Vegetation mapping and characterization in West Siang District of Arunachal Pradesh, India – a satellite remote sensing-based approach

Vegetation mapping is a primary requirement for various management and planning activities at the regional and global level. It has assumed greater importance in view of the shrinkage and degradation in forest cover. Usage of remotely sensed data for mapping provides a cost-effective method. In the...

Full description

Saved in:
Bibliographic Details
Published inCurrent science (Bangalore) Vol. 83; no. 10; pp. 1221 - 1230
Main Authors Singh, T. P., Singh, S., Roy, P. S., Rao, B. S. P.
Format Journal Article
LanguageEnglish
Published Current Science Association 25.11.2002
Subjects
Artificial satellites
Coniferous forests
Deciduous forests
Degraded forests
Forest cover
Mixed forests
RESEARCH ARTICLE
Riverbanks
Subalpine forests
Tropical forests
Vegetation
Online AccessGet full text

Cover

More Information
Summary:Vegetation mapping is a primary requirement for various management and planning activities at the regional and global level. It has assumed greater importance in view of the shrinkage and degradation in forest cover. Usage of remotely sensed data for mapping provides a cost-effective method. In the present study vegetation cover assessment has been done using remotely sensed data in West Siang District of Arunachal Pradesh. Standard method was adopted for ground data collection by establishing the correlation between satellite data and various vegetation types. Ground data were collected extensively and sufficient information was obtained. Vegetation classification was performed using traditional methods of image recognition. The discrimination among the various forest types is restrained on satellite data owing to the environmental set-up, intermixing of species/vegetation and topography. However, to achieve higher accuracy, other methods have been considered. Hybrid approach of classification has been adopted where modification of spectral classification with the aid of ancillary data set has been found useful. The study area has been classified into twenty-three categories. The vegetation cover types extracted from classification showed good relationship with altitudinal zones. Correspondence with field-gathered GPS points for vegetation classes showed 85.29% overall accuracy. Hybrid classification approach gives an opportunity to refine the classification to acceptable limits for various activities related to management and planning.
ISSN:0011-3891