Bridging Cloud-Induced Gaps in MODIS Leaf Area Index Products Using High-Frequency Geostationary Satellite Observations

Leaf Area Index (LAI) is a fundamental variable for characterizing vegetation canopy structure and ecosystem functioning. However, persistent cloud contamination substantially constrains the spatiotemporal continuity of LAI products derived from polar-orbiting satellite observations. In contrast, geostationary satellites provide dense temporal sampling, offering a valuable observational basis for filling data gaps in polar-orbiting satellite products. To bridge this observational gap, this study develops an observation-complementary framework that integrates high-frequency Himawari-8 Advanced Himawari Imager (AHI) observations with MODIS LAI products to achieve LAI reconstruction in cloud-prone regions. Within this framework, AHI observations are first used to retrieve LAI at a spatial resolution of 2 km using a Random Forest model trained with a high-quality MODIS-based benchmark dataset, and are then downscaled to 500 m using MODIS-derived spatial texture information. The reconstructed AHI LAI is subsequently integrated into low-quality or missing MODIS pixels to generate a fused AHI-MODIS LAI dataset over China. Validation against in-situ measurements and cross-comparison results show that the fused product maintains reasonable numerical consistency and stable temporal trajectories. Data availability increases from 67.60% to 90.20% under the 4-day compositing cycle. These results indicate that geostationary–polar-orbiting observation complementarity provides a practical pathway for improving the continuity and usability of optical LAI products in persistently cloudy regions, while remaining subject to the constraints of optical observations and cross-scale reconstruction assumptions.