Estimation of aboveground biomass and carbon stocks in Matang Mangrove Forest Reserve using allometry model and remote sensing technique

Among other forest types, mangrove forest is one of the most important ecosystems. Being able to act as a “sponge” for carbon pools is one of the biological services provided by mangrove forest. Tree biomass estimates and analyses are essential for carbon accounting and other feasibility studies including bioenergy. Mangrove plays an important role in ecosystems which could mitigate the climate change through biomass and carbon storage. Due to concerns on global climate change and carbon sequestration a reasonable method for estimating tree biomass and carbon stocks are highly demanded. Thus, this study was design to estimate the aboveground biomass (AGB) of mangrove trees using allometry model and remote sensing technique. A total of 150 plots from 17 compartments were sampled in 2014. Based on remote sensing technique, SPOT-5 data were used to estimate and mapping aboveground biomass and carbon stocks. Four types of vegetation indices (VI’s) were selected and tested in this study. The estimation of AGB was further refined using integration analysis from direct method data. To achieve the objective of this study, optical image and sampling data had been processed and analysed, and validated by using regression model. By using non-destructive method and allometry model, the results shown that average AGB per plot was about 168.93 ton ha-1. The maximum value was 462.40 ton ha-1 and the minimum was 24.35 ton ha-1 respectively. While average value of carbon storage was 84.47 tonC ha-1. Then, exponential regression model had performed to estimate the AGB along with optical image. The findings demonstrate a good relationship between measured AGB and selected vegetation indices (NDVI and GEMI-NDVI). The others two like SAVI and GNDVI had proved weak relationship with measured AGB. However, the results indicate that there were slightly increase for about 3% with the using multi-exponential regression analysis. Study had further refined by using multi-exponential of integration method from incorporating both predicted NDVI and GEMI-NDVI with coefficient of determination (R2 = 0.74) which had proved that the increment of 1% compared with multi-exponential of direct method and the overall result for RMS error was 85.24 ton ha-1. This shows that the average estimation of AGB was 130.36 ton ha-1. Therefore, total amount of AGB for the whole study area in Kuala Sepetang (South) (9,884 ha) approximately of 1.3 million tonnes. The amount of AGB was 1.80% slightly overestimated compared with previous study by using destructive sampling. Thus, the studies had suggested that the non-destructive sampling by using common allometric equation is still effective and reasonable to be used for the estimation on AGB in mangroves forest. The correlation and regression analysis were done separately between AGB and vegetation indices by using direct method and integration method. Results had shown that the multi-exponential integration method is highly in accuracy and strongly correlated with the field data among the others correlation. In conclusion, the study indicates that the common allometric equation for calculating AGB was applicable for all mangrove species instead of species-specific equation. With the using of remote sensing technique and multi-exponential integration analysis, the estimated biomass and carbon stocks were slightly increased for about 1%. The regression model developed may be useful for estimating the AGB of areas that are not reachable and in low cost. The study using satellite imagery data was an attempt to improve the estimates by integration method as a final outcome for better conversion of biomass to carbon stock content.

Preview Text FH 2017 11 IR.pdf Download (1MB) | Preview

Actions (login required)

View Item