基于生物光学模型的巢湖悬浮物浓度反演

根据2009年6月巢湖32个样点实测的遥感反射率、悬浮物浓度、吸收系数及散射系数等数据,分析巢湖水体各组分的吸收、散射等固有光学特性,确定悬浮颗粒物单位散射系数、后向散射概率等固有光学参数,构建基于生物光学模型的悬浮物浓度反演模型,并利用准同步获取的环境1号卫星CCD影像数据反演巢湖悬浮物浓度.结果表明,555 nm处悬浮颗粒物单位散射系数的平均值为0.48 m2/g,以555 nm为参考波长,建立指数衰减模型对悬浮颗粒物单位散射系数进行参数化,模型的决定系数可以达到0.99;此外,在760～900 nm(Band4)范围内,后向散射概率不具有波长依赖,其值稳定在0.051.利用所得到的表观及固有光学量构建巢湖水体遥感反射率模型,反演巢湖悬浮物浓度,得到实测值与反演值之间的相对误差随着浓度的增加而呈现下降的趋势,平均相对误差为17.25%,由此表明该方法适用于反演悬浮物浓度较高的湖泊水体;利用两景环境1号卫星CCD影像数据反演得到的巢湖悬浮物浓度主要在0～100 mg/L之间变化,其中6月13日巢湖悬浮物浓度40 mg/L的水域占到总面积的54.37%,而6月15日巢湖61.62%的水域悬浮物浓度40 mg/L,且这2 d巢湖悬浮物的分布与当时的气候变化一致.

Estimating of Suspended Matter Concentration Based on Bio-optical Model in Chaohu Lake

Remote sensing reflectivity (R(rs)), total suspended matter (TSM), absorption and scattering properties of 32 sampling sites in situ were measured in June 2009. The inherent optical properties (IOPs), such as absorption and scattering properties of different components in Chaohu Lake, were analyzed according to these data. And specific scattering coefficient of suspended matters and the backscattering ratio which were the indispensable inherent optical parameters were identified. Then an inversion model based on bio-optical model was constructed and quasi-synchronous HJ-1 CCD images were used to inverse the concentration of suspended matters in Chaohu Lake. Results showed that, the average specific scattering coefficient of suspended matters at the wavelength of 555 nm was 0.48 m2/g. A power model about specific scattering coefficient of suspended matters was constructed by choosing 555 nm as a reference wave band with determination coefficient of 0.99. In addition, the backscattering ratio didn't rely on the wavelength from 700 nm to 900 nm (Band 4), and its value is stable at 0.051. A remote sensing reflectivity model was constructed by IOPs and apparent optical properties (AOPs) to obtain TSM in Chaohu Lake. The trend of relative error between measured values and inversed values decreased with the concentration of TSM increasing. And the average relative error value was 17.25% by using this model. This showed that this method was applicable to inverse TSM in inland lake with high concentration suspended matters. Based on two CCD image data of HJ-1 satellite, the inversion of suspended matter concentration mainly changed between 0 and 100 mg/L in Chaohu Lake. The suspended matter concentration higher than 40 mg/L covered 54.37% area of Chaohu Lake in June 13 and that of 61.62% in June 15. As well as, the distribution of suspended matter concentration inversed by using two HJ-1 CCD images of Chaohu Lake kept correspondence with climate changes at that time.