You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Proceedings, XXth congress

Seyed Zeynalabedin Hosseini', Sayed Jamaleddin Khajeddin',
Hossein Azarnivand*,
1-College of Natural Resources & Desert Studies, University of Yazd, P.O.Box 89195-741 Yazd, Iran
TeleFax:+98351-8210312 , E-mail: hosseini_sz@yahoo.com
2- College of Natural Resources, Isfahan University of Technology
3- College of Natural Resources, University of Tehran
KEYWORDS: ETM’, Vegetation indices, Fusion, Percentage vegetation cover, Correlation analysis, Stepwise regression.
The major land cover in Iran is rangeland, so it is not possible to assess the vegetation cover of these vast areas through field
work. During the recent years, many researchers have used various satellite data with vegetation indices as a tool to evaluate the
rangelands vegetation cover and to estimate the yield and percentage cover. This study was carried out in Chamestan area rangelands
(12000ha) in Mazandaran province, Iran. Landsat ETM" data with acquisition date of 18th July, 2000 were registered to digital
topographic maps with scale of 1:25000. Multi spectral bands were registered to TM8 panchromatic band using data fusion method
in two ways: “Spectral Response” and “Hue, Intensity, Saturation (HIS)” techniques. In addition, 26 vegetation indices were applied
to Landsat data as well. Field surveys and available vegetation maps confirmed the limited number of vegetation types that were not
exceed to four major vegetation forms at the studied areas. Totally on 40 points, 400 quadrates were placed to estimate vegetation
cover. To arrange these quadrates, on each randomly selected point, 10 quadrates each 10m*10 m were placed on perimeter of a
supposed circle with 20m radius. This supposed circle covers (3*3) 79 pixels of ETM' image. Inside of each quadrate the percentage
vegetation cover was estimated and the dominant species were recorded. Correlation and stepwise regression analysis were applied to
field data and vegetation indices. These indices were the products of the Landsat data bands and the fused bands. The results
illustrated that vegetation indices alone did not have enough ability to estimate percentage cover on studied rangelands. But applying
of vegetation indices to ETM" bands along with fused bands, leads to get reliable results on estimating of rangelands percentage
INTRODUCTION (merging) was done with two methods: “Spectral Response”
and “Hue, Intensity Saturation (HIS)”. In spectral response
method, 2, 3 and 4 bands were merged with pan band. In HIS
During the recent years, implication of satellite data on method all of the bands were merged with pan band. Also pixel
vegetation assessment and land cover mapping has been size of thermal and multi spectral bands were equaled to pan
extended. Field surveying and data collection can be carried out band. Subsequently a lot of 26 vegetation indices were tested
on a part of studied area, then the collected data relation with (Table 1).
satellite data can be modeled to assess the vegetation cover on Field surveys and available vegetation maps confirmed that only
the whole region. According to this point some works have been four major vegetation types can be found in the studied area.
managed. Among them, Longford and et al. (1997) used TM Totally, on 40 randomly selected sample points, 400 quadrates
data to map land cover of Caucau in Colombia. Khajeddin were randomly placed to estimate vegetation cover in a manner
(1995) use Landsat MSS data to assess the vegetation cover of that on each sampling sit, 10 quadrates (10*10m) were placed
Jazmourian area, Iran. Sepehri (2002), applied TM data and on the perimeter of supposed circles with 20m radius to cover 9
vegetation indices, to assess the percentage vegetation cover of ~~ ETM+ pixels (Figure 1). Inside of each quadrates the
rangelands in Jahannema protected area, Iran. Also Leeuwen percentage vegetation cover was estimated and the dominant
and Huete (1996) have employed NDVI and MNDVI on litter species were recorded.
determination and vegetation assessments.
(Table 1) The tested vegetation indices.

The studied area is located in Mazandaran province, northern
part of Iran, between 36° 14'to 36° 17 E and 52° 0'to 52° 15 N IR2 (TM4-TM7)/(TM4+TM7)
about 12000ha where the mean annual precipitation is 1200mm.
Landsat ETM+ images of 18th July, 2000 were registered to MIR TMS/TM7
1:25000 digital topographic maps by nearest neighbor ur
MND (TM4-(1.2* TM3)(TM4* TM3)
resampling algorithm which the RMSe was less than one pixel.
To use the panchromatic band, the multi spectral and thermal MSI IMS/IMA

bands were registered to the pan band, then data fusion