Chlorophyll-a is a key indicator for assessing lake water quality, but traditional point-based water quality monitoring has limitations. Remote sensing technology can provide large-scale and dynamic water quality monitoring. Therefore, utilizing multispectral remote sensing technology and considering the spectral band characteristics of the Sentinel-2 satellite, three-band and four-band chlorophyll-a concentration inversion models were constructed based on ground-measured data. The leave-one-out cross-validation method was adopted to validate the model accuracy. The optimal model was applied to monitor the chlorophyll-a concentration in Chaohu Lake from 2019 to 2023. GIS spatial analysis and statistical methods were combined to systematically analyze the spatiotemporal variation characteristics of chlorophyll-a concentration in Chaohu Lake. The results showed that the three-band model based on b₄/b₅/b₇ achieved the highest inversion accuracy. For the Sentinel-2A and Sentinel-2B three-band models, the determination coefficients (R²) were 0.763 and 0.766, the mean relative errors (M_MRE) were 16.87% and 16.66%, and the root mean square errors (R_RMSE) were 4.26 μg/L and 4.22 μg/L, respectively. Spatiotemporal analysis revealed that over the past five years, the chlorophyll-a concentration in Chaohu Lake exhibited significant spatial heterogeneity characterized by higher levels in the west and lower levels in the east, and seasonal fluctuations characterized by higher concentrations in summer and autumn and lower concentrations in spring and winter. An overall decreasing interannual trend was observed, and the inversion results were consistent with the water quality monitoring data for Chaohu Lake. The improved chlorophyll-a concentration inversion model established in this study demonstrates good accuracy and strong applicability. The proposed research method is feasible and effective, providing robust technical support for the monitoring, evaluation, and management of inland Type II water bodies such as Chaohu Lake using remote sensing.