Answers are grounded in this paper and cite their sources.
Reconstructing high spatiotemporal resolution water quality products is fundamentally constrained by the tradeoffs between satellite spatial and temporal resolutions. To address this limitation, we propose the spatial and temporal linear prediction and coarse-resolution constraint fusion method (STLCCFM). This product-level fusion framework constructs a fine-resolution temporal prediction from available Landsat retrievals and subsequently constrains the overall magnitude using target-date moderate-resolution imaging spectroradiometer retrievals, with Gaussian residual redistribution applied. Unlike conventional spatiotemporal fusion (STF) methods, STLCCFM decouples temporal reconstruction from the need for strictly paired coarse-fine reference images, enabling the use of temporally sparse or spatially incomplete inputs. Validation using water transparency (Secchi disk depth, ZSD) products in Honghu Lake and Poyang Lake showed that STLCCFM achieved stronger agreement with the target-date Landsat-derived references than the compared benchmark methods. Additional experiments, including extended temporal intervals, single-sided fine-resolution inputs, and controlled geometric missing scenarios, were conducted to examine the contributions of individual components and the sensitivity of the framework to reduced fine-resolution data availability. The Taihu Lake chlorophyll-a experiment provided preliminary evidence that STLCCFM can be extended beyond ZSD retrievals. Although the fusion performance remains dependent on the quality and consistency of the input retrievals, STLCCFM provides a flexible framework for reconstructing asynchronously acquired satellite water-quality products at high spatiotemporal resolution.
No takes yet. Share an insight, caveat, or question.
Luo et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: