The crisis of climate change has drawn attention of ecologists all over the world to explore ways that could effectively enhance the sequestration of carbon in forests and woodlands. This necessitates forest inventory, along with knowledge of techniques that are quick and manageable. The present study was carried out in Miombo woodlands of western Tanzania namely, Kitwe and Mgaraganza. The aim of the present study was to investigate the effectiveness of remote sensing in estimating the stock of carbon in Miombo woodlands. Two types of data namely, ground-truthing and satellite imagery were used. Ground-truthing data were obtained by measuring the diameter at breast height (DBH) of all trees in 30 and 20 concentric plots in Kitwe and Mgaraganza forests, respectively. The average DBH of trees in each forest was fitted in biomass allometric models to estimate the ground-truthing vegetation biomass. On the other hand, Landsat images of the two forests were used to compute the Normalised Vegetation Index (NDVI). The computed NDVI were regressed with the ground-truthing vegetation biomass to get the remotely sensed vegetation biomass which was assumed to be 50% carbon. The coefficients of determination between the ground-truthing above-ground biomass and the NDVI values were statistically significant at P<0.05. The above-ground carbon stock obtained by ground-truthing in Mgaraganza and Kitwe forests was 3 and 2 times higher than that of satellite remotely sensed data respectively. The above-ground carbon stock obtained from satellite remote sensing gives some impression thus a basis for remote sensing in Miombo woodlands
World Bank through Mkwawa University College of Education