River Weathering Data and Machine-Learning Workflows
Derived geochemical data, watershed predictors, and R code linking inverse modeling with machine learning to study rock-weathering fluxes.
Aug 8, 2026
The Global RIver-lanD (GRID) project connects river hydrogeochemical observations with watershed information to support research on weathering, water quality, and river biogeochemical cycles.
Explore released datasets, analysis code, teaching resources, and opportunities to collaborate.
GRID inventory
As of October 19, 2026
- Sites
- 154,577
- Hydrogeochemical variables
- 91
- Watershed properties
- 139
- Watershed property data sources
- 19
These totals describe the assembled GRID inventory. Individual releases cover specific subsets. Not every hydrogeochemical variable or watershed property is available at every site. View coverage and contents.
Featured research output
River weathering data and machine-learning workflows
An August 2026 GRID output connects river chemistry and watershed context with geochemical inverse modeling and machine-learning analysis across the contiguous United States.
Datasets and research outputs
Versioned releases document the data, code, and scope of individual studies.
No matching items
News and research highlights
No matching items
Help shape the next stage
GRID-SI is a proposed development program to improve chemistry documentation, catchment links, and reusable workflows. Read the development roadmap or discuss a contribution.