Identification of Satellite Indicators for Predicting CyanoHABs in Kansas

Abstract

Cyanobacterial Harmful Algal Blooms (HABs) degrade water quality by producing harmful toxins and causing significant diel changes in water column pH and dissolved oxygen concentrations, leading to degradation of water quality that disrupts food webs, and has negative ecological, social, recreational, and economic impacts. In Kansas, the ubiquity and impact of HABs necessitates large-scale monitoring and prediction of the events, however, no clear indicator or predictor for HABs has been established. The goal of this study is to determine if surface observations from satellite retrievals contain information about the development of HAB events and if this information can be assimilated into a 1-D lake model to improve the prediction of HAB events in Kansas. Different environmental variables are explored to determine their suitability as predictors or indicators of HABs, and with identified candidates, nonlinear regression and regression tree models are created at Cheney Reservoirs with cyanobacteria data. To evaluate their transferability, they are then tested with sediment core pigment data primarily at Marion Reservoir. These results showed that MODIS land surface temperature satellite data paired with NLDAS precipitation, windspeed, and shortwave radiation gave the most promising results for application at both Cheney and Marion. These results can be used for future assimilation into lake models to help with better prediction and modeling of blooms.