Wind and rainfall regulation of the diffuse attenuation coefficient in large, shallow lakes from long-term MODIS observations using a semianalytical model

摘要

Knowledge of the variation and regulation of the diffuse attenuation coefficient is important for management of lake water quality and ecological environments. Data were collected seven times (378 samples) from 2006 to 2011 in Lake Taihu to improve the existing semianalytical model of diffuse attenuation coefficient at 490 nm (K_d(490)) by measuring in situ K_d(490) and remote sensing reflectance (R_(rs)). This semianalytical model was applied to determine the long-term variation of K_d(490) and its driving factors from Moderate Resolution Imaging Spectroradiometer (MODIS) images. Our semianalytical model of K_d(490) improves the retrieval accuracy of K_d(490) from 29.60 to 22.58 (root-mean-square percentage error) and 24.28 ± 18.54 (mean value ± standard deviation) to 19.29 ± 14.34 (mean absolute relative error). The Kd (490) from maps produced from MODIS images clearly present the temporal and spatial variations of K_d(490). This indicates the suitability of using the improved semianalytical model and MODIS to observe K_d(490) in turbid shallow inland waters. Long-term MODIS observation of K_d(490) shows distinct seasonal and spatial variations. Temporally, the highest and the lowest K_d(490) were found in winter and in summer, respectively. Spatially, high K_d(490) values were primarily distributed in the southwestern and northwestern Lake Taihu, and relatively low K_d(490) values were distributed in Meiliang Bay and Gonghu Bay (the eastern part of the lake and East Bay were excluded in this study due to invalid retrieval results of K_d(490)). Many covarying points were found between K_d(490) and maximum wind speed from long-term observations, suggesting that wind-induced sediment resuspension is the principal driver of high K_d(490). Correlation was also observed between K_d(490) and heavy rainfall events, suggesting the regulating effect of heavy rainfall.