A Community Dataset for Large-Scale River Nitrogen Modeling in the United States

Area of Science:

• Environmental Science
• Hydrology
• Data Science

Background:

• Water quantity predictions are advanced by benchmarks like CAMELS.
• Large-scale water quality predictions, particularly for nutrients, are hindered by inadequate datasets.
• Existing datasets suffer from limited human-impacted system representation, missing nutrient inputs, incomplete metadata, and sparse monitoring.

Purpose of the Study:

• To introduce IWAND-Nitrogen (Integrated Watershed Attributes and Nutrient Data for Nitrogen), a comprehensive dataset for the contiguous United States.
• To overcome limitations of existing water quality benchmarks by enhancing spatial and temporal coverage and anthropogenic gradient representation.
• To establish a new benchmark for the nutrient research community, facilitating advancements in nutrient modeling and insights.

Main Methods:

• Integrated 574,767 nitrate records from 1,877 catchments (1980-2023) with high sampling frequency (median 272 samples/gauge).
• Linked water quality data with 93 watershed attributes, eight nitrogen input forcings (basin-averaged and gridded), and eleven climate forcings.
• Ensured catchments had at least 200 measurements for robust analysis.

Main Results:

• Developed IWAND-Nitrogen, a dataset significantly expanding spatial and temporal coverage compared to existing benchmarks.
• Enhanced representation of systems across various anthropogenic gradients.
• Provided a rich resource for studying nutrient dynamics at catchment to national scales.

Conclusions:

• IWAND-Nitrogen addresses critical data gaps in large-scale water quality prediction, especially for nutrients.
• The dataset serves as a valuable benchmark for advancing nutrient modeling and research.
• Facilitates new insights into nutrient processes across diverse environmental settings.