A spatial‒temporal seamless reconstruction method for high-resolution 16 m daily land surface albedo products

Accurate reconstruction of fine-resolution, spatially temporally continuous land surface albedo is critical for energy balance studies and climate change modeling. This paper proposes a dual-constrained optimal image matching strategy (DC-OPS) for reconstructing seamless 16 m daily albedo by integrating 500 m MODIS and 16 m GF-1 WFV products through the flexible spatiotemporal data fusion (FSDAF) model. The strategy combines temporal proximity and texture similarity to identify optimal reference image pairs, enhancing reconstruction accuracy. Validated in the Huailai region and the Tibetan Plateau (TP), the results show high accuracy. In Huailai, DC-OPS yielded an RMSE of 0.0304 against original GF-1 images and 0.0282 against in situ measurements. In the TP region, the method maintained a reliable RMSE of 0.0377 despite extreme data scarcity. Comparative analysis demonstrated that DC-OPS outperforms traditional methods such as harmonic analysis of time series (HANTS) and neighborhood similar pixel interpolator (NSPI), especially in capturing rapid surface changes such as snowfall. The strategy exhibits robust applicability across diverse landscapes, providing a foundation for large-scale applications by enhancing spatial temporal continuity.