Iranian Journal of Remote Sensing & GIS
Year:2013 | Volume:4 | Issue:4|Papers Count:7

Routing issues have many applications in GIS. The main goal in routing is to find the best route passing a series of determined positions. This process has an effective role in many of spatial decision makings. A lot of studies have been accomplished about the optimization of different kinds of routing problems. Traveling Salesman Problem is one of the very old issues in applied sciences. Emerging of new technologies such as GIS caused a lot of new applications for the TSP. A lot of new approaches have been suggested using these technologies to solve TSP. Evolutionary algorithms are one of the methods have been used to solve various optimization problems. Researches have shown that combining local search methods with the genetic operators can lead to better results in solving TSP. In this paper, a new and innovative method is presented to solve this routing problem. In this method using the concept of mass center for a set of station points a growing convex polygon is assigned to them. During this process a reliable solution, which is a route, is produced. Besides, a local search method based on the genetic algorithm and a nearest neighbor algorithm has been implemented to evaluate the results of the new developed method in solving TSP. Evaluation of the results of the genetic method, local search and nearest neighbors indicate that the new method can provide acceptable solutions during very short time.

One of the physical quantities that could be acquired from remote sensing data is surface reflectance in different regions of electromagnetic reflected spectra. Almost all the urban pixels that imaged by low/medium spatial resolution sensor systems represent composite radiance fields emanating from several distinct features with different reflectance's within the sensor’s field of view. A general approach for describing urban environments is using classification methods and unmixing models, and among the useful models is spectral mixture analysis in which the mixed pixel reflectance is a linear combination of the reflectance and fractions corresponding to land cover types within the pixel. This model can be grouped to spectral unmixing and spatial unmixingnes. The objective of spatial unmixing models is to determine the reflectance spectrum of the classes in each mixed pixel. In this paper, spectral reflectance of important urban classes is determined using spatial unmixing model. To this end, spatial information from IKONOS imagery and spectral information from Hyperion data have been employed. Spatial resolutions of these images are 4m and 30m, and the numbers of bands are 4 and 242, respectively. The validity of the proposed method has been substantiated through comparison of original Hyperion image with the reconstructed image. Afterwards, using the obtained spectral reflectance, an image with 4m spatial resolution and 136 bands has been produced and variability of urban land covers has been taken into account.

Accessing the long-term rainfall statistics in most of regions is restrained due to the lack of rainfall stations and drawbacks in recording data. Developing the artificial neural networks leads to reconstruct the missing information in dispersed meteorological stations even within the years in which the recording stations were absent. In this research the average monthly rainfall data including 36600 collected records for 305 meteorological stations were employed to reconstruct the missing monthly average rainfall values. The selected study area is a vast part in north west of Iran; including the provinces Ardebil and West East Azerbaijan and the models have been run for a decade between 1995 and 2004. The network structure has been based on two hidden layers and trained using the back propagation error algorithm. The longitude, latitude, elevation, slope, month number (1-12) and the average monthly rainfall values for 1 to 26 nearest neighbors for each station have been used as the model input variables. The results showed that using 5 nearest neighboring stations gives the highest precision and the obtained correlation coefficient during 1995-2004 is 0.86 for the training data. The corresponding value of reconstructed data for West and East Azerbaijan and Ardebil provinces were 0.82, 0.78 and 0.70 during 1985-1994 respectively. The reason of such differences in the mentioned provinces arises from the fact that the number of stations taking part in training process decreases from West Azerbaijan to Ardebil. The interpolated maps have then been created using different geostatistics methods based on the output predicted estimates resulted from neural networks. Consequently, this method allows investigating the spatiotemporal distribution of rainfall along with creating the monthly interpolated maps for years 1985 to 2004 in the study area. For validating the interpolation methods, the cross validation has been employed and the Mean Biased Error and Mean Absolute Error and Root Mean Square Error have been compared.

The study of distribution of temperature patterns associated with different land uses and their influences on the temperature range is of great importance in understanding the urban microclimate. Nowadays it can be said that human activities and changes in the natural landscape of the city, lead to variance of temperature between the downtown and its suburbs. Satellite imagery can be used for measuring and monitoring these changes. The purpose of the current research is to study the temporal and spatial changes of urban thermal patterns in Shiraz, through satellite images and determine locations with high temperatures and based on point in order to map and analyze city land use. The used Method is based on Satellite imagery LANDSAT TM and ETM+ sensor to extract the thermal pattern of the user as well as the mapping to land uses monitor changes in the city for 24 years has time frame. The results of this study showed that surfaces soil barren and devoid of vegetation the highest temperature in the city suburb of user's texture than other compact urban areas is higher than the temperature, thermal and rings with urban traffic areas and are adaptable. Well as for the coldest area of vegetation is adapted to the period also witnessed a decline trend of user changes to User Account in favor of the city are barren.

Careful assessment of biomass for forest management and understanding the role of forests as carbon sources is very important. Since 50% of forest biomass contains Carbon, estimating the forest biomass in terms of the amount of energy stored in the review of forests and climatic change is highly regarded. The best way for estimating biomass is the developed models which are (based upon) the local collected data. Recently the remote sensing techniques with minimum ground measurements have been played an important role in forest biomass estimation. Lack of the local equations for estimating biomass and ambiguity precision of the global equations, Leads the study toward the development of biomass equations for the forests of Guilan region, in northern Iran. To develop these equations, in a field work, 28 plots (each plot 30*30m) have been selected and the various parameters such as height and diameter of trees have been measured. Non-destructive method is used to estimate the biomass of plots. Then, by using the optical and radar ALOS satellite images, the relationship between remote sensing data and the estimated biomass of plots are determined. The multivariate regression analyses are used to establish the mentioned relationship. The best model based on the highest coefficient ((RÙ2=0.731 and the lowest Root Mean Squared Error (RMSE=13.88(ton/he)) are determined.

In this study, 14 spectral (vegetation) indices have been calculated from different groups, using spectral bands. The correlation of these vegetation indices with canopy cover has then been determined for three rangeland types and one forest type. The percentage of canopy cover in all vegetation types is measured using step-point method in radial direction (6000 points per vegetation type). The correlation of these vegetation indices with canopy cover is then assessed at the scale of vegetation type and across the vegetation types using SPSS and Erdas Imagine softwares. Different groups of vegetation indices, including slope-based, distance-based, orthogonal transformations and plant-water sensitive ones are extracted from Landsat TM satellite reflectance data and evaluated against field cover data. The results indicated that vegetation indices had higher relationships within vegetation types up to 81% (p<0.001) but the correlations across vegetation types was decreased due to diversity of vegetation spectra and the effects of background features. Overall, the results indicated that each spectral index was suitable for mapping a specific vegetation type because of structural and biological differences in vegetation cover as well as background soil and these characteristics must be considered for selecting the best indices for mapping and stratification of vegetation types.

Considering the canopy closure as one of the main biometric parameters in forest management, it is important to find a suitable method to estimate it. Comparing to field measurements, studying this characteristic on the remotely sensed data is more precise and accurate because of its view from top. For this purpose, dot grid is among the most applicable methods applied on remotely sensed data. It was aimed to evaluate the accuracy and precision of this method on UltraCam-D aerial imagery. A 30 ha study -area was selected in Zagros forests in Kohgilouye BoyerAhmad province and it was fully callipered. Afterwards, different dot grids were obtained based on sampling error percents (5 to 20) and applied in GIS. The statistical comparison of the results showed that changing the number of points in the dot grids did not affect the accuracy of them and the results of all grids (except 9 and 14%) did not show statistical difference with the ground truth, although their precision increased by increasing the number of points. Generally, it can be concluded that dot grid has the efficiency of estimating the canopy closure percent but there is no logical relation between the precision and accuracy of it.