| As the main surface water resources,inland lake reservoirs are one of the main sources of drinking water for human beings,which is related to human survival and development.The change in water storage capacity is the basic indicator of environmental changes caused by climate change and human activities.The elevation and area of water level are important indicators to measure the change of lake storage.Therefore,monitoring the changes of water level and area of lake reservoirs has always been a hot spot for many researchers.With the rapid development of satellite altimetry and remote sensing technology,space-borne laser altimeter and optical camera can obtain the vertical height and area changes of surface targets in a large range and with high efficiency,which provides a large number of observation data for monitoring the water level and area of lake reservoirs.Based on the above advantages,this paper mainly extracted the information of water level and area in the inland lake reservoir by using space-borne single-photon laser and long-timeseries optical remote sensing image data,and estimated the change of its storage volume.The main research contents are as follows:(1)A denoising algorithm based on two-step statistical histogram method is proposed.Using the altimetry data of the Ice,Cloud and Land Elevation Satellite-2(ICESat-2),the paper first analyzes the main sources of background noise of its data product ATL03.On this basis,a twostep denoising method including "coarse denoising based on sliding window" and "fine denoising based on distance statistical histogram" is proposed to realize effective signal extraction of photon data.The single photon laser data extracted by this method can be used as the vertical information of the lake reservoir to construct the area elevation model and retrieve the water level elevation of the lake reservoir without elevation data.(2)It has realized the water body information extraction based on the modified normalized difference water index(MNDWI)and iterative global threshold,and completed the water body information extraction based on multi-spectral remote sensing images.A total of 320 cloudless satellite images of Landsat from 1984 to 2020 are used as experimental data.First,image batch preprocessing is implemented based on IDL(Interactive Data Language).Second,water bodies are extracted from the preprocessed images and their area is calculated.Finally,the change of longterm temporal area is analyzed.(3)The area elevation(A-E)model of the lake reservoir is constructed to realize the water level inversion in any time phase.The percentage of water occurrence is calculated based on the image extracted from the water body index,and a graph of percentage of water occurrence is generated.The area enclosed by different percentage of water occurrence is regarded as the area(A)in the A-E model.Then based on a quadratic polynomial of one variable,combined with the percentage of water occurrence and photon elevation,an area-elevation model was constructed.The corresponding water level elevation is inverted through the image areas of different time phases obtained in(2),and the inversion accuracy is analyzed.(4)Based on the Landsat images from 1984 to October 2020,the area of Miyun Reservoir is analyzed,and the change process is as follows: increase,decrease,small range fluctuation,increase.The area of the reservoir reached the largest in 1994 and the smallest in 2004.Based on the A-E model constructed by single-photon laser data of Icesat-2 and water occurrence percentage from 2018 to 2020,a total of 130 temporal water level values from 2009 to 2020 were retrieved,and the retrieved water level values have a very strong correlation with the measured data on that day.The annual increase and decrease of water storage are estimated,and the increase and decrease are statistically analyzed.Finally,the factors causing the change are discussed in combination with precipitation,temperature and human activities.This paper has 37 pictures,13 tables,and 84 references. |
PDF Full Text Request