Development And Application Of An Automatic Remote Sensing Algorithm For Submerged Vegetation In Shallow Lakes

Lakes are one of the most important freshwater resources on earth,and they have important ecological functions and economic value.In recent years,due to the influence of human activities,the area of the lake has shrunk severely and the eutrophication of the water has become serious.Aquatic vegetation,especially submerged vegetation,as an important component and main primary producer in lake ecosystems,can play a role in purifying water quality,maintaining the stability of lake ecological environment and biodiversity.Using remote sensing to classify aquatic vegetation in inland lakes and dynamic monitoring of submerged vegetation can provide decision support for lake environmental governance and ecological restoration.Based on this,this research mainly conducts the research of the following contents and draws corresponding conclusions:(1)Based on Landsat 8 OLI remote sensing images and field survey data,an automatic dynamic threshold aquatic vegetation decision tree classification algorithm was established.Using this algorithm,lake aquatic vegetation was divided into two types:one is vegetation that grows above the water surface,including emergent vegetation and floating-leaved vegetation,and the other is submerged vegetation that grows underwater.In this algorithm,the above two types of aquatic vegetation are extracted from the water body using a combination of NDWI index and FAI index,and then the two types of aquatic vegetation are distinguished using the reflectivity characteristics of the SWIR band.In order to make the algorithm have good applicability and portability for long-time image classification,spatial interpolation and other methods are adopted,so that the classification threshold can be automatically changed according to the spectral reflectance characteristics of the water area of the study area,which greatly reduces The artificial intervention of the algorithm and the dependence of the algorithm on the sample points improve the classification accuracy and calculation time of the algorithm.(2)In order to verify the availability of the algorithm on different Landsat sensor data,this paper uses the algorithm to classify aquatic vegetation for Landsat 7 ETM+and Landsat 8 OLI images at 8-day intervals,and uses field survey data that is close to the image acquisition time to verify the classification results.The results show that the classification algorithm proposed in this paper has good classification results on the data obtained by different sensors of Landsat,which lays the foundation for the next algorithm application.(3)Water depth is the main factor restricting of the remote sensing identification of submerged vegetation.In order to explore the maximum depth of submerged vegetation that can be identified by the algorithm and the sensitivity of the algorithm to the concentration of light-sensitive parameters of water,this paper uses simulated water reflectance spectra and corresponding water diffusion attenuation coefficients(K_d)at different Chl-a and TSS concentrations,combined with the reflectivity attenuation function of vegetation under water and the linear mixing principle of pixels,obtains the simulated spectral characteristics of submerged vegetation at different depths and different vegetation coverage.By calculating the FAI index of these simulated spectra,the following conclusions are drawn:The algorithm proposed in this paper can identify aquatic vegetation 0-0.20 m underwater;the algorithm is not sensitive to Chl-a concentration of water.(4)Based on Landsat 5 TM and Landsat 8 OLI remote sensing data,using the aquatic vegetation classification algorithm proposed in this paper,we obtained the spatial and temporal distribution data of submerged vegetation in four large lakes in Hubei Province from 1986 to 2018.Combined with meteorological data such as precipitation,temperature and wind speed,and human activity information such as enclosure fisheries,the long-term classification results were analyzed,and it was concluded that precipitation,temperature and human activities may be the main factors affecting the growth of lake submerged vegetation.