摘要
针对太湖水质富营养化、水华现象时常发生且难以实时快速监测等问题,文中基于最小二乘原理的6种函数拟合模型(linear、quadratic、exponential、geometric、hyperbolic、logsquare),应用Landsat卫星影像的波段比值和归一化差值叶绿素a指数,结合同时期太湖实测叶绿素a浓度进行反演和验证,通过精度对比,筛选最佳拟合模型,并将模型应用于同季度2009年、2010年、2015年、2020年的4幅影像中进行太湖叶绿素a反演,对比分析太湖富营养化水域的时空变化特征。结果表明:应用最小二乘原理的quadratic函数模型,拟合遥感影像第4波段与第2波段反射率归一化差值和实测值的效果最佳,反演和验证的均定系数分别为0.883 3、0.827 9。2019年太湖水域严重富营养区域占比5.30%,所占面积为123.92 km^(2),其余均为富营养区域。模型应用于2010年、2015年、2020年进行反演可得严重富营养区域占比分别为6.38%、2.49%、4.60%,所占面积分别为149.16 km^(2)、58.19 km^(2)、107.49 km^(2),其余均为富营养区域。
Eutrophication and algal blooms in Taihu Lake occur frequently and are difficult to be monitored in real time.In this study,six function fitting models(linear,quadratic,exponential,geometric,hyperbolic,logsquare) based on the principle of least squares were used,and the band ratio and normalized difference chlorophyll a index of Landsat satellite images were used,combined with the actual measurement of Taihu Lake during the same period.The concentration of chlorophyll a was inverted and verified,and the best fitting model was screened through accuracy comparison,and the model was applied to the four images of the same quarter in 2009,2010,2015,and 2020 to invert the chlorophyll a in Taihu Lake,and the temporal and spatial variation characteristics of trophic waters in Taihu Lake was compared and analyzed.The results show that the quadratic function model based on the principle of least squares has the best effect on fitting the normalized difference and measured values of the reflectivity between the fourth and second bands of remote sensing images,and the average coefficients of inversion and verification are 0.883 3 and 0.827 9,respectively.In 2019,the severely eutrophic areas in the Taihu Lake waters accounted for 5.30%,covering an area of 123.92 km^(2),and the rest were eutrophic areas.The model is applied to inversion in 2010,2015,and 2020,and the proportion of severely eutrophic regions is 6.38%,2.49%,and 4.60%,and the area is 149.16 km^(2),58.19 km^(2),107.49 km^(2),and the rest are rich Nutrition area.
作者
苑春雨
刘昌华
YUAN Chunyu;LIU Changhua(School of Surveying and Land Information Engineering,Henan Polytechnic University,Jiaozuo 454000,China)
出处
《黑龙江工程学院学报》
CAS
2023年第3期20-27,34,共9页
Journal of Heilongjiang Institute of Technology
基金
国家级大学生创新创业训练计划项目(202011765025)。
