Share Email Print

Proceedings Paper

Application limit of Landsat ETM images to detect Saxaul plant community in desert ecosystems
Author(s): Adel Sepehry; Hassan Hassanzadeh
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Application of satellite remote sensing imageries in studying desert ecosystems is crucial normally due to desert extend, its harsh environment and difficult access which make studying and monitoring of desert ecosystems a cumbersome task. Black Saxaul (Haloxylon Aphyllum), as a resistant plant, is widely planted in Iran’s central deserts to prevent sand dune movement and protect arable lands, roads and buildings from sand debris. An attempt was made to study the effect of Saxaul plantation in Kavir-e-Omrani, in reducing wind erosion and stabilizing sand dunes after 30 years of its plantation in an area of 31627 hectares. Landsat 7 ETM+ imagery acquired on March, 2002 was used to study Saxaul community extent and its canopy cover percentage classes to be related to the field measurements of soil sedimentation depth along prevailing wind direction and through canopy cover percentage gradient of Saxaul community. It was therefore necessary to have canopy cover percentage classes of the community obtained via classification of the ETM images, and their derivative bands. All ETM bands were registered to 1:50000 topographic maps of the area and 10 GCPs obtained by filed measurements using GPS. Correction was made on digitized maps using linear transformation and nearest neighbor method for resampling with RMS error of less than 8 meters. 53 sampling units of 90m by 90m were field checked and canopy cover percentage, density (number of Saxaul per unit area) and ground canopy cover of accompanying plants were measured. Soil samples of ground surface were obtained within each sampling units for lab analysis of soil texture, EC (electric conductivity) and ESP (exchangeable sodium percentage). Coordinates of the corners of sample units were recorded using GPS so that positional discrepancies of sample units were minimized. 16 different vegetation indices, including RVI, NDVI, SAVI, MSAVI and other indices were created. Mean DN values of all ETM bands (except panchromatic band) and 16 derivative images related to 53 sampling units were extracted for statistical analysis. Principal Component Analysis and Correlation Analysis showed no meaningful correlation between canopy cover percentage classes of Saxaul and DN values of all ETM bands as well as 16 derivative images. To examine capability of ETM bands and their 16 derivative images to differentiate between Saxaul sites having more than %75 of canopy cover and sites with bare soil, discrimination of the two sites were tested via Student T test. A series of unsupervised classification was also performed on FCC image with 2, 3, 16 and unlimited number of classes using ISODATA clustering method to find if Saxaul plantation could be classified as a distinct class. Results showed no distinction between the two sites. Visual investigation of all images proved the statistical results. Despite the fact that the image was acquired in a season with highest Saxaul greenness and LAI, It was found that ETM images are unable to detect Saxaul plant community. It seems, inability of ETM to discriminate Saxaul plant community from surrounding bare soil is due to Saxaul prevailing bark percentage comparing to its LAI, its reduced leaf surface area as well as its pubescent leaf structure which seems to let beneath soil reflectance prevail upper plant cover reflectance.

Paper Details

Date Published: 24 February 2004
PDF: 8 pages
Proc. SPIE 5232, Remote Sensing for Agriculture, Ecosystems, and Hydrology V, (24 February 2004); doi: 10.1117/12.507910
Show Author Affiliations
Adel Sepehry, Gorgan Univ. of Agricultural Sciences and Natural Resources (Iran)
Hassan Hassanzadeh, Gorgan Univ. of Agricultural Sciences and Natural Resources (Iran)

Published in SPIE Proceedings Vol. 5232:
Remote Sensing for Agriculture, Ecosystems, and Hydrology V
Manfred Owe; Guido D'Urso; Jose F. Moreno; Alfonso Calera, Editor(s)

© SPIE. Terms of Use
Back to Top