Inversion Of Water SSC And Decision Tree Analysis Of Poyang Lake Based On GF-1 Image

Poyang Lake is China’s largest freshwater lake.Water suspended sediment is one of the important water quality parameters,which is a key factor in the problems such as erosion and sedimentation caused by water environment monitoring and sediment transport.Traditional water quality testing methods require a lot of manpower,material and financial resources,but the detection of water quality by remote sensing can overcome these shortcomings,and the effectiveness and wide performance are natural advantages of the method.High-resolution satellite(GF-1)is China’s first satellite which can observe the earth with high-resolution,the combination of high spatial resolution and high time resolution is the perfect feature of it,the revisit cycle is only 4 days,so in the monitoring of Poyang Lake water quality changes have a very important practical value.In this paper,the author used GF-1 remote sensing data and combined with Poyang Lake hydrological elements to observe and analyze the suspended sediment concentration in Poyang Lake watershed.The Poyang Lake Basin is a meso-scale area where the world’s few administrative boundaries are highly consistent with the natural boundary of the basin.Therefore,it is an important area to study the relationship between "human-to-earth".This study can provide scientific decision-making basis for environmental protection and sustainable economic development in the region.The main contents of this paper are:1.Study of water suspension sedimentation characteristics in the Poyang Lake Basin;2.Study of spectral reflectance and suspended sediment concentration model in the Poyang Lake Basin;3.Study of Remote Sensing Information Processing Method for Suspended Sediment in the Poyang Lake Basin;4.Construction and study of Quantitative inversion model of suspended sediment concentration in the Poyang Lake Basin;5.Decision tree analysis with pixel ratio and slope as constraint condition.Results and conclusions:1.The spectral reflectance curves are characterized by bimodal and valley;2.The single-band model,the band-ratio model and the three-band model constructed by the GF-1 image band data and the ground measured suspended sediment concentration data were used to calculate the peak ratio model and the first-order differential model based on the measured spectral data and ground water quality data.By comparing the correlation coefficient between each model and the measured suspended sediment data,the image ratio model B4/B3 has the largest correlation coefficient with the measured value,reaching 0.8510;3.The statistical regression model is established according to the best inversion model,and the inversion and precision evaluation of the Poyang Lake waters are carried out.The results obtained inversion are consistent with the measured data;4.Based on the ratio of the pixel as the constraint condition,the decision tree model is established to classify the decision tree,and the results are compared and analyzed.The results show that the decision tree classification is effective.