This research has developed a method for detecting changes in vegetation cover and state, accounting for change direction, magnitude and extent, in regions of frequent cloud cover or data gaps. The study integrated high-temporal, low spatial-resolution data from MODIS with moderate spatial-resolution sensors (Landsat TM / ETM+ and ASTER), to negate the issue of cloud cover that frequently creates significant gaps even in long-term data-sets such as the Landsat series. Absolute correction of the moderate resolution imagery was implemented using 6SV in order to achieve comparable surface reflectance measurements. Variations from the normal observed behaviour of MODIS vegetation indices time-series data were identified, and utilised to target further change analysis in Landsat and ASTER imagery, giving indications of the potential temporal and conditional changes in vegetated land cover; vital information required for the monitoring of key protected habitats. Results indicate that both abrupt conversions and gradual conditional changes are identifiable over annual time-steps with the ability to report these changes annually.
|Journal||IEEE International Geoscience and Remote Sensing Symposium Proceedings|
|Publication status||Published - 28 Jul 2014|
- remote sensing
- vegetation mapping
- spatial resolution