| In China, the lake eutrophication is serious and the lake alga is frequent occurrence, that make people’s manufacture and life greatly impact, and then the problem of the lake alga is widely concerned by people. Recently, it has become a hot research to dynamically monitor the lake alga’ formation and distribution by means of remote sensing. The signal coming from the inland water is weak and heavily affected by atmosphere. The signal is only less than 20% of the signal of satellite remote sensing, at the same time, it is constantly changing due to the fast reproduction of alga and the water flowing. All of those makes it more difficult to quantitatively extract algae blooms information by satellite remote sensing. Currently, in order to improve extraction accuracy, it was a very critical issue for satellite remote sensing to monitor the lake alga that how to effectively carry on the atmospheric correction.The quantitatively monitoring level of water quality parameters, such as chlorophyll a concentration, is closely related to the inversion accuracy of the watersurface reflectivity. Based on atmospheric radiative transfer theory, the main error sources affecting the accuracy of atmospheric correction of satellite remote sensing over inland water were systematically analyzed. The stable pixel atmospheric correction method was proposed based on the satellite image information to remedy the problems of the existing methods such as dependence for the meteorological data, low correction accuracy and so on. And the application effect for the atmospheric correction method was test and evaluated. On this basis, the chlorophyll a concentration for the past five years based on FY-3 MERSI data in Taihu Lake were retrieved by the proposed method in this study. And the change characteristics over time and the relationship with meteorological factors of the lake alga in Taihu Lake were analyzed.The main research contents and innovation are as follows:1) Atmospheric correction error transfer characteristics and the response analysis of each error factor by aerosol disturbance over inland water. The error transfer equation of satellate remote sensing over inland water was constructed in theory, and the change characteristics of the retrieval error of water surface reflectivity with the variation of each error factor, atmospheric conditions and water conditions were quantitatively analyzed by mathematical modeling. The analysis results pointed out that the main error of atmospheric correction was coming from the correction error of atmospheric upward reflectivity whose error magnification was up to 13 times for water reflectivity, the second was coming from the atmospheric downward and upward transmissivity who were almost equivalent to the error of water surface reflectivity, and the correction error of atmospheric downward reflectivity had only little effect on the error of water surface reflectivity. By simulation, the influence of each error factor from aerosol(such as aerosol type and aerosol optical depth(AOD)) disturbance and the influence of the error of atmospheric transmissivity from other atmospheric composition were quantitatively analysized. The results pointed out that the large correction errors were introduced in inland water environment by the role of aerosol. To improve the accuracy of atmospheric correction, the sensitivity of the correction results for aerosol disturbance should be reduced or eliminated for the mospheric correction algorithm constructed.2) The Stable Pixel atmospheric correction method construction. For the error response characteristics on water surface reflectivity from different error factors, the Stable Pixel atmospheric correction method for inland water(such as lakes) was constructed from the point of view of error magnification, and the error transfer process of remote sensing for Stable Pixel atmospheric correction method was analyzed and evaluated in theory. The results showed that the correction error of atmospheric upward reflectivity, the error of satellite measurements system, and the error of water reflectivity from other parameters information unrelated to retrieved parameter in the target pixel and the wave characteristics consistent with the target pixel were eliminated by the method, and then the relative differences of chlorophyll a concentration between each pixel in satellite images were preserved.3) Spectral response model construction for chlorophyll a concentration according to the field synchronous measurement data. Using the field synchronous measurement spectral data and chlorophyll a concentration in Taihu Lake in a year from May 2008-May 2009 from Shanghai Institute of Technical Physics of the Chinese Academy of Sciences, the spectral response models such as single-band model, band-ratio model, the normalized-index model and three-band model were established based on chlorophyll a sensitive bands. The result showed that there were a higher correlation between the models of band-ratio and three-band and the chlorophyll a concentration. At the same time, the single-band model based on near-infrared and the band-ratio model based on near-infrared and red band were constructed according to the spectral bands of 250 m spatial resolution of MODIS data and FY-3 MERSI. The results showed that both had high correlation with chlorophyll a concentration, but the correlation of latter was higher than the former as the same atmospheric correction method was applied for the satellite images.4) The Stable Pixel atmospheric correction method test. The application effect of the Stable Pixel atmospheric correction method was tested respectively from three aspects such as the error analysis such as relative error and absolute error and so on with measured data, the consistency on data value and space distribution with measured data and the sensitivity to the error of atmospheric parameters. The test results showed that the retrieval results of chlorophyll a concentration based on Stable Pixel atmospheric correction method had higher accuracy than the correction method for 6S radiative transfer model, higher consistency on the data value and space distribution with measured data across the whole lake and less sensitivity for aerosol error. Besides,the application stability of spectral model for chlorophyll a concentration was improved using the Stable Pixel atmospheric correction method.5) The analysis of the time evolution law and the relationship with meteorological factors of the lake alga in Taihu Lake. Based on the Stable Pixel atmospheric correction method proposed in the study, the Chlorophyll a concentrations were retrieved five years in a row using the FY-3 MERSI data from 2010 to 2014 in Taihu Lake. And the varying law with the time of the area of the lake alga and the correlation with meteorological factors were analyzed. The results showed that the occurrence and development of the lake alga were related with the overall climate conditions of the year, and wind speed and temperature had a strong driving action for the growth of the lake alga. The weather conditions such as high temperature and calm wind were conducive to reproduction and diffusion of the lake alga.The study work in the topic can provide a strong support for satellite remote sensing to real-time dynamicly monitor and analyze the lake alga. |
PDF Full Text Request