Using Machine Learning to Identify Hydrologic Signatures With an Encoder–Decoder Framework

Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and th...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Water resources research Ročník 59; číslo 3
Hlavní autori: Botterill, Tom E., McMillan, Hilary K.
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Washington John Wiley & Sons, Inc 01.03.2023
Predmet:
Base flow
Climate
Climatic data
Coders
Coefficients
data collection
deep learning
Dynamics
encoder–decoder
flow indices
Flow mapping
flow metric
Hydrologic models
hydrologic signature
Hydrology
Information processing
Learning algorithms
Machine learning
Maximum flow
meteorological data
Model accuracy
Modelling
Neural networks
Outflow
Precipitation
prediction
Rainfall-runoff relationships
Runoff
Runoff models
Seasonal variations
Seasonality
Sensitivity analysis
Signatures
Stream discharge
Stream flow
time series analysis
Watersheds
ISSN:0043-1397, 1944-7973
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Hydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and that are derived directly from the data. We compare the learned signatures to classical signatures, interpret their meaning, and use them to build rainfall‐runoff models in otherwise ungauged watersheds. Our model has an encoder–decoder structure. The encoder is a convolutional neural net mapping historical flow and climate data to a low‐dimensional vector encoding, analogous to hydrological signatures. The decoder structure includes stores and fluxes similar to a classical hydrologic model. For each timestep, the decoder uses current climate data, watershed attributes and the encoding to predict coefficients that distribute precipitation between stores and store outflow coefficients. The model is trained end‐to‐end on the U.S. CAMELS watershed data set to minimize streamflow error. We show that learned signatures can extract new information from streamflow series, because using learned signatures as input to the process‐informed model improves prediction accuracy over benchmark configurations that use classical signatures or no signatures. We interpret learned signatures by correlation with classical signatures, and by using sensitivity analysis to assess their impact on modeled store dynamics. Learned signatures are spatially correlated and relate to streamflow dynamics including seasonality, high and low extremes, baseflow and recessions. We conclude that process‐informed ML models and other applications using hydrologic signatures may benefit from replacing expert‐selected signatures with learned signatures. Key Points We built an encoder–decoder network to learn an optimal equivalent of hydrologic signatures Using learned signatures as input to a process‐informed ML model improves prediction accuracy over benchmark configurations We interpret learned signatures by correlation with classical signatures and by sensitivity analysis of their impact on model store dynamics
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0043-1397
1944-7973
DOI:10.1029/2022WR033091