Retrieval of Chlorophyll-a Concentration in the Turbid and Eutrophic Taihu Lake

杨煜,李云梅,王桥,乐成峰,孙德勇,黄昌春
DOI: https://doi.org/10.3724/sp.j.1047.2009.00597
2009-01-01
Abstract:Semi-empirical models are the main models to retrieve chlorophyll-a concentration in remote sensing.The water is so turbid and eutrophic in Taihu Lake that the accuracy of some semi-empirical models varies highly.The datasets were collected over a considerable range of optical properties,trophic status,and geographical locations in Taihu Lake.So,it is very useful to find a model which can retrieve chlorophyll-a concentration in a good accuracy and has the best ability to adapt time changing.This study applied a recently developed model of the form 1/Rrs(λ1)-1/Rrs(λ2)×Rrs(λ3) where Rrs(λi) is the remote-sensing reflectance at the wavelength λi.The objectives of this paper are(a) to build the reflectance first-derivative model、reflectance ratio model and three-band model using the data measured from June to October in 2005 to find which model is the best,(b) to find out the special bands which are suitable in Taihu Lake for the three-band model,and(c) to examine these three models by the dataset measured in November in 2006 and 2007.Strong linear relationships were established between analytically measured chlorophyll-a concentration and both the three-band model 1/Rrs(λ666)-1/Rrs(λ693)×Rrs(λ734) and the two-band model R(λ705)/R(λ680).Accuracy for all these models were measured by R2 and RMSE.The result is that the three-band model is the best one of them.Its normal RMSE is less than 20% of the change which was shown in the dataset.And it has the best ability to adapt time changing that the normal RMSE of 2006 and 2007 is 11.17 and 9.74 respectively.We can also find that the bands used in the three-band model are available from some satellites.So,the three-band model has a good future to retrieve chlorophyll-a concentrations.
What problem does this paper attempt to address?