Enhancing Land Surface Temperature Reconstruction: An Improved Interpolation of Mean Anomalies Based on the Digital Elevation Model (DEM-IMA)

Surface temperature is a key parameter in scientific studies, encompassing areas, such as resource environment, climate change, and terrestrial ecosystems. The moderate-resolution imaging spectroradiometer land surface temperature (MODIS LST) products play a critical role in research related to land surface temperature (LST). However, these products are often plagued with data loss or distortion, attributable to atmospheric conditions or technical impediments. Unfortunately, there is a shortage of fast LST reconstruction methods that consider both the temporal relationships between close images and the LST variation characteristics in complex, heterogeneous terrain. To address this, the present study proposed a novel method, the improved interpolation of the mean anomalies based on the digital elevation model (DEM-IMA), seeking to fill in missing temperature values. The model suggested in this research was evaluated by comparing it to conventional methods, such as interpolation of mean anomalies (IMA) and gap fill (GF), using a combination of simulation data and actual satellite data. The results suggest that the DEM-IMA model enhances LST reconstruction, particularly for heterogeneous landscapes. The approach effectively restored missing data, displaying a remarkable level of accuracy overall. It surpassed both the IMA and GF methods in the task of filling small, medium, and large cloud gaps in day and night LST data. It reduced the root-mean-square error by 17%, with its accuracy higher at night than during the day. The findings of this study have the potential to provide valuable technical support for enhancing the utilization of MODIS LST products and for conducting quantitative analysis and assessment of regional climate resources with greater effectiveness.