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

TempEst-NEXT National Stream Water Temperature Forecasts
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 208.4 MB
Created: Apr 01, 2026 at 4:22 p.m. (UTC)
Last updated: Apr 01, 2026 at 4:51 p.m. (UTC)
Citation: See how to cite this resource
Content types: CSV Content 
Sharing Status: Public
Views: 27
Downloads: 0
+1 Votes: Be the first one to 
 this.
Comments: No comments (yet)

Abstract

This resource contains a Jupyter Notebook producing weekly water temperature forecasts for a large sample of rivers nationally. As configured, the emphasis is on smaller rivers to minimize data processing. The resource also contains all required preliminary data and an archive of weekly forecasts, which will be updated as more forecasts are run over time. These data and related items of information have not been formally disseminated by NOAA, and do not represent any agency determination, view, or policy.

Included files and directories:

- forecast.ipynb: runs the forecasts. The workflow should be fully automatic.
- coefs_hrrr.pickle: pre-fitted TempEst-NEXT model.
- README.md: general information.
- requirements.txt: pip requirements file.
- forecast_points.json: GeoJSON containing forecast locations.
- full_forecasts: archive of actual forecast results in CSV format. See README.md for details.
- forecasts: site-specific forecasts, including the GFS-forecasted daily max air temperature.
- hist_data: GridMet-based preliminary site data.
- hrrr_data: sample of HRRR forecast archives for a subset of sites, used to adjust climatology.
- final_hist: corrected historical data used for the model.

Directories are uploaded as zip archives, except for the forecast archive.

Subject Keywords

Coverage

Spatial

Coordinate System/Geographic Projection:
WGS 84 EPSG:4326
Coordinate Units:
Decimal degrees
Place/Area Name:
Contiguous United States
North Latitude
49.0000°
East Longitude
-68.0000°
South Latitude
26.0000°
West Longitude
-124.0000°

Temporal

Start Date:
End Date:

Content

README.md

National Water Temperature Forecast Notebook

This Jupyter Notebook generates weekly water temperature forecasts for a large sample of rivers nationally. As configured, the emphasis is on smaller rivers to minimize data processing. If data are not already downloaded, the notebook will spend some time processing preliminary data; however, prepared data will be available in Releases.

General Information

Warning: this is an experimental stream water temperature forecast for the upcoming week. Results are not guaranteed in any capacity. Typical errors are approximately 2.5 C (4 F); any given prediction may have considerably higher errors. For example, in the southern Rocky Mountains during the first forecast run (April 2026), the model, while it did predict high temperature anomalies, was unable to fully account for the unseasonal loss of snowpack and therefore underpredicted temperatures by several degrees.

These data and related items of information have not been formally disseminated by NOAA, and do not represent any agency determination, view, or policy. This research is funded by the NOAA Cooperative Institute for Research to Operations in Hydrology.

The code used to generate these forecasts is available on GitHub at https://github.com/mines-ciroh/natl-temp-forecast (requires support for ecCodes, for which Linux is recommended). Additionally, required preliminary data for the 913 default forecast sites are available on the GitHub Releases page, so the only data retrieval required is the GFS forecast download. Actual weekly forecasts are made available on CUAHSI HydroShare at http://www.hydroshare.org/resource/b8852529788a437a8d697e9b0435b99a.

Usage

  1. git clone https://github.com/mines-ciroh/natl-temp-forecast, or equivalent.
  2. pip install -r requirements.txt.
  3. If using the prepared data, download InputData.zip from Releases and unzip it in the same directory as the notebook.
  4. Launch Jupyter and open forecast.ipynb.
  5. Run the notebook.

Output Variables

Forecast outputs contain the following columns:

  • date: The date. YYYY-MM-DD format.
  • tmax: Daily maximum air temperature (GFS forecast) (deg C).
  • id: Site ID (HUC-12).
  • prediction: Predicted daily mean stream water temperature (deg C).
  • seasonal: Predicted day-of-year (seasonal) mean stream water temperature (deg C).
  • anomaly: Predicted daily departure (anomaly) from seasonal mean stream water teperature (deg C).
  • area: Watershed area in square meters.
  • lat: Prediction point latitude in decimal degrees north.
  • lon: Prediction point longitude in decimal degrees east.

Credits

Funding Agencies

This resource was created using funding from the following sources:
Agency Name Award Title Award Number
National Oceanic and Atmospheric Administration Cooperative Institute for Research to Operations in Hydrology NA22NWS4320003

How to Cite

Philippus, D. (2026). TempEst-NEXT National Stream Water Temperature Forecasts, HydroShare, http://www.hydroshare.org/resource/b8852529788a437a8d697e9b0435b99a

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