Application Of Remote Sensing Monitoring In Water Quality And Landuse Of Upstream Huangpu River

Upstream Huangpu River including Dianshan Lake is the water supply source of Shanghai, the water quality is very important for people's health and the city's economic growth. With the rapid growth of economy and industrialization, water pollution problem like eutrophication are getting more and more serious, which requires frequent monitoring and integrated management. However, conventional method of water quality monitoring is time-consumed and money-consumed with only limited water quality data provided at the measurement sites. In addition, it requires reasonable sample collection and transportation. Remote sensing (RS) monitoring can not only save manpower and material resource but also realize the development of water quality monitoring from point to big area.The author firstly analyzed correlation between water quality parameters and hyper-spectral reflectance, based on which water quality models of Dianshan Lake and Huangpu River were established. For Dianshan Lake, chlorophyll-a (chla) model was established by R^*₇₀₈ and R^*₆₇₇ with R² of 0.6979, and TN model was established by R^*₆₅₅ with R² of 0.5112. For Huangpu River, models for chla, turbidity and COD was accurate. Chla model was linear model by R₇₃₈/R₆₇₅ with R² of 0.8515. Turbidity model was linear model by R^*₅₂₀ and R^*₇₂₈ with R² of 0.5146. COD model was established by R₇₆₃/R₆₇₅ with R² of 0.6187. The models used for retrieve water quality of different time, the result show that they can accurately retrieve water quality except for some samples.Multispectral models for Dianshan Lake using TM data were then estblished. Both chla model and TP model were linearly regressed by TM1^*TM4 and TM2^*TM4 with R² of 0.346 and 0.754, respectively. Turbidity model was established by (TM3-TM1)/ (TM4-TM3) and TM1 with R² of 0.553. COD model was established by (TM3-TM2)/ (TM3-TM1) and TM1 with R² of 0.410. TN model was established by (TM3-TM2)/ (TM4-TM1) and TM1 with R² of 0.370. Based on the above models, spatial distribution of water quality parameters in March 3th of 2006, March 21th of...