al, , a, 2021Frame et al, 2020;Nearing et al, 2020a;Klotz et al, 2021;Gauch et al, 2021), and because one of the objectives of this paper (Section 2.2) is to expand on the high flow (FHV) analysis by benchmarking on annual peak flows.…”

Section: Fms VIImentioning

confidence: 99%

“…as a land surface component, kinematic wave overland flow, and Muskingum-Cunge channel routing. NWM-Rv2 was previously used as a benchmark for LSTM simulations in CAMELS byKratzert et al (2019a andFrame et al (2020). Public data from NWM-Rv2 is hourly and CONUS-wide -we pulled hourly flow estimates from the USGS gauges in the CAMELS data set and averaged these hourly data to daily over the time period October 1, 1980 through September 30, 2014.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Deep learning rainfall-runoff predictions of extreme events

Frame

et al. 2021

Preprint

Self Cite

Abstract. The most accurate rainfall-runoff predictions are currently based on deep learning. There is a concern among hydrologists that data-driven models based on deep learning may not be reliable in extrapolation or for predicting extreme events. This study tests that hypothesis using Long Short-Term Memory networks (LSTMs) and an LSTM variant that is architecturally constrained to conserve mass. The LSTM (and the mass-conserving LSTM variant) remained relatively accurate in predicting extreme (high return-period) events compared to both a conceptual model (the Sacramento Model) and a process-based model (US National Water Model), even when extreme events were not included in the training period. Adding mass balance constraints to the data-driven model (LSTM) reduced model skill during extreme events.

“…We trained all our machine learning models with the neuralhydrology Python library (https://doi.org/10.5281/zenodo.4688003, . All code to reproduce our models and analyses is available at https://doi.org/10.5281/zenodo.4687991 (Gauch, 2021). The trained models and their predictions are available at https://doi.org/10.5281/zenodo.4071885 (Gauch et al, 2020a).…”

Section: B3 Multi-timescale Input Single-timescale Outputmentioning

confidence: 99%

Rainfall–runoff prediction at multiple timescales with a single Long Short-Term Memory network

Gauch

Hydrol. Earth Syst. Sci.

et al. 2021

Self Cite

164

Abstract. Long Short-Term Memory (LSTM) networks have been applied to daily discharge prediction with remarkable success. Many practical applications, however, require predictions at more granular timescales. For instance, accurate prediction of short but extreme flood peaks can make a lifesaving difference, yet such peaks may escape the coarse temporal resolution of daily predictions. Naively training an LSTM on hourly data, however, entails very long input sequences that make learning difficult and computationally expensive. In this study, we propose two multi-timescale LSTM (MTS-LSTM) architectures that jointly predict multiple timescales within one model, as they process long-past inputs at a different temporal resolution than more recent inputs. In a benchmark on 516 basins across the continental United States, these models achieved significantly higher Nash–Sutcliffe efficiency (NSE) values than the US National Water Model. Compared to naive prediction with distinct LSTMs per timescale, the multi-timescale architectures are computationally more efficient with no loss in accuracy. Beyond prediction quality, the multi-timescale LSTM can process different input variables at different timescales, which is especially relevant to operational applications where the lead time of meteorological forcings depends on their temporal resolution.