Calibrating a Chlorophyll and phycocyanin model in oligotrophic lakes with Sentinel-2 satellite data

This video shows an attempt to use chlorophyll-a (Chl-a) measurements to estimate Chl-a from high definition satellite images of New York’s Finger Lakes. Chl-a, the green photosynthetically active pigment found in phytoplankton, is often used as a proxy for algal biomass, including in cyanobacteria blooms. Established satellite-based methods of estimating Chl-a have spatial resolutions too coarse to be useful in most New York lakes. Conversely, data from the Multi-Spectral Instrument sensor on the Sentinel-2 satellites have 10m resolution and have useful radiometric bands. Given the rapid growth rates of algal blooms, the timing of in-lake data sets with satellite overflights was important in providing high quality data for calibration of satellite images. When paired with in-lake measurements of Chl-a in two Finger Lakes (detailed elsewhere in this conference), a simple empirical relationship can be produced, allowing for future estimation of lakewide Chl-a with 10m resolution. Initial results show surface Chl-a (µg_L) = 11.57·(log10 (Band 3 _ Band 2)) +1.35. The relationship holds for the relatively low Chl-a concentrations in the open water of these oligotrophic lakes, but complications arise from water clarity, bathymetry, macrophyte growth, and phytoplankton community composition. For these reasons, this method is limited to deeper areas of lakes (depth greater than Secchi depth). Other limitations are the 10-day return frequency of each Sentinel-2 satellite, cloud cover, and wind glint. The work could enhance detailed estimates of algal biomass in lakes as small as 1.4 hectares, with implications for water quality assessments and harmful algae bloom studies.