Study On Aquatic Vegetation Identification And Huangtai Algae Bloom Coverage Monitoring Of Lake Wuliangsuhai Based On Remote Sensing Images

Water resources in lakes are one of the important components of global water resources.Lakes have multiple ecological functions and various service values.In recent years,the water quality of lakes has deteriorated,and the eutrophication status of lakes has been severe.Remote sensing monitoring of aquatic vegetation in lakes can comprehensively reflect the eutrophication status of lakes.In this paper,lake Wuliangsuhai,the largest freshwater lake in the Yellow River basin,was selected as the research area.Based on China's GF-1 WFV 16m data and American Landsat data,the spectral curves of aquatic vegetation were synthetically analyzed,and a concave-convex detection function was established for the first time to accurately identify and extract aquatic vegetation in lake Wuliangsuhai.In order to reveal the evolution process of the water environment in lake Wuliangsuhai,the coverage of Huangtai algae bloom in lake Wuliangsuhai,which is sensitive to eutrophication,was retrieved.At the same time,the factors affecting the change of Huangtai algae bloom were analyzed.The main conclusions are as follows:(1)Based on GF-1 16m resolution data in July and August 2015 which has abundant vegetation information,the aquatic vegetation in lake Wuliangsuhai was divided into three types:emergent vegetation,submerged vegetation,and Huangtai algae bloom.It is difficult to identify and accurately extract water and vegetation that grow underwater on the basis of conventional vegetation indices.The concave-convex detection function which can overcome the shortcomings of low classification accuracy of traditional classification method was established at the very first time to distinguish the water body and submerged vegetation.In combination with classification decision tree method,emergent vegetation and Huangtai algae bloom were classified simultaneously.Through the verification of model classification results based on simultaneously measured in situ data,the classification accuracy in July and August were 92.59%and 91.79%,respectively,showing that GF-1 data can completely meet the accuracy requirement of aquatic vegetation extraction.With the addition of 4-day short revisit period and high spatial resolution,it is a reliable data source for monitoring the long term growth of aquatic vegetation.(2)Based on Landsat data,we constructed a dimidiate pixel model based on a new AVI index and calculated the coverage of Huangtai algae bloom in lake Wuliangsuhai for each pixel.According to the verification of inversion accuracy using the in situ coverage data of Huangtai algae bloom on July 25-27,2016,the fitting accuracy was R~2=0.8661,indicating that the inversion model can be used for the coverage extraction of Huangtai algae bloom.Subsequently,based on the coverage inversion model,the monthly coverage,annual average coverage,and multi-year average coverage of Huangtai algae bloom in 2006-2016 was estimated.The results showed that the outbreak of Huangtai algae bloom was severe in 2008,2009 and 2010.The annual average coverage of Huangtai algae bloom reached14.43%,14.06%,and 12.56%,respectively.In 2011,the coverage of Huangtai algae bloom fell sharply,then the coverage of Huangtai algae bloom continued to decrease,reaching a minimum of 0.33%in 2013.The spatial distribution of the Huangtai algae bloomcoverageinlakeWuliangsuhaishowedthephenomenonof high in east and low in west.Huangtai algae bloom was mainly concentrated in the central and eastern lake,and the distribution of Huangtai algae bloom in the southern lake was also intensive,while the outbreak of Huangtai algae bloom in the western lake was relatively rare.(3)Based on inversion coverage data of Huangtai algae bloom in 2006-2016,nutrient salt data and meteorological data in lake Wuliangsuhai,correlation analysis had been carried out.The results showed that there was no significant correlation between the annual average coverage of Huangtai algae bloom and the annual average concentration of nitrogen(N)and phosphorus(P).However,when the N and P concentrations in March and May was respectively correlated with the average coverage of Huangtai algae bloom in June and August,they both showed a significant positive correlation,and the Pearson correlation coefficients were 0.629 and 0.642,respectively,indicating that the water nutrient concentration in March and May promoted the growth of Huangtai algae bloom.In meteorological factors,the effect of precipitation on the coverage of Huangtai algae bloom showed a two-month lag.A significant correlation also existed between the coverage of Huangtai algae bloom and the temperature of this month,the temperature of one month ago,the temperature of two months ago and the temperature of three months ago,respectively.The Spearman correlation coefficients were 0.411,0.497,0.438 and 0.446,respectively,indicating that temperature had a great influence on the growth of Huangtai algae bloom.Wind speed in spring can promote the growth of Huangtai algae bloom.There was a significant positive correlation between the coverage of Huangtai algae bloom in this month and the wind speed of three months ago,and the Spearman correlation coefficient was 0.495.While the monthly coverage of Huangtai algae bloom was significantly negatively related to the wind speed in the month,which is mainly due to the disturbance of wind after the outbreak of Huangtai algae bloom will affect the stable growth of Huangtai algae bloom on the water surface.