Remote Sensing Inversion Of Chlorophyll A Concentration In Karst Plateau Lakes ——A Case Study Of Pingzhai Reservoir

Chlorophyll-a（Chla）is not only an important index to measure the primary productivity and eutrophication degree of water,but also an important factor to maintain the health of reservoir ecosystem and affect the apparent reflectance spectrum of water.Pingzhai Reservoir is the source reservoir of the central Guizhou Water Diversion Project and an important water source in the central Guizhou region of Guizhou Province.It is crucial to ensure the ecological security of the middle and lower reaches of Guizhou province and the safety of drinking water supply in the central Guizhou region.With the continuous and rapid economic development,the natural ecological environment in this area has been under great pressure,especially the direct discharge of untreated industrial and agricultural waste water,which has also caused serious pollution to Pingzhai Reservoir,and the water quality is worrying.In this paper,Pingzhai Reservoir is taken as the research object.Three times of water surface spectral measurement and synchronous water quality sampling are carried out in August 2020,March 2021 and July 2021.Remote sensing reflectance of water body is normalized.The inversion model of Chlorophyll-a concentration was carefully analyzed from the aspects of verification of atmospheric correction results,selection of sensitive window and sensitive band,and analysis of model accuracy.Based on Sentinel 2 MSI images from 2018 to 2021,a Chla estimation model of Pingzhai Reservoir was constructed.The variation of water Chlorophyll-a concentration and the influence of different environmental factors on its spatial differentiation were analyzed,and the effects of different remote sensing models and the model constructed in this paper on the estimation accuracy of Chlorophyll-a concentration were verified.It provides a powerful technical means for the observation of Chlorophyll-a concentration in Pingzhai Reservoir.The main conclusions are as follows:（1）The comparison between the measured remote sensing reflectance and Sen2cor atmospheric corrected remote sensing reflectance shows that Sentinel 2 MSI images corrected by Sen2cor can be used to estimate Chla concentration in the study area.Based on Sentinel 2 MSI data and measured spectral data,the concentration of Chla in Pingzhai Reservoir was inverted.In sentinel 2 MSI single-band and band combination algorithm,the binomial function model based on B8×（B7-B5）of Sentinel 2 MSI achieved acceptable to high fitting accuracies,which explained 89%of the variation of Chla.The determination coefficient R~2 was 0.849,and the root mean square error（RMSE）of the verification result was 6.50%,indicating that the established model could better estimate the concentration of Chla in Pingzhai Reservoir.（2）The Chlorophyll-a of Sentinel 2 MSI images showed a clear spatial distribution pattern.In space,the Chla in Nayong River and Dam area were higher than that in other areas,followed by Zhangwei and Baishui Rivers,and Shuigong and Hujia Rivers that were relatively low.Considering time,it is high during autumn and low during winter,and the maximum value appeared during August.TN、NH₄⁺-N、DO and are important environmental factors affecting the spatial variation of Chla.During summer,autumn and winter,the contribution of each factor from large to small was Nayong River,Dam and other regions,and more significant during autumn.The correlation was not obvious in spring.In addition,Chla spatio-temporal variations are also effected by human activities.（3）The Chla concentration inversion model constructed in this paper has good accuracy in the estimation of Chla in Pingzhai Reservoir in different periods by using the Chla of long-term monitoring points for verification.Moreover,the accuracy of Chla inversion model constructed in this paper is better than the estimation ability of Chla in Pingzhai Reservoir based on different Sentinel 2 MSI remote sensing inversion models.Meanwhile,the inversion effect of the model in the estimation of Chla concentration in Hongfeng Lake and Baihua Lake is also good.