Development of ocean color algorithms for estimating chlorophyll-a concentrations and inherent optical properties using gene expression programming (GEP)

研究成果: Article同行評審

3 引文 斯高帕斯(Scopus)

摘要

This paper proposes new inversion algorithms for the estimation of Chlorophyll-a concentration (Chla) and the ocean's inherent optical properties (IOPs) from the measurement of remote sensing reflectance (Rrs). With in situ data from the NASA bio-optical marine algorithm data set (NOMAD), inversion algorithms were developed by the novel gene expression programming (GEP) approach, which creates, manipulates and selects the most appropriate tree-structured functions based on evolutionary computing. The limitations and validity of the proposed algorithms are evaluated by simulated Rrs spectra with respect to NOMAD, and a closure test for IOPs obtained at a single reference wavelength. The application of GEP-derived algorithms is validated against in situ, synthetic and satellite match-up data sets compiled by NASA and the International Ocean Color Coordinate Group (IOCCG). The new algorithms are able to provide Chla and IOPs retrievals to those derived by other state-of-the-art regression approaches and obtained with the semi- and quasi-analytical algorithms, respectively. In practice, there are no significant differences between GEP, support vector regression, and multilayer perceptron model in terms of the overall performance. The GEP-derived algorithms are successfully applied in processing the images taken by the Sea Wide Field-of-view Sensor (SeaWiFS), generate Chla and IOPs maps which show better details of developing algal blooms, and give more information on the distribution of water constituents between different water bodies.

原文English
頁(從 - 到)5417-5437
頁數21
期刊Optics Express
23
發行號5
DOIs
出版狀態Published - 2015

All Science Journal Classification (ASJC) codes

  • Atomic and Molecular Physics, and Optics

指紋 深入研究「Development of ocean color algorithms for estimating chlorophyll-a concentrations and inherent optical properties using gene expression programming (GEP)」主題。共同形成了獨特的指紋。

引用此