Data for: "Environmental drivers of river ecosystem metabolism and their implications for enhanced weathering across the contiguous United States"

This dataset contains the input and compiled data supporting the analysis in [Yating Li et al., Year, Journal]. The study models river metabolism (gross primary production, ecosystem respiration, and net ecosystem production) across river monitoring sites in the continental US, and evaluates the potential influence of enhanced weathering on river metabolism using interpretable random forest models with SHAP and partial dependence analysis. Contents: River metabolism model outputs and drivers (GPP, ER; depth, light, discharge, water temperature), derived from Appling et al. (2018) - USGS water chemistry data (pH, alkalinity, total nitrogen, total phosphorus) - USGS fertilizer and manure nutrient loading data, used to compute a nutrient input index for each site's upstream watershed - NLCD land cover classification and tree canopy cover data, summarized per site - Daily discharge statistics (L-moments: mean, L-CV, L-skewness, L-kurtosis), computed per site - Stream order, upstream drainage area, and river width, derived from Maavara et al. (2025) - The final compiled, model-ready dataset used to train the random forest models The code used to process this data and reproduce the analysis is available at: https://github.com/YatingLi0616/River_metabolism_project See the README in the linked repository for the full data processing pipeline, including the expected directory structure for the files in this archive.