Journal of RS and GIS for Natural Resources
Year:2013 | Volume:4 | Issue:2|Papers Count:9

The mean aerosols samples in three periods of ten stations were taken from Bushehr region, to characterize the spatial variability and concentration of As, Cd, Co, Fe, Ni, Pb and V. The geostatistics and geographic information system (GIS) techniques were applied, and the disjunctive kriging was used to map the spatial patterns of the seven heavy metals. Meanwhile, Principal component analysis (PCA) and correlation matrix (CA) were used for the data processing. The results of Nug / Sill ratios for the seven metals showed that spatial dependent is moderate (0.25-0.75), that indicative the effects of natural factors such as parent material and topography. Meanwhile, the disjunctive kriging technique was used to quantify their concentration distribution. Combined with the results of PCA, 7 heavy metals could be divided into 3 factors. D1 was the metals, i.e., As, Co, Ni, Pb, V. Cd was in D2, Fe in D3. This results show the concentrations of 7 heavy metals were mainly controlled by the external factors. These results will contribute to the management of regional environment.

The present study aimed to investigate the changing trends in a part of Marivan County in the area of 18077 ha. Buffer classes of residential, agricultures and lake determined as effective criteria. Land use maps for 1985 and 2005 determined by the supervised classification method. Change area detected with overlaying two land use map. Change area was extracted for each of effective parameters distance buffers measured using ArcGISÒ 9.3 software. The results showed that during the time period of 1985-2005, forest and agriculture land use decreased and residential areas increased. The results also showed that, agriculture land use decreased, and residential areas increased. Furthermore, 1503 ha of forests during 16 years changed to other land uses. 73.3% of forest changes to agriculture lands, a 24.14 % change to range lands and 2.56% of forest lands changed to a residential area, deforested areas have a significant difference in one-percent level with distance from residential areas. Most deforestation areas are about 0-200 meters distance, which includes 602 hectares. Furthermore, deforestation areas in agricultures have 0.870 relevant correlations, and it was decreasing. Deforested area on the lake distances is increasing and has a consequential difference on the five-percent level. Results showed that 900-1200 buffer has the most deforestation.

Land use is among the most important aspects of natural resource management and reviewing environmental changes. The aim of the study, remote sensing combined with GIS techniques, LandSat TM and ETM+ satellite data (1987-2010), were used to detect Jiroft land use changes. Categorizing of land use classes, analysis methods and their changes, were carried out in ENVIÒ 4.7 and ArcGISÒ 9.1 software. Analysis of the images is showing that over the 23 years studied, percent land use changes of four classes (Residential +7.90, Agricultural +42.10, Orchards -17.93 and Bare land -32.07). The results indicate that, the neural network and the support vector machine technique were the least accurate at all the techniques tested. While these techniques, in terms of visual interpretation, showed a better separation in the mentioned uses compared to other techniques. The results showed the level of residential land use has significantly increased. Evaluating the spatial patterns of different land use indicates that the highest changes have occurred in agricultural and orchard areas. The results also shown the importance of land use in areas with agricultural potential refers to reduce the change trend. Improper management of regional hydrology, provision of agricultural inputs and lack of economic stability are factors affecting land use changes in the study area.

This paper proposes ability of CA Markov model in detection and simulating land use/cover change over 65 km of Aghbolagh river. Land use/cover map was prepare using 1956, 1969, 1998 aerial photos and 2006 satellite image’s. Then river, riparian area, agriculture and bare lands around it were illustrated and land use/cover maps of each year were drowned and overlaid through cross routine to evaluate model validation. Finally, next 10 years conditions of these classes were forecasted using CA-Markov model. Results of forecasting future changes based on 1969 and 1998 maps showed that in 2016, 60 percent of riparian area, 26.6 percent of river and 58.8 percent of bare lands and 98.1 percent of arable land will be intact as the current land use. In other words, arable lands will have the most stability while river is the most vulnerable land cover to change.

One of the water management methods that had wide application in recent years is the artificial recharge. The aim of this study is to determine the suitable sites for artificial recharge in Oshtorinan plain located in Boroujerd City. The effective factors of artificial recharge were identified includes: geology, slope, the thickness of the unsaturated zones, groundwater quality, aquifer transmissivity, hydraulic gradient, storage coefficient, distance from surface water sources, and land use. In first, raster maps for different layers were prepared and arranged using the paired comparison method and Analytic Hierarchy Process (AHP). Raster layers were integrated using the Weighted Index Overlay Method. Unsuitable areas according to the proximity to wells, springs and Qantas were eliminated by using the Boolean model, and the artificial recharge map was prepared. Three sites were determined for artificial recharge in Oshtorinan Plain. The potential artificial recharge map shown, 17% of the region area had quite suitable, 21% was suitable, 31% was moderately suitable, 18% was unsuitable and 13% was prefect unsuitable for artificial recharge. Finally, through field visiting four sites were determined for artificial recharge in Oshtorinan Plain.

Land use/cover (LULC) classification using remote sensing images is one of the most important applications of remote sensing technology. Recently, a lot of algorithms have been developed for mapping LULC. The aim of the study was to compare the maximum likelihood and fuzzy approach to mapping land use/cover using LandSat satellite images. After preparing images (radiometric and geometric corrections), with field visiting, land use/cover map were prepared. For determining accuracy assessment training samples were used. Results relating to the overall accuracy of the classification show that the fuzzy classification method possessing kappa coefficient of 99% has higher accuracy than the method of maximum likelihood algorithm possessing kappa coefficient of 98%.

The aim of the study was to produce rangeland vegetation types using LISS-III and ASTER satellite sensors in Deylam area, Bushehr province, Iran. Studying an area has dry Climate and located in the coastal region with 15915 hectare. Geometric corrections of images were applied using ground control points and georeferenced images with RMSE less than one pixel, then images co-registered with together (RMSE<0.2 pixel). The atmospheric corrections of images were applied using subtraction of dark object's method. Image spatial resolution enhanced using fusion with a panchromatic band of IRS P6. Image processing includes classification of images using supervised classification (Maximum Likelihood and Neural Network methods) with 50 training area (each sample is an average of nine plots) to producing rangeland vegetation types, and determining of the accuracy of producing maps with 25 ground truth samples. The results show that both sensors can produce suitable vegetation types map in two years, and didn’t differentiate between producing vegetation type's maps with sensors. Overall accuracy for LISS III and ASTER are 91% and 84.3% for ML and 71.3% and 65.6% for NN classification methods sequentially. The satellite images cannot determine exactly the vegetation type boundary; therefore, the produced maps completed with visual interpretation of images.