Checking for non-preferred file/folder path names (may take a long time depending on the number of files/folders) ...

Streamflow composition in U.S. rivers is shifting toward recent precipitation
Authors:
Owners: This resource does not have an owner who is an active HydroShare user. Contact CUAHSI (help@cuahsi.org) for information on this resource.
Type: Resource
Storage: The size of this resource is 67.1 MB
Created: May 11, 2026 at 5:39 p.m. (UTC)
Last updated: Jul 06, 2026 at 2:35 p.m. (UTC) (Metadata update)
Published date: Jul 06, 2026 at 2:35 p.m. (UTC)
DOI: 10.4211/hs.8c5e6c59f1ac411caf809a735cd6f6b1
Citation: See how to cite this resource
Sharing Status: Published
Views: 254
Downloads: 18
+1 Votes: Be the first one to 
 this.
Comments: No comments (yet)

Abstract

This data product is related to a journal article by Chuqiang Chen and Admin Husic entitled "Streamflow composition in U.S. rivers is shifting toward recent precipitation".

Abstract:
The fraction of streamflow derived from recent precipitation—termed ‘event water’—profoundly impacts water quality and flood risk. Here, we show that over the past four decades, this fraction has significantly increased in 27% of 754 U.S. catchments. We achieve this first continental-scale reconstruction by applying a deep learning model to a solute tracer dataset spanning nearly 24,000 storm events. Significant increases in event water fraction occur nearly twice as often as significant increases in streamflow, indicating that the composition of streamflow is often changing even where its magnitude is stationary. Using explainability methods, we identify intensifying precipitation and urbanization, together with the loss of forest cover, as the primary drivers. These results reveal a widespread shift towards the rapid conversion of precipitation to streamflow by watersheds, signaling a need to adapt water management strategies for a future of accelerating climate and land-use pressures.

Subject Keywords

Content

README.txt 

# Decadal reconstruction of event water fraction across CONUS

## This repository provides code for the reconstruction of event water fraction across 754 sites from 1986 to 2024 and creating figures in the main text of the paper. Information of study sites, annual reconstructed event water and streamflow, annual climate and land-use drivers, and trend results are also provided.

## Repository Structure

* **0. DATA/**
	This includes information of study sites, annual reconstructed event water and streamflow, annual climate and land-use drivers, and trend test results.

* **1. Data Preparation/**
	Retrieve meteorological, catchment attribute, streamflow, and specific conductance data. Conduct basin delineation.

* **2. Hydrograph separation/**
	Perform hydrograph separation.

* **3. Model Training/**
	Train event water and streamflow LSTM models.

* **4. Post Processing/**
	Aggregate hourly flow and meteorological variables to annual scale. 

* **5. Analysis/**
	Perform Mann-Kendall test.

* **6. Figures/**
	Create figures in the main text of the paper.

##License

This code is licensed under the MIT License. See LICENSE.txt for details.

Credits

Funding Agencies

This resource was created using funding from the following sources:
Agency Name Award Title Award Number
U.S. National Science Foundation CAREER: Dynamic connectivity: a research and educational frontier for sustainable environmental management under climate and land use uncertainty 2438017

How to Cite

Chen, C., Husic, A. (2026). Streamflow composition in U.S. rivers is shifting toward recent precipitation, HydroShare, https://doi.org/10.4211/hs.8c5e6c59f1ac411caf809a735cd6f6b1

This resource is shared under the Creative Commons Attribution CC BY.

 http://creativecommons.org/licenses/by/4.0/
CC-BY

Comments

There are currently no comments

New Comment

required