Categorization and Identification of Acacia Trees Based on Species in El Ain Natural Forest Reserve by Using MultiTemporal Landsat Imagery and Spectral Angel Mapper Classification

Abstract

Identification and classification of Acacia trees cover based on their species in arid and semi-arid areas through
remotely sensed images involves various considerations, processes and techniques. Efforts to remotely sense arid land
vegetation are often hindered by high reflectance of the soil background, mixtures of green and senescent grasses, and
the occurrence of shrubs in grasslands. Elain forest reserve area, in North Kordofan state, Sudan, endures intensive
different heterogeneity of Acacia trees cover and is highly sensitive to climate fluctuations and human intervention.
Mapping of tree based on their species in these areas needs more input of high advance techniques under condition of
appropriate forest management strategies. The main objective of current paper was to determine the applicability of the
spatial and spectral resolution of Landsat imagery and sup pixel classification in combination with ancillary data and
field sampling to distinguish and discriminate Acacia species distribution at Elain Forest Reserve in North Kordofan
State. The multi-temporal landsat imagery (1987,1994 and 2016) , Minimum Noise Fraction (MNF) and Pixel Purity
Index (PPI) were applied to extract different spectral signatures of Acacia trees in the study area. The Spectral Angel
Mapper Classification (SAM)- was applied using selected endmmbers from PPI results . The paper provides a reliable
comparable method of performing Multispectral Processing of landsat data using ENVI for Hyperspectral
classification and mapping Acacia trees based on their species in arid regions .