Process-oriented estimation of column-integrated algal biomass in eutrophic lakes by MODIS/Aqua

•A process-oriented algorithm was proposed to remotely derive algal biomass.•The nonuniform Chl-a profile could be described using a power decay function.•Temperature and NH4+-N were the main factors changing algal biomass temporally.•Algal biomass served as a better indicator for remotely assessing water quality.•A lookup table was generated for forecasting algal bloom probability. Algal blooms happen widely in eutrophic lakes. The satellite-derived surface algal bloom area and chlorophyll-a (Chl-a) concentration are commonly used as indicators to judge the water quality. However, vertical migration of phytoplankton may seriously affect the assessment accuracy of water quality with the surface status as an indicator. In fact, the water column-integrated algal biomass is better for describing the water quality. To remotely estimate the algal biomass considering the vertical Chl-a profile (uniform or otherwise), this paper proposed a novel process-oriented algorithm applied successfully in Lake Chaohu, China. First, we built a decision tree to identify the pixel-based Chl-a profile types using the floating algae index, surface Chl-a concentration, and wind speed. Then, the nonuniform profile was expressed as a power decay function (Chl-a(z)=n1×zn2) and parameterized using the surface Chl-a concentration. Finally, the algal biomass was derived from MODIS/Aqua satellite data. The estimation results were acceptable and the bias was −19.95%. The monthly mean algal biomass varied significantly and exhibited two peaks in spring and summer. The annual mean algal biomass increased from 2003 to 2010 (Pearson r = 0.82; p