Optimized AVHRR land surface temperature downscaling method for local scale observations: case study for the coastal area of the Gulf of Gdańsk

Abstract

Satellite imaging systems have known limitations regarding their spatial and temporal resolution. The approaches based on subpixel mapping of the Earth's environment, which rely on combining the data retrieved from sensors of higher temporal and lower spatial resolution with the data characterized by lower temporal but higher spatial resolution, are of considerable interest. The paper presents the downscaling process of the land surface temperature (LST) derived from low resolution imagery acquired by the Advanced Very High Resolution Radiometer (AVHRR), using the inverse technique. The effective emissivity derived from another data source is used as a quantity describing thermal properties of the terrain in higher resolution, and allows the downsampling of low spatial resolution LST images. The authors propose an optimized downscaling method formulated as the inverse problem and show that the proposed approach yields better results than the use of other downsampling methods. The proposed method aims to find estimation of high spatial resolution LST data by minimizing the global error of the downscaling. In particular, for the investigated region of the Gulf of Gdańsk, the RMSE between the AVHRR image downscaled by the proposed method and the Landsat 8 LST reference image was 2.255 ∘ C with correlation coefficient R equal to 0.828 and Bias = 0.557 ∘ C. For comparison, using the PBIM method, it was obtained RMSE = 2.832 ∘ C, R = 0.775 and Bias = 0.997 ∘ C for the same satellite scene. It also has been shown that the obtained results are also good in local scale and can be used for areas much smaller than the entire satellite imagery scene, depicting diverse biophysical conditions. Specifically, for the analyzed set of small sub-datasets of the whole scene, the obtained RSME between the downscaled and reference image was smaller, by approx. 0.53 ∘ C on average, in the case of applying the proposed method than in the case of using the PBIM method.

Spatial-temporal evolution of vegetation coverage and its relationship with terrain and human factors in the upper reaches of Ganjiang River Basin, China

“…The measurement methods for FVC include ground surveys and remote sensing estimation. The latter is suitable for large-scale and long-term FVC assessments and has been widely used (Dardel et al, 2014;Chybicki and Łubniewski, 2017;Ranjan and Gorai, 2022). Vegetation indexes, such as normalized difference vegetation index ((NDVI) (Ivanov et al, 2008;Zhan et al, 2021) and enhanced vegetation index (EVI) (Zeng et al, 2022), are measure of surface reflectance that provides a quantitative estimation of vegetation growth.…”