Generation of All-Weather Modis-Like Land Surface Temperature Based On Data Fusion Method

Currently, the most commonly used technique for obtaining land surface temperature (LST) is thermal infrared remote sensing (RS). However, this technique is affected by clouds and cannot obtain complete spatiotemporal LST. To solve this problem, this paper has generated a data fusion model. First, the physical model (Weather Research and Forecast Model) was used to generate LST source data. Then, combined with multisource RS data, a data-driven method (random forest) was used to improve the accuracy of the LST. Finally, all-weather MODIS-like LST with a spatial resolution of 1 km were generated. The results showed that the mean absolute error (MAE), root mean square error (RMSE) and correlation coefficient (ρ) in the case of more (fewer) clouds were ranked as follows: MAE < 1 K (< 2 K), RMSE < 2 K (< 2 K) and ρ > 0.9. The data fusion model can generate high precision all-weather LST data.