Deep Quantitative-inversion Study Of Qinghai Lake Based On Remote Sensing Technology

Global climate change is a hot issue in the government and the scientific community as well as the general public. The Qinghai Tibetan Plateau is an important component, driving factor and magnifier of the global climate change due to its special terrain condition. In recent years, the Qinghai Tibet Plateau presents the global trend of climate warming, which shows that the air temperature and the earth-surface temperature rise, glaciers are melting, so is frozen soil, the annual regional precipitation increases or decreases and the year precipitation regional characteristic increases or reduces.Climate change and water environmental factors have mutual feedback. Therefore, it is very important to study the water environmental factors and changes in the Qinghai Tibet Plateau.Lake is an important part of water environmental factors, which indirectly affect the climate through the participation of the natural water cycle. At the same time, the lake is very sensitive to the climate change and it is an important information carrier to reveal the global climate change and regional response. Therefore, it is a necessary prerequisite and basis for the study of climate change in the Qinghai Tibet Plateau about how to monitor the water content and its changes in the lake in real time. Previous studies on the lake water volume in the Qinghai Tibet Plateau, mainly focus on the number and area of the lake. Although these statistic figures could reflect the water environmental status and its changes indirectly, it also can indirectly indicate the climate change of the Qinghai Tibet Plateau, but it can not accurately and quantitatively determine the amount of water environmental change. How much volumes of water storage is in the Plateau lake? How to accurately and quickly measure the water depth and water storage capacity of large and super large lakes? How much is the the change with the water volume? How does it evolve in time and space scale? This science and technology problem is especially important and urgent in the background of global climate change.Under this background and based on the theory of remote sensing color, we adopt research measures including LandSat-8 OLI images as the data source, remote sensing technology and quantitative methods to build the model of Qinghai Lake deep quantitative inversion which is based on the remote sensing technology and it uses Qinghai Lake as the research object and combines ultrasonic instrument to measure water depth in order to provide scientific basis and technology support for depth measurement of large lakes plateau and provide a new perspective for research and evaluation for the Tibetan Plateau climate change. The main work and conclusions and achievements include:1. On September 17 th and 18 th 2015, we got 21 sampling points of the 63 water depth data and 150 water 15 hyperspectral data acquisition point by the way of renting a boat from Qinghai Lake administration and using SQ-SFCC hand-held ultrasonic sounder、ASD Fieldspec 4 Portable spectrum instrument to measure the depth of Qinghai lake and collect water hyperspectral. Regressing the sampling depth data points and Multi band model, Green / Blue ratio model, Red / Blue ratio model, Red / green ratio model, Blue logarithmic model, Green logarithm model, the calculation results of albedo-independent Bathymetry algorithm,(bottom albedo-independent Bathymetry algorithm, called "B algorithm"),we establish water depth inversion of remote sensing model. Calculate every model of minimum, maximum, mean, standard deviation,correlation coefficient square R2 and select the better model. The calculation results show that the water depth model based on B algorithm is superior to other models and it is the best model for the inversion of Qinghai lake water depth.2. In order to further improve the accuracy of model estimation, this paper attempts to partition a depth. Partitioning method is based on the B algorithm to calculate the results of the water depth regression model, defining the water depth of 20 meters below as “shallow water” of remote sensing inversion and the water depth of 20 meters above as “deep water” of remote sensing inversion. The partition of water depth model was achieved by actually measuring water depth and the simulation regression analysis of B algorithm. Deep water model: Z=29.71*X-0.17, X is for the B algorithm simulation value, R2=0.95, the average absolute error for the overall is 0.28 m, the overall average relative error is 2.01%. The shallow water model is: Z=11.17*X +0.92, R2=0.53, overall absolute error is 0.82 m, the average relative error is 2.05% overall. The average absolute error model is not more than 1 meter, which has been able to meet the general application and the requirements of scientific research for water depth remote sensing estimation of the largest salty lake in China.3. With the help of the theories and methods of spatial analysis, we revealed depth of the Qinghai Lake more than 25 m through the Qinghai Lake deep DTM analysis. There are 3 deep-space centers respectively deep water area in the northern mountains, deep water area in Haixin mountain and Erlangjian deep water area. Its area is respectively 750, 576, 78km~2, which indicates that the deep water area of Qinghai Lake is mainly distributed in the northern and central and western part of the lake mainly because of being supplied by the Buha River, Shaliu River.In this paper, we first use the combination of the main and passive remote sensing technology to study the depth of the lake in Qinghai Tibet Plateau.