Evaluating MODIS-vegetation continuous field products to assess tree cover change and forest fragmentation in India – A multi-scale satellite remote sensing approach uri icon

abstract

  • Monitoring the changes in forest-cover and understanding the dynamics of the forest is becoming increasingly important for the sustainable management of forest ecosystems. This paper uses temporal MODIS Vegetation Continuous Field (MODIS-VCF) to monitor the tree cover change in the Indian region over a period of 6 years (2000-2005). Pixel-based linear regression model is developed to identify rate of deforestation and fragmentation at landscape level. The regression parameters viz., slope, offset and variance are used to identify threshold between forest and non-forest classes. The classification algorithm resulted into change area, no change area, positive change and negative changes. MODIS-VCF raw product of 2005 was validated using the field data and showed a coefficient of determination (R-2 = 0.85) between percent tree cover and individual plot wise canopy cover information. The results were overlaid with UNEP protected area boundary. On a long-term basis, the forest cover change was monitored using medium spatial resolution (Landsat and IRS) satellite data to identify the rate of deforestation and fragmentation at landscape level. The developed approach is efficient and effective for regional monitoring of forest cover change. It could be automated for regular usage and monitoring. (C) 2017 National Authority for Remote Sensing and Space Sciences. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license.
  • Monitoring the changes in forest-cover and understanding the dynamics of the forest is becoming increasingly important for the sustainable management of forest ecosystems. This paper uses temporal MODIS Vegetation Continuous Field (MODIS-VCF) to monitor the tree cover change in the Indian region over a period of 6 years (2000?2005). Pixel-based linear regression model is developed to identify rate of deforestation and fragmentation at landscape level. The regression parameters viz., slope, offset and variance are used to identify threshold between forest and non-forest classes. The classification algorithm resulted into change area, no change area, positive change and negative changes. MODIS-VCF raw product of 2005 was validated using the field data and showed a coefficient of determination (R2 = 0.85) between percent tree cover and individual plot wise canopy cover information. The results were overlaid with UNEP protected area boundary. On a long-term basis, the forest cover change was monitored using medium spatial resolution (Landsat and IRS) satellite data to identify the rate of deforestation and fragmentation at landscape level. The developed approach is efficient and effective for regional monitoring of forest cover change. It could be automated for regular usage and monitoring

publication date

  • 2017
  • 2017
  • 2017