Modeling Carbon Emissions from Deforestation in Context of REDD+ Using GIS and Remote Sensing

Abstract

Due to increase in greenhouse gases emissions including carbon dioxide (CO2) at faster rate recently results rise in natural disasters and global warming. To resolve this issue protection and conservation of forests is necessary but these are destroyed at faster rate because of increase in anthropogenic activities. Hence to decrease the emission of carbon as a result of deforestation REDD+ came up by measuring carbon stock and AGB and the carbon in a particular forest is Measured, Reported and Verified (MRV) by developing countries. To obtain AGB and carbon stock collection of forest inventory data was done which includes DBH and height data and then these parameters were putted in allometric equation. In this study AGB estimated was 148.79 t/ha while carbon stock in the study area was 69.93 t/ha. Then from above ground biomass/carbon below ground biomass/carbon was estimated which were 38.68 t/ha and 18.18 t/ha. By adding the Above ground and below ground biomass/carbon we obtained total biomass/carbon of 187.48 t/ha and 88.12 t/ha. The total carbon was converted into CO2 equivalent and its mean value was about 322.5 t/ha. Total carbon (AGC + BGC + Soil Carbon) in study area was 94.44 t/ha. The carbon sequestration capacity calculated was 82.07 ± 13 t/ha. In fifteen years carbon emissions were 6.96 Mt CO2 e and according to departmental working plan data total carbon emission were 3.589 Mt CO2 e. Soil samples were collected in forest area using grid sampling and their lab analysis was done and then IDW interpolation was done to obtain soil carbon map of study area the mean soil carbon value found was 6.33 t/ha. Landsat data and supervised classification algorithm were used in order to assess change in forest cover. In fifteen years about 27209.64 ha area has been deforested which is 16.88% and annual rate of deforestation was 2.51 %. Assessment of Sentinel-2 VIs was done in estimation of above ground biomass while using linear regression model relationship between AGB and various VIs was assessed. From regression models map of predicted biomass were developed. The study reveals that regression model of AGB and RERVI was best because of its high R-square value of 0.68 and P-value less than 0.01 and low RMSE value of 35.23 t/ha. The Study suggest that from sentinel-2 derived VIs biomass/carbon could be assessed more accurately due to high spatial resolution (10m), less saturation effect and three Red-edge bands specifically for studying vegetation and this method seems to be accurate, best, economical and will assist REDD+ in SFM (Sustainable forest management) and lessening Carbon dioxide losses.