| Abstract | The purpose of this study is to develop a new vegetation classification methodology,
especially with a focus on the monitoring of primary and secondary forest resources in Indo-
China Peninsula using integrated multi-resolution satellite data.
This new classification methodology will assist in the processes of monitoring and
analysis of primary forest, secondary forest, and other vegetation resources in the study area.
which includes Thailand, Myanmar, Laos, Cambodia, and Vietnam.
A new vegetation index named Normalized Thermal-NDVI (NT -NDV1), which is a
combination of optical a.nd thermal bands, has been developed. This new vegetation index
makes it possible to achieve better classification accuracy of primary forest, secondary forest,
active agriculture land, harvested land, and bare soil, which could not be clearly classified by
existing NDVI. Furthermore, this new vegetation index is stable over large area, so it is
possible to apply same thresholds through out the study area.
An integration technique for low and high resolution satellite data has been developed.
High resolution remotely-sensed data (Landsat TM 30 m. resolution) is used as ground truth to
decide precise thresholds for the classification of low resolution data. Because the same
thresholds can be applied for the whole Indo-China Peninsula, and can be decided from only
few of high resolution remotely-sensed data, this new technique can simplify the process of
detemiining thresholds which needs very ambiguous decision when only low resolution data is
used.
Low spatial resolution but high temporal resolution and large coverage remotely-sensed
data (NOAA AV1-IRR LAC 1.1 km. resolution), which is suitable for monitoring purpose of
large area, is used as the main source of data to cover the whole study area.
Finally, vegetation classification for Indo-China Peninsula was obtained using newly
developed NT-NDV1 and multi-resolution satellite data integration technique.
In conclusion, this study contributes to the development of vegetation classification
methodology particularly for monitoring of primary and secondary forest resources in Ludo-
China Peninsula. |