Remote sensing and in-situ data to support chlorophyll-a mapping in Blue Mesa Reservoir from Sentinel-2 imagery

This data set includes estimates of aquatic chlorophyll a concentration and reservoir temperature for Blue Mesa Reservoir, CO. A Random Forest modeling approach was trained to model near-surface aquatic chlorophyll a using near-coincident Sentinel-2 satellite imagery and water samples analyzed for c...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tyler V King, Robert A Bean, Evan J Gohring
Format: Dataset
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This data set includes estimates of aquatic chlorophyll a concentration and reservoir temperature for Blue Mesa Reservoir, CO. A Random Forest modeling approach was trained to model near-surface aquatic chlorophyll a using near-coincident Sentinel-2 satellite imagery and water samples analyzed for chlorophyll a concentration. The trained chlorophyll a model was applied to Sentinel-2 imagery to produce maps of modeled chlorophyll a concentrations at 10 m spatial resolution for May through October for 2016 through 2023. Chlorophyll a concentrations for three sections (basins) of Blue Mesa Reservoir were extracted from the raster data to produce time-series of modeled chlorophyll a concentration summary statistics (e.g., median, standard deviation, 90th percentile, etc). Water temperatures were approximated using the provisional Landsat surface temperature (PST) product collected with sensors on board Landsat 5, 7, 8, and 9 for May through October between 2000 and 2023. PST values for Landsat 8 and Landsat 9 were scaled to match in-situ water temperature observations in the top 1 m of the water column using a multivariate linear regression model. A harmonized water temperature record was produced by adjusting Landsat 7 PST values to align with the adjusted Landsat 8 values for images collected within one day of each other. Similarly, Landsat 5 PST values were adjusted to match the adjusted Landsat 7 values. The modeled chlorophyll a and temperatures had root mean square errors of 1.9 micrograms per liter and 0.7 degrees Celsius, respectively. This data release includes three components with tabular and raster data: 1) Tabular .csv format in-situ and remotely sensed chlorophyll a data from Blue Mesa Reservoir, Colorado May through October 2016 - 2023: Data used to train the chlorophyll a model (chl_model_training.csv), and modeled chlorophyll a time series (chl_rs_values.csv) 2) Raster format remotely sensed aquatic chlorophyll a for Blue Mesa Reservoir, Colorado May through October 2016 - 2023: Raster data include 167 geotiffs of modeled chlorophyll a concentrations in zipped directory (chl_conc_ug_L.zip) 3) Tabular in-situ and remotely sensed temperature data from Blue Mesa Reservoir, Colorado May through October 2000 - 2023: Data used to train the temperature model (temp_model_training.csv) and modeled temperature timeseries (temp_rs_values.csv)
DOI:10.5066/p9xulqas