CPMF: An Integrated Technology for Generating 30-m, All-Weather Land Surface Temperature by Coupling Physical Model, Machine Learning, and Spatiotemporal Fusion Model

Although thermal remote sensing is the optimal method to measure large-scale land surface temperature (LST), its application has been severely constrained due to cloud contamination and the tradeoff between temporal and spatial resolutions. The integrated technology of LST gap filling and downscaling is an effective method to break through these limitations. In this study, we proposed an integrated technology of gap filling and downscaling to generate daily 30-m all-weather LST by coupling a physical model, machine learning (ML), and spatiotemporal fusion model, termed CPMF. CPMF comprises three modules: 1) estimating 1-km LST based on the surface energy balance theory (SEB-LST1 km); 2) generating spatially complete 1-km LST coupling ML (CRLST1 km); and 3) all-weather 30-m LST from the CRLST1 km combining the spatiotemporal fusion downscaling and ML downscaling in an equal-weighted manner (CPMF-LST30 m). Then, satellite data, reanalysis data, airborne data, and in situ LST data were used to evaluate the CPMF’s performance. Results showed that: 1) SEB-LST1 km correlates well with clear-sky MODIS-LST (mean Pearson’s ${R}~\approx ~0.70$ and mean RMSE $\approx ~3.62$ K); 2) CRLST1 km has a high correlation with MODIS-LST and reanalysis-LST, outperforming other four existing gap-filling products; 3) CPMF-LST30 m achieves good accuracy, with Pearson’s R of 0.86–0.96 (RMSE <3.40> ${R} \,\, =0.66$ ( ${P} \,\, \lt 0.01$ ) with airborne LST, and ${R} \,\, =0.97$ (RMSE =4.25 K) with in situ LST, surpassing single-method downscaling; and 4) sensitivity analysis highlighted the importance of SEB-LST and CRLST in ML models, confirming the efficacy of the proposed physical model. CPMF provides a one-stop service for producing high-quality, long-term, all-weather LST data at a 30-m resolution.