Dataset Supporting: Predicting Harmful Algal Bloom Risk in a Small Inland Reservoir Using Integrated Sentinel-2 and In Situ Data

Description

This repository contains the processed dataset supporting the study, "Predicting Harmful Algal Bloom Risk in Barr Lake, Colorado Using Integrated Sentinel-2 and In Situ Data (2019–2025)." The dataset integrates in situ chlorophyll-a measurements, Sentinel-2-derived Normalized Difference Chlorophyll Index (NDCI) values, environmental variables, ecological zone classifications, and satellite–field matchup metadata collected from Barr Lake, Colorado, USA, between 2019 and 2025. Each observation represents a field sampling event matched to the closest valid Sentinel-2 Level-2A image acquisition within a ±5-day temporal window following quality-control procedures. The repository includes the processed dataset, a README file describing data collection and processing procedures, and a data dictionary containing variable definitions, units, and coding information. The dataset supports research on harmful algal bloom monitoring, satellite-assisted water-quality assessment, and probabilistic HAB-risk modeling in inland reservoirs.

Steps to reproduce

1. In situ chlorophyll-a and environmental measurements were collected from Barr Lake, Colorado, USA, during bloom-season monitoring campaigns (July–September) between 2019 and 2025. 2. Chlorophyll-a concentrations were determined following U.S. EPA Method 445.0 using fluorometric analysis. 3. Sentinel-2 Level-2A surface reflectance imagery (COPERNICUS/S2_SR_HARMONIZED) was accessed through Google Earth Engine. 4. Cloud-contaminated observations were excluded using the Sentinel-2 Scene Classification Layer (SCL), and valid water pixels were identified following quality-control procedures. 5. The Normalized Difference Chlorophyll Index (NDCI) was calculated from Sentinel-2 Band 5 (705 nm) and Band 4 (665 nm) reflectance values. 6. Field observations were matched to the closest valid Sentinel-2 acquisition within a ±5-day temporal window. 7. A binary harmful algal bloom exceedance variable (HAB50) was generated using a chlorophyll-a threshold of 50 µg L⁻¹. 8. The processed dataset was used for probabilistic harmful algal bloom risk modeling and statistical analyses described in the associated manuscript.

Institutions

• University of Colorado Boulder

  Colorado, Boulder
• Colorado Parks and Wildlife

  Colorado, Denver

Categories

Licence