Feasibility Study of Retrieving Water Quality and Particle Size from Hyperspectral Remote Sensing Reflectance

• Main Authors: Chin-HowLin, 林俊豪
• Other Authors: Cheng-Chien Liu
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/13944878533969042484

碩士 === 國立成功大學 === 地球科學系碩博士班 === 101 === Remote sensing of satellite technology in recent years has been widely used in marine resources exploration and development, and on the use of telemetry data that we estimate water quality has been considerable progress in the previous study such as Quasi-Analytical Algorithm (QAA)、Garver-Siegel-Maritorena model (GSM) and GA-SA(Genetic algorithms and semi-analytical algorithm). QAA in applications can estimate inherent optical properties(IOP) with high accuracy, but it can’t directly obtain the required water quality parameters; In GSM IOP is determined primarily by the impact of chlorophyll concentration, and it doesn’t apply to the higher concentration of sand in the water; In GA-SA, the accuracy of the IOP slightly inferior than QAA, but has a direct solver water quality parameters advantages, however, its inverse suspended material accuracy is unsatisfactory, therefore, this study intends to use Downhill Simplex Method (DSM) algorithm to establish water quality inverse mode. In this study, we want to take a different composition in the IOP of the water bodies in the field under different ambient light of the remote sensing reflectance (Rrs) , which requires large amounts of data for analysis. Although in situ data with accuracy, but its collection is quite waste time and labor, and we have a limited number of information right now, therefore, this study for each water body, establish a complete simulation database, and then based on this information to inverse water quality parameters and particle size. Simulation database created using Hydrolight radiative transfer model to simulation with different IOPs of the water body, selected during the simulation contains chlorophyll-a concentration(C), chlorophyll backscatter ratio (bbph /bph), colored dissolved organic matter absorption coefficient(F) and attenuation coefficient(r), suspended particles concentration(d) and its backscatter ratio(bbd/bd) of six water quality and optical parameters, and the sun angle, cloud cover, average visibility, wind speed and other kinds of environmental parameters, then we use various combinations of situation arising Rrs , and Rrs into the established patterns in order to obtain the inverse water quality parameters and the optical properties of the best resolve. Then used with the original Hydrolight simulated that used water quality and optical parameters were compared to verify the accuracy of the inverse. Finally, the inverse model obtained particle backscatter coefficient (bbp) to discussed in the biotic and abiotic suspended solids size and bb/b correlation, and then analyze the particle size distribution.