Forest cover change detection and spatial modelling for identifiction of errosion prone areas in Kaptai reservoir watershed of Bangladesh

Abstract

Erosion has become an important issue for Kaptai watershed, the largest man made freshwater body in Bangladesh which generales a significant share of Bangladesh's power supply by usiog hydro-power. It contributes significantly to the production of freshwater fish, flood oonlrol, tourism and agriculture. Over lhe past few decades forest depletion has become a main problem here because of human intervention, replacing the forest cover with shllUng cultivation which eventually increasing soil erosion and thus contributing large amounlS ofscdirncll1lltion to the reservoir posing 3 serious threat to the power plant as well as navigability. Studies have been done \0 assess the sedimentation in Kaptai Lake have found that the expected tife time ofthc lake has considemhly reduced. This study is an effort to look into this met and II)' to understand the spatial variation of erosion that leads to siltation. The study has two major goals - first to nwnitor the change ofvcgctation and other major land coven; OYl:rlast two decades (1980-2000) and ~econd to develop a simple spatial model to identify potential erosiOll prone areas in Kaptai watershed. Land cover maps have been produced from satellite image~ of Landsat program. NOVI and bi-spectra! plot techniques were u~ed to identify land covers. Appropriate NOVI has been set chosen based on ground tnllt! data, high resolution satellite image (IRS LlSS 111) and sccondlll'Y land cover maps. Two land cover map~ have been prepared for two nominal years (1980 and 2000). Finally map~ have been compared with each other. It has been found lIlat non-furest land has increased by 10.6% \\'hile low density forest area has decreased by 18.4% but high density fOrest has increased by around 7.9%. However overall forest coverage (Iow+high density) has decreased by 10.6% which is highly significant for20 years time period (1980-2000). A simple potential erosion model has been developed to assess the distribution of potential erosion prone areas in the watershed. This model cmphasize~ potential erosion because it does not estimate the actual erosion; rather it indicates areas that are potentially i, erodabte based on ~npe, soil, land cover and rainfall in the watershed. Modelling result shows that 45% of the area has low erosion risk while 17% has high erosion risk. The rest 36% falls into llVerage erosion category. This is an indication of how much area is to bc conserved for erosion management in Kaptai Watershed. This study is a small preliminary effort to look into the erosion problem spatially. But this might be useful as a starting point for further in depth spatial study of this problem to figure out possible remedies.