Samaneh Sadat Mahzooni-Kachapi; Ataollah Ebrahimi; Pejman Tahmasebi; Mohammad Hassan Jouri
Volume 29, Issue 4 , January 2023, , Pages 608-626
Abstract
One of the most important methods of extracting information from satellite data is various image classification techniques. The current research was conducted to separate and classify plant ecological units by classification tree analysis algorithm on satellite images and also visual interpretation of ...
Read More
One of the most important methods of extracting information from satellite data is various image classification techniques. The current research was conducted to separate and classify plant ecological units by classification tree analysis algorithm on satellite images and also visual interpretation of Google Earth images in one of the semi-steppe rangelands of Chaharmahal and Bakhtiari province. After applying the classification algorithm in Idrisi TerrSet software, the software generated the error matrix, and then based on the values inside this matrix, the extracted statistics were evaluated. The results of visual interpretation showed that finally, seven types of plant ecological units that were different in terms of structural features were identified and expressed as descriptive statistics, including Astragalus verus, Bromus tomentellus, Scariola orientalis, Astragalus verus-Bromus tomentellus, Astragalus verus -Stipa hohenikeriana, Bromus tomentellus-Stipa hohenikeriana, and Stipa hohenikeriana. The results also showed that the overall accuracy and kappa coefficients of 0.92% and 0.89 for Landsat 8 images and 0.94% and 0.92 for sentinel 2 images were achieved. Based on the obtained results, it was found that satellite images and aerial images have a suitable separation capability to prepare a map of plant ecological units. Since the visual interpretation of Google Earth images is a time-consuming and expensive method, it can be concluded that satellite images, especially Centile 2, can be used as a tool due to their high resolution and high-resolution images, Practical and usable, provide accurate information and details of the earth's surface phenomena and be used as a suitable source for preparing a map of plant ecological units.
Jamal Imani; Ataollah Ebrahimi; Bahram Gholinejad; Pejman Tahmasebi
Volume 28, Issue 4 , November 2021, , Pages 640-651
Abstract
In the present study, we compared different sampling patterns and different plot dimensions to estimate the percentage of canopy cover and forage production in rangeland habitats around Choghakhor Wetland. The choice of sampling method was based on the opinion of the researcher. ...
Read More
In the present study, we compared different sampling patterns and different plot dimensions to estimate the percentage of canopy cover and forage production in rangeland habitats around Choghakhor Wetland. The choice of sampling method was based on the opinion of the researcher. Sampling was performed in three different plant communities, in two patterns of six and three plots. Different dimensions of the plot (including 1×1, 1×2, 2×2, and 3×3 m2) were used as a nest to estimate the production and the percentage of canopy cover. Sampling was in each community within 30 sampling units along three transects. The canopy cover of the species was estimated by estimating and producing them by double sampling. Species density was calculated by counting bases in 2×2 plots. In each population, the type of distribution of dominant species was determined by statistical tests. The results showed that two different sampling patterns and four different plot dimensions have significant differences in estimating the percentage of cover and plant production (P≤0.01). According to the interaction of the community and the pattern, plot dimensions with pattern and plant community, patterns and plot dimensions with each other, there was no significant difference. The effect of plant community with plot dimensions on a cover percentage at (p≤0.05) and production at (p≤0.01) is significant. In Gundellia tournefortii-Couisinia Bakhtiari plant community with random distribution patterns, two different sampling patterns in 1×1 and 1×2 m2 plots were significantly different, and in other dimensions, no significant difference was observed. The two different sampling patterns in the Daphnea mucronata-Astragalus adsendence community with the uniform dispersion pattern were significantly different only in plot 1×1 (p≤0.05) and شdid not differ in other sizes. The mentioned patterns in the population of Melica persica-Agropyron trichophorum with the heap pattern distribution in statistical plots of 3×3 m2 did not show a statistical difference, but in other dimensions showed a significant difference (p≤0.05).
Jamal Imani; Ataollah Ebrahimi; Pejman tahmasebi; Bahram Gholinejad
Volume 24, Issue 2 , July 2017, , Pages 429-440
Abstract
The use of satellite data is one of the proper methods, which makes studying ecosystems less costly. This research was carried out to determine the correlation among the vegetation cover of dominant species in three sites with the NDVI index. For this purpose, a study was conducted on three different ...
Read More
The use of satellite data is one of the proper methods, which makes studying ecosystems less costly. This research was carried out to determine the correlation among the vegetation cover of dominant species in three sites with the NDVI index. For this purpose, a study was conducted on three different vegetation types. In each region, a range was determined for sampling. Then, within this range and in the horizontal direction, 30 sampling units of 30 x 30 m were selected along three 900-meter transects in a random-systematic manner. In each sampling unit, nested quadrates with dimensions of 1 × 1, 2 × 2 m were placed. In order to prevent geometric errors, the 900-meter ranges were 60 meters apart. Within each sampling unit, geographic coordinates were recorded with GPS. The number of individuals and canopy cover percentage of dominant species were recorded separately in the plots. Then, the correlation of canopy cover percentage with NDVI index was obtained by processing Landsat 8 images. The distribution pattern was also determined using species density data. The results showed that in all species, the correlation coefficient of NDVI index was higher in the plots with higher area. Also, the correlation coefficient with one quadrat increased towards five quadrats. Due to the lack of high correlation between the total canopy cover percentage of a quadrat and the NDVI index, the use of one quadrat inside the pixel is not recommended in any way. Selecting the type of sampling depends on distribution pattern and species size as well as access to facilities and correlation coefficient acceptability. For species with clumped distribution, more quadrats are needed with proper distribution inside the pixel on the ground, so that sampling could be a good representation of the total pixel. In uniform distribution, fewer samples are needed since the whole pixel is the same in terms of plant growth.
Zahra Abolalizadeh; Ataollah Ebrahimi
Volume 22, Issue 1 , June 2015, , Pages 21-30
Abstract
Sabzkouh protected area is located in central Zagros. This area, with a variety of natural ecosystems and landscapes, is very rich from the point of diversity of climate, topography, habitats and wildlife. Its ecosystems, like other semi-arid ecosystems of Iran, have undergone changes in their structure ...
Read More
Sabzkouh protected area is located in central Zagros. This area, with a variety of natural ecosystems and landscapes, is very rich from the point of diversity of climate, topography, habitats and wildlife. Its ecosystems, like other semi-arid ecosystems of Iran, have undergone changes in their structure in recent years. Nowadays, Markov chain models are widely used in ecological studies to predict these changes quantitatively. For this purpose, satellite images and geographic information system were used to predict the future condition of the region using CA-Markov model. Finally, the land cover map of the region for the next fifteen years (2003-2018) was predicted by CA-Markov model. According to the prediction map, an increase in the area of agricultural lands, shrub/brush rangelands and forests of the region was observed, while the area of forbs rangelands and bare lands decreased. The results of these predictions provide a valuable tool to the managers and policy makers of natural resources and environment