Answers are grounded in this paper and cite their sources.
Satellite remote sensing is critical for lake monitoring but is often hindered by data gaps arising from clouds and sensor malfunctions. To explore the feasibility of current gap filling solutions for inland lakes, this article took Lake Dianchi (2020–2023) in southwestern China as a representative case and comprehensively evaluated the performance of the data interpolating empirical orthogonal functions (DINEOF) algorithm for reconstructing missing remote sensing observations under complex weather conditions. The impacts of missing data proportion, gap size, spectral waveband, and spatial resolutions on the algorithm were assessed using medium-to-low resolution datasets (MODIS and Ocean and Land Color Instrument; 500 m) and medium-to-high resolution datasets (OLI and MSI; 30 m). The results showed that DINEOF remained reliable for medium-to-low resolution data when the missing proportion was below 70% (R2 > 0.65, root mean square error <0.0013 sr−1), beyond which the reconstruction accuracy declined sharply. Performance is more sensitive to the overall missing proportion than to individual gap size, and their combined effects lead to nonlinear degradation. Despite reduced accuracy under high missing conditions, DINEOF outperformed a conventional Kriging–Savitzky–Golay approach, which even yielded negative R2 values. Clear spectral dependence is observed, with better performance in the red and green bands than in the blue band. In contrast, DINEOF performs poorly for medium-to-high resolution data, showing lower accuracy and pronounced patch-like artifacts. Overall, DINEOF is effective for reconstructing medium-to-low resolution lake data within a threshold of missingness, but more advanced methods are needed for higher resolution data in complex aquatic environments.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: