Chlorophyll-a Detection for 317 Maine Freshwater Lakes Through the Use of Archived Data from LANDSAT Satellite Imagery, 23,000 in situ Samples, and Google Earth Engine
Abstract
Chlorophyll-a, the preponderant photosynthetic pigment found in cyanobacteria and eukaryotic life, is a key measurement used to classify the relative productivity, health, and clarity of freshwater lakes and to document long-term trends in lake water quality. Strong relationships between chlorophyll-a and surface reflectance have been used in the remote sensing community to provide semi-autonomous measurements of lake water quality for many years, yet few studies have fully utilized the entirety of the satellite archive in conjunction with the annals of in situ lake observations. To address this challenge, we used Google Earth Engine to match roughly 23,000 water samples assembled by the Maine Department of Environmental Protection over the past 30 years with satellite imagery from Landsat 5, 7 ETM+, and 8, in order to validate algorithms from the scientific literature that have been used to estimate chlorophyll-a concentrations. Of the roughly 23,000 chlorophyll-a samples across 317 lakes taken by observers dating back to 1974, roughly 7% were taken within two days of a satellite flyover in clear conditions, resulting in a total validation data set of over 1,800 samples. We document the results from linear regressions between in situ samples and remotely sensed reflectance data to assess the performance of each algorithm, examine the benefits of three atmospheric correction techniques on the data set, and also address algorithm performance for each sensor. The results of this study mark one of the largest validation data sets to date for chlorophyll-a detection using remotely sensed methods and provide the research community with a framework for constructing long-term multi-lake spatially extensive water quality analyses.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.H31K2076P
- Keywords:
-
- 1819 Geographic Information Systems (GIS);
- HYDROLOGYDE: 1855 Remote sensing;
- HYDROLOGYDE: 1856 River channels;
- HYDROLOGYDE: 1857 Reservoirs (surface);
- HYDROLOGY