A model for determining satellite-derived primary productivity estimates for Lake Michigan

A new MODIS based satellite algorithm to estimate primary production (PP) has been generated and evaluated for Lake Michigan. The Great Lakes Primary Productivity Model (GLPPM) is based on previous models that required extensive in situ data but it can utilize remotely sensed observations as input for some model variables and therefore allows greater spatial resolution for primary productivity estimates. The Color Producing Agent Algorithm (CPA-A) is utilized to obtain robust chlorophyll a values and the NASA KD2M approach is used to obtain the diffuse attenuation coefficient (Kd). Only incident PAR and carbon fixation rates are additionally needed to generate the primary productivity estimate. Comparisons of the satellite derived PP estimates from single monthly images to average monthly field measurements made by NOAA/GLERL found good agreement between estimates. Satellite derived PP estimates were used to estimate a preliminary Lake Michigan annual primary production of 8.5 Tg C/year. The new algorithm can be easily adapted to work on all the Great Lakes and therefore can be used to generate time series dating back to late 1997 (launch of SeaWiFs). These time series can contribute to improved assessment of Great Lakes primary productivity changes as a result of biological events, such as Dreissenid mussel invasions, climatic change and anthropogenic forcing