About This Item
- Full TextFull Text(subscription required)
- Pay-Per-View PurchasePay-Per-View
Purchase Options Explain
Share This Item
Extended Abstract: Improving U.S. Gulf Region Streamflow Predictions from the U.S. National Water Model with Machine Learning
Hydrological science is in the midst of an exciting revolution with rapid improvement coming in two ways: (1) The domain of hydrology models is expanding to a continental scale and beyond and (2) the predictive accuracy of hydrology models is improving at an accelerating rate with the integration of machine learning (ML) and, more specifically, deep learning (DL). We present results throughout the U.S. Gulf region (Gulf) from post-processing daily streamflow predictions in the U.S. National Water Model (NWM), a large domain hydrologic forecasting system, with long short-term memory (LSTM) networks, a DL method that is particularly well suited to model hydrologic processes. We compared the discrepancies between the NWM and the LSTM to show the relatively high predictive skill of the NWM in Mississippi and Louisiana, and the potential for improving predictions in Alabama, Georgia, and Florida using ML. Figures 1 and 2 provide LSTM workflow and post-processing improvement versus the NWM.
Pay-Per-View Purchase Options
The article is available through a document delivery service. Explain these Purchase Options.
|Watermarked PDF Document: $14|
|Open PDF Document: $24|