Streamflow Hydrology Estimate Using Machine Learning (SHEM)

Abstract

Continuity and accuracy of near real‐time streamflow gauge (streamgage) data are critical for flood forecasting, assessing imminent risk, and implementing flood mitigation activities. Without these data, decision makers and first responders are limited in their ability to effectively allocate resources, implement evacuations to save lives, and reduce property losses. The Streamflow Hydrology Estimate using Machine Learning (SHEM) is a new predictive model for providing accurate and timely proxy streamflow data for inoperative streamgages. SHEM relies on machine learning (“training”) to process and interpret large volumes (“big data”) of historic complex hydrologic information. Continually updated with real‐time streamflow data, the model constructs a virtual dataset index of correlations and groups (clusters) of relationship correlations between selected streamgages in a watershed and under differing flow conditions. Using these datasets, SHEM interpolates estimated discharge and time data for any indexed streamgage that stops transmitting data. These estimates are continuously tested, scored, and revised using multiple regression analysis processes and methodologies. The SHEM model was tested in Idaho and Washington in four diverse watersheds, and the model's estimates were then compared to the actual recorded data for the same time period. Results from all watersheds revealed a high correlation, validating both the degree of accuracy and reliability of the model.

Identifying Subsurface Drainage using Satellite Big Data and Machine Learning via Google Earth Engine

“…Random forest machine learning (RFML) is a supervised classification algorithm that constructs a multitude of decision trees and predicts class labels, using a random subset of training samples and variables (Breiman, 2001). The RFML has become popular within the remote sensing and hydrology communities due to its accuracy (Belgiu & Drăguţ, 2016;Gómez et al, 2016;Petty & Dhingra, 2018). For land surface and crop type monitoring, the RFML has been shown to produce higher accuracies than other ML techniques such as Maximum Likelihood Classifier, Neural Network, and Support Vector Machine (Gómez et al, 2016;Ma et al, 2017;Ok et al, 2012).…”

Links between Teleconnection Patterns and Water Level Regime of Selected Polish Lakes

“…Even more advanced computational methods (like machine learning and artificial neural networks) can be used to predict future values of given hydro-and meteorological variables. These methods were used for instance to forecast runoff values for the effective reservoir management [91], to predict evaporation [92,93] or to establish a rainfall-runoff models for forecasting the river flow [94][95][96][97].…”

Featured Collection Introduction: National Flood Interoperability Experiment II

“…Essential to a robust national water forecasting system is the ability to assimilate existing streamflow measurements. As part of their contribution to the NFIE research, Petty and Dhingra () developed a machine learning program that would provide immediate estimates of streamflow for gages out of operation based on the operable stations within the same region. Their work was tested in watersheds in Washington and Idaho and showed a very high degree of accuracy.…”