A method to estimate land surface temperature from Meteosat Second Generation data using multi-temporal data

Area of Science:

• Earth and Environmental Sciences
• Remote Sensing
• Atmospheric Science

Background:

• Land surface temperature (LST) is crucial for understanding land-atmosphere interactions.
• Accurate LST retrieval is essential for various environmental and climate studies.
• Existing methods may have limitations in certain conditions.

Purpose of the Study:

• To develop and validate a novel method for inverting LST from multi-temporal thermal infrared remote sensing data.
• To assess the accuracy of the proposed method using simulated and satellite-derived datasets.
• To evaluate the method's performance under clear-sky conditions.

Main Methods:

• The study integrates the split-window algorithm with a diurnal temperature cycle model.
• Validation involved simulated diurnal brightness temperatures from MSG2-SEVIRI using MODTRAN 4.
• Comparison was made with the MOD11B1 V5 LST product.

Main Results:

• The proposed method achieved a Root Mean Square Error (RMSE) of 1.2K for simulated data with accurate inputs.
• Most errors for simulated data were within ± 2K.
• Comparison with MOD11B1 V5 showed an RMSE of 3.0K, with most errors within ± 6K.

Conclusions:

• The developed method provides a viable approach for LST inversion from thermal infrared remote sensing data.
• The method demonstrates good accuracy under clear-sky, mid-latitude, daytime conditions.
• Further validation is recommended for diverse geographical areas and atmospheric conditions, including cloudy scenarios.