Application of Gon's algorithm to monitor productivity using Sentinel-2 Multi-spectral Instrument in Lake Kyoga, Uganda

Lake Kyoga is vulnerable to anthropogenic activities and climatic drivers. It requires regular monitoring due to the potential threat of algal bloom and cyanobacteria. Chlorophyll-a is a proxy parameter that shows the lake's trophic state, hence primary productivity. The objective of this study was to compare Gon's method and ocean color algorithms and select suitable remote sensing techniques to monitor chl-a concentration in water bodies. Google Earth Engine along with Sentinel-2 Multi-Spectral Instrument (MSI) imagery were used to retrieve satellite data. In situ, chl-a data for model calibration and validation was collected from 29 sampling locations in the lake according to the seasons. We calibrated Gon's algorithm and then used Sentinel-2 MSI imagery to produce chl-a maps. We noticed a major difference between the suitability of ocean color and Gons algorithms in estimating chl-a in Lake Kyoga. Ocean color was less accurate because it assumes that the blue-green band ratio linearly responded to chl-a abundance, and this made it suitable for relatively clear waters such as oceans or oligotrophic lakes, not Lake Kyoga. The Gons algorithm was more applicable in turbid productive waters such as Lake Kyoga and estimated chl-a accurately. These results are very important for real-time monitoring of climate-sensitive lakes such as Lake Kyoga, which are also subjected to severe catchment degradation.