Validation of Alpine Lake Surface Temperature Measurements derived from NOAA-AVHRR and MODIS data

Numerous lakes in the European Alps contain a sufficient large volume and areal extent of water to influence significantly local weather patterns with their surface temperature. Up to now there is no area-wide consistent temperature data set for the alpine lakes available. Also, in situ measurements can not take into account the high variability of water in nearshore zones and their limited spatial coverage are not adequate for resolving the spatial distribution of temperature in the lakes. Data derived from imaging sensors such as AVHRR and MODIS can provide an area wide operational lakes surface water temperature (LSWT) information. An operational near real-time processing scheme for NOAA-AVHRR data was developed, from which LSWT measurements based on the MCSST algorithm are derived. For the years 2002 and 2003, day and night data from NOAA-12, -15, - 16 and -17 and from MODIS on board TERRA and AQUA have been compared with in situ data of three different sized lakes in the Alps. The validation revealed a bias around 0.5-1.5 K for both AVHRR and MODIS derived LSWT. As expected, low wind speed and high insolation situations could be identified as the main contributors to the bias. Overall, this study shows the feasibility of the AVHRR LSWT processing scheme and MODIS data for application, such as LSWT climatology and assimilation for numerical weather prediction models, on water bodies of a few square kilometer extent.