- R 100%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
| data | ||
| scripts | ||
| .gitignore | ||
| icon.ico.png | ||
| LICENSE | ||
| MACH.Rproj | ||
| README.md | ||
|
|
Multi-Attribute Catchment Hydrometeorological dataset |
Overview
MACH is large-sample hydrometeorological dataset spanning 1,014 watersheds across the United States. It integrates daily meteorological forcings, streamflow observations, and a comprehensive suite of watershed attributes into a unified basin-scale resource.
The dataset covers 1980-2023:
- 1980-2023 (44 years) for all 1,014 catchments
- 1948-2023 (75 years) for 395 basins
MACH is designed to support reproducible, basin-scale hydrologic analysis across diverse climatic and physiographic regimes.
Scope and Contents
MACH includes:
- Daily meteorological forcings
- Daily streamflow observations
- Static watershed attributes
- Topography
- Soils
- Geology
- Hydrology
- Land cover
- Anthropogenic influences
- Hydroclimatic indices and derived metrics
Designed to support
- Climate and runoff analyses
- Hydrologic model calibration and benchmarking
- Sensitivity and attribution studies
- Long-term watershed change assessment
- Large-sample hydrology research
Basin framework
MACH adopts basins identifiers consistent with the MOPEX and CAMELS frameworks to facilitate cross-dataset comparison and interoperability.
However, all meteorological forcings, streamflow records, watershed attributes, and derived indices have been newly processed, harmonized, and validated for this release. All variables are aggregated and standardized at the basin scale to ensure spatial consistency across catchments.
Primary data sources
MACH integrates information derived from:
- National meteorological forcing datasets including Daymet Version 4 and Gleam 4.2
- Streamflow observations from the USGS National Water Information System
- Attributes from the NHDPlusV2, SSURGO, MRLC, NID, and GIMMS-3G+
Data Access
👉 The MACH dataset is publicly archived on Zenodo. Each release is versioned and permanently archived. Future updates will include extended time series. Please use the DOI corresponding to the version used in your analysis.
Version 4.0 release date 2/19/2026
Feedback and issue reporting
If you encounter inconsistencies or have suggestions for improvement, please open an issue in this repository: https://github.com/k-sink/MACH/issues. Constructive feedback, feature requests, and reproducibility reports are welcome.
About this repository
This repository contains:
- Documentation for the MACH dataset
- Data processing workflows and scripts
- Release notes and version history
- Issue tracking and community feedback
The full dataset is archived on Zenodo.
Contact information
Please contact Katharine Sink at katharine.sink@utdallas.edu with any questions or feedback regarding MACH.