BayCEER Workshop 2008

Regional-scale mapping of tropical forests using integrating MODIS Vegetation Continuous Field, Landsat and IRS data for systematic monitoring of forest cover change and fragmentation analysis in Indian region

In this paper, we demonstrate an approach that uses MODIS Vegetation Continuous Field (MODIS 44B-VCF) at regional level to monitor tree cover change in the Indian region over 6 year (2000 – 2005) period. We used pixel based linear regression model to identify its changes in the form of slope, offset and variance and then classified using threshold method into forest and non-forest classes, which has resulted to pixel change as no-change and change as positive and negative We validated MODIS-VCF raw product using field data gathered in different tree crown cover as part of DOS-DBT biodiversity-landscape project and the results shows a reasonable relationship (86.6%) between tree canopy cover with tree crown cover. The change pixels where masked with GLCF land-cover map to estimate percent tree cover change within the forest region for further assessment. Spatial pattern analysis for the change areas was implemented to understand forest dynamics that describes the landscape characteristics. Results were overlaid with UNEP protected area boundary for natural resource management and wildlife protection. On a long-term basis, we monitor these forest cover change using high spatial resolution data (1972 – 2004) to identify rates of deforestation and fragmentation at landscape level for sustaining the region’s biodiversity and managing the forest ecosystems. The current study propose Rapid Ecological Alerts (RET’s) using the forest area change and the approach can be made automation for every year.