Relation Between Bulk Temperature Based on in Situ Temperature Profiles and NOAA-AVHRR Sea Surface Temperature Data for Lake Constance

Abstract

The determination of phytoplankton primary production is one key application of remote sensing data in Lake Constance. One of the main influencing factors of the maximum photosythetic rate in waters is the bulk temperature. From satellite data like NOAA-AVHRR, the skin temperature can be retrieved in a high temporal resolution. However, as parameter for primary production, the bulk temperature in the mixed water layer is needed. In this study, an extensive multi-temporal set of AVHRR data is used to set up a relationship between lake skin temperature from satellite data and the bulk temperature from in situ temperature profiles. Temperature profiles exist for a fixed station in Lake Constance in 20 min repetition over the last twelve years. From about 200 cloudless AVHRR-scenes over the last six years, the skin temperature was derived using the split window technique. Different approaches of this technique are being intercompared for their accuracy for Lake Constance. In a future step, the bulk temperature in 2 to 4 m and the depth of the mixed layer will be determined for input in a primary production model. Multiple regression analysis as well as a neural network approach will be used for the prediction of the bulk temperature from the skin temperature for Lake Constance.