การประมาณค่าดัชนีพื้นที่ผิวใบของป่าชายเลนโดยใช้การสำรวจระยะไกลแบบไฮเปอร์สเปกตรัล

Abstract

Mangroves are the part of the coastal ecosystem. They are very important ecosystems and society. A well-recognized benefit of mangrove forests is their fundamental role in coastal protection from erosion the nurturing of the juvenile, as well as the food source. Therefore, the investigation and monitoring of mangroves thus become a very important. This study is aimed to estimate the Leaf Area Index (LAI) of the tropical mangroves by hyperspectral remote sensing data. The hyperspectral satellite image (i.e., EO-1 Hyperion) was used to create the models for estimates the LAI of dense (LAI = 2.94 - 4.97 m2/ m2) and diverse species of mangrove forest at Pa Khlok Bay, Phuket province. The regressions analysis was used to find the relationship between dependent variable (LAI) and independent variables (i.e., the selected of 10 vegetation indices). From 51 model, the results showed that the Support Vector Machine method provided the highest accuracy for estimated the LAI (R-square = 0.657, RMSE = 0.292 m?/ m?) and the averaged of estimated LAl was 3.94 m?/ m?. This study shows the capability of the hyperspectral data in LAl estimation of dense and diversity of mangroves application. We hope that the methodology presented in this research can be used as a guideline for study in other area or can be used for mangrove forest management in the future.