| Autor: |
Putra Suardana, A. A. Md. Ananda, Anggraini, Nanin, Nandika, Muhammad Rizki, Aziz, Kholifatul, As-syakur, Abd. Rahman, Ulfa, Azura, Wijaya, Agung Dwi, Prasetio, Wiji, Winarso, Gathot, Dimyati, Ratih Dewanti |
| Předmět: |
|
| Zdroj: |
Forest & Society; Apr2023, Vol. 7 Issue 1, p116-134, 19p |
| Abstrakt: |
Carbon dioxide (CO2) is one of the greenhouse gases that cause global warming, with the highest atmospheric concentration. Mangrove forests absorb CO2 three times higher than terrestrial forests and tropical rainforests. Moreover, mangrove forests can be a source of Indonesian income in the form of a blue economy. Therefore, an accurate method is needed to investigate mangrove carbon stock. The utilisation of remote sensing data with the results of the above-ground carbon (AGC) detection model of mangrove forests based on multispectral imaging and vegetation index can be a solution to obtain fast, cheap, and accurate information related to AGC estimation. This study aimed to investigate the best model for estimating the AGC of mangroves using Sentinel-2 imagery in Benoa Bay, Bali Province. The Random Forest (RF) method was used to classify the difference between mangrove and non-mangrove by treating several parameters. Furthermore, a semi-empirical approach was used to assess and map the AGC of mangroves. Allometric equations were used to calculate and produce AGC per species. Moreover, the model was built with linear regression equations for one variable x and multiple regression equations for more than one x variable. Root Mean Square Error (RMSE) was used to assess the validation of the model results. The results of the mangrove forests area detected in the research location around 1,134.92 ha, with an Overall Accuracy (OA) of 0.984 and a kappa coefficient of 0.961. This study highlights that the best model was the combination of IRECI and TRVI vegetation indices (RMSE: 11.09 Mg/ha) for a model based on red edge bands. Meanwhile, the best results from the model that does not use the red edge band were the combination of TRVI and DVI vegetation indices (RMSE: 13.63 Mg/ha). The use of red edge and NIR bands is highly recommended in building the AGC model of mangrove forests because they can increase the accuracy value. Thus, the results of this study are highly recommended in estimating the AGC of mangrove forests because it has been proven to increase the accuracy value of previous studies using optical images. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
