Data for "Streamflow prediction in poorly gauged watersheds in the United States through data-driven sparse sensing"
| Authors: | |
|---|---|
| Owners: | Kun Zhang |
| Type: | Resource |
| Storage: | The size of this resource is 38.7 MB |
| Created: | Mar 29, 2023 at 2:53 a.m. |
| Last updated: | Apr 04, 2023 at 1:04 p.m. (Metadata update) |
| Published date: | Apr 04, 2023 at 1:04 p.m. |
| DOI: | 10.4211/hs.49b0f3b0f6924b2d917b3659fb03926b |
| Citation: | See how to cite this resource |
| Sharing Status: | Published |
|---|---|
| Views: | 864 |
| Downloads: | 21 |
| +1 Votes: | Be the first one to this. |
| Comments: | No comments (yet) |
Abstract
This archive includes data used in Zhang et al.'s paper "Streamflow prediction in poorly gauged watersheds in the United States through data-driven sparse sensing", which has been published in Water Resources Research (WRR) (https://doi.org/10.1029/2022WR034092) The archive contains 1) raw data (daily-scale CAMELS streamflow data and watershed attributes) and 2) MATLAB scripts used to perform data-driven sparse sensing and generate sample figures. The streamflow data used in this study was retrieved from the Catchment Attributes and Meteorology for Large-sample Studies (CAMELS) dataset (https://ral.ucar.edu/solutions/products/camels) The MATLAB code used for data-driven sparse sensing was retrieved from the Github repository by Krithika Manohar (https://github.com/kmanohar/SSPOR_pub) and customized for this study.
Subject Keywords
Coverage
Spatial
Temporal
| Start Date: | 01/01/1981 |
|---|---|
| End Date: | 12/31/2010 |
Content
readme.txt
This archive contains several MATLAB data and script files:
1) DataDSS.mat:
This MATLAB data file contains the data used in this study. It contains the following variables:
A) Data (18x3 cell array): CAMELS daily streamflow time-series in 543 watersheds across CONUS. 18 rows refer to 18 2-digit HUCs; Three columns refer to 1) gauge ID, 2) datetime array, and 3) streamflow time-series in the dimension of 365x30 (number of days x number of years) (mm/d).
B) Attributes (543x60 table): CAMELS watershed attributes in 543 watersheds across CONUS.
C) regions (543x1 matrix): Index (ranging from 1 to 9) representing the classification of watersheds into nine regions classified by US Census Bureau.
2) DSS.m:
This MATLAB script is a function that can perform data-driven sparse sensing given appropriate input
Input: x -- time-series for training (e.g., streamflow in nearby watersheds) (normalized)
Y -- time-series for prediction (e.g., streamflow in target watershed) (normalized)
y -- time-series for prediction (e.g., streamflow in target watershed) (original)
R -- number of columns and samples to be considered in prediction
Output: yQR -- predicted time-series
yPivot -- optimal times for sampling
nse -- NSE between the measured and predicted time-series
nsem -- modified NSE between the measured and predicted time-series
3) DSSmaster.m:
This MATLAB script is used to demonstrate how to use the data-driven sparse sensing to predict streamflow time-series in ungauged watersheds using CAMELS dataset as an example. Sample figures (different from the paper) can be plotted.
4) Others
pretreat.m is a script used to normalize the dataset using Z-score method.
daboxplot.m is a script retrieved from Povilas Karvelis (https://github.com/frank-pk/DataViz).
Credits
Funding Agencies
This resource was created using funding from the following sources:
| Agency Name | Award Title | Award Number |
|---|---|---|
| Army Corps of Engineers (USACE) Engineer Research and Development Center (ERDC) | Novel Technologies to Mitigate Water Contamination for Resilient Infrastructure | W9132T2220001 |
How to Cite
This resource is shared under the Creative Commons Attribution CC BY.
http://creativecommons.org/licenses/by/4.0/
Comments
There are currently no comments
New Comment