David Warner Peter C. Esselman Brett Hobson Brian Kieft David Bennion 2019 <p>Accurate methods to track changes in lake productivity through time and space are critical to fisheries management. Chlorophyll <i>a</i> is the most widely studied proxy for ecosystem primary production, and has been the topic of many studies. The main sources of chlorophyll <i>a</i> measurements are ship-based measures or multi-spectral satellite data. Autonomous underwater vehicles can survey large spatial extents approaching the scale of satellite data, but with the accuracy of ship-based water sampling methods. We use several statistical measures to compare measures of chlorophyll <i>a</i> collected in Lake Michigan with spatiotemporally matched satellite-derived measures of chlorophyll <i>a</i> from the MODIS Aqua multi-spectral sensor using NASA’s OC3 and the Great Lakes Fit algorithms. Our findings show a near one to one relationship between AUV data and both satellite-derived data sets when the AUV data are coarsened to the resolution of the satellite data. A comparison of satellite-based chlorophyll <i>a</i> to AUV-derived chlorophyll summarized in discrete water depth bins suggested that, based on decreasing coefficients of determination, satellite estimates of chlorophyll accounted for the most variability in chlorophyll <i>a</i> concentrations in the upper 10 m of the water column, even though satellite sensors may detect past this depth.</p> application/pdf 10.1016/j.jglr.2019.04.003 en Elsevier A comparison of chlorophyll a values obtained from an autonomous underwater vehicle to satellite-based measures for Lake Michigan article