Validating MODIS Land Surface Temperatures Using In-situ Skin Temperature Data across Greenland

Given the current and future potential for sea level rise caused by increasing ice melt from the Greenland Ice Sheet (GIS), accurate land surface temperature (LST) measurements from the GIS are invaluable to climate change research. LST data is used to model and predict a multitude of polar and global geophysical processes, such as surface energy balance and ice mass loss. Remotely-sensed LST measurements, such as those calculated using the Moderate Resolution Imaging Spectroradiometer (MODIS), provide information from the vast polar regions where ground-based measurements are few and difficult to obtain. This remotely sensed LST data must be validated using ground-based measurements in order to ensure the accuracy of the MODIS LST algorithms and their applications. Previous validation projects have found a cold bias in the MODIS LST data over Greenland when compared to in-situ near surface air temperature measurements. However, due to the presence of near surface inversions, MODIS LST data should be validated by ice surface (skin) temperature measurements when possible. To identify if a cold bias is present, we compare LST data from MODIS and skin temperature measurements collected in 2015 by automatic weather stations operated by the Programme for Monitoring of the Greenland Ice Sheet (PROMICE) at 20 sites in the ablation zone of the GIS. Our initial results aggregated from across the ice sheet indicate the presence of a mean cold bias of 2.1°C ± 5.2°C (uncertainty reported is one standard deviation) in the MODIS LST data, and a root mean squared error of 5.6°C. Using specific humidity data from PROMICE and solar zenith angle and LST data from MODIS, we find that this bias is more pronounced at lower temperatures and higher humidity. We plan to expand our analysis to investigate geographic factors that control the MODIS bias across the GIS, such as elevation and latitude. These results can inform future polar temperature studies and refinement of the MODIS LST algorithms, particularly in low temperature and high humidity cases where we find the largest biases.