یکی از مشکلات اصلی مناطق خشک و نیمه‌خشک حاکمیت پدیده بیابان زایی است. بنابراین، شناخت و پیش‌بینی عوامل مؤثر در پیشرفت پدیده بیابان‌زایی می‌تواند در مدیریت بهتر این مناطق مؤثر واقع شود. هدف از این تحقیق ارزیابی صحت مدل شبکه عصبی مصنوعی در پیش‌بینی روند بیابان‌زایی و انتخاب مؤثرترین معیار بیابان‌زایی در دشت دهلران با استفاده از مدل ایرانی ارزیابی وضعیت بیابان‌زایی (IMDPA) است. در این روش دو معیار آب و اقلیم به عنوان عوامل مؤثر در بیابان‌زایی انتخاب شدند. برای معیار اقلیم سه شاخص بارش سالانه، شاخص SPI و تداوم خشک‌سالی و برای معیار آب پنج شاخص افت آب، نسبت جذب سدیم، کلر، هدایت الکتریکی و کل مواد محلول در آب ارزیابی شد. با استفاده از مدل مذکور هر شاخص امتیازدهی شد. سپس با میانگین هندسی نقشه‌های معیار و شدت بیابان‌زایی در نرم‌افزار ArcGIS®93 برای دوره مورد نظر تهیه شد. در نهایت داده‌ها به شبکه عصبی مصنوعی جهت پیش‌بینی وارد شدند. نتایج نشان‌دهنده کارایی بالای مدل‌های شبکه عصبی مصنوعی در پیش‌بینی روند بیابان‌زایی بود به گونه‌ای که دقت شبکه بالای 80 درصد و میانگین مربعات خطا کمتر از یک بدست آمد. همین‌طور بر اساس نتایج بدست آمده برای دوره پیش‌بینی شده مهم‌ترین معیارهای احتمالی تأثیرگذار بر شدت بیابان‌زایی منطقه به ترتیب معیارهای اقلیم و آب با متوسط‌ وزنی 2 (متوسط زیر کلاس 1، 2 و 3)، 84/1 (متوسط زیر کلاس 1 و 2) رتبه‌بندی گردیدند. تفاصيل المقالة

Rangelands are important for plant productivity, livestock production, wildlife,conservation of soil and water resources, and etc. One of the main problem of rangeland isthat has not been used based on its potential that leads to more degradation of rangelands.The purpose of this study was to evaluate the range management capability of Badrehregion in Ilam province, Iran, using ANP (Analytic Network Process) and GIS(Geographic Information Systems) techniques. For this regard, firstly, the network ofeffective factors in evaluation was designed. Four clusters including vegetation cover,topography, pedology, and geology were divided into number of sub-criteria. Fordetermining the relations among these clusters and sub-criteria, a number of questionnairesdistributed among the experts and used to obtain their judgments about the relativeimportance of each criterion in rangeland capability. In the next step, based on the limitedsuper matrixes the final weight of nodes was calculated. The weights of nodes inevaluating process were extracted by calculating the geometric mean of the questionnaireweights, as well. After determining the weights of nodes, they were transformed to datalayers. Finally, ecological capability map for range management was provided using WLC(Weight Linear Combination) technique in GIS. The results showed that 3.00, 21.76,58.46, 16.79 percent of the study area had very good (or excellent) condition (as firstclass), good condition (second class), fair condition (third class), and poor condition(fourth class) for capability of range management, respectively. تفاصيل المقالة

The major aim of processing satellite images is to prepare topical and effectivemaps. The selection of appropriate classification methods plays an important role. Amongvarious methods existing for image classification, artificial neural network method is ofhigh accuracy. In present study, TM images of 1987, and ETM+ images of 2000 and 2006were analyzed using artificial fuzzy ARTMAP neural network within Mehrgan region,Kermanshah province, Iran, with an area of 5957 ha changes in range cover state in thisbasin during 3 periods of time from 1987 to 2000 and 2000 to 2006 were examined. In thisstudy, initially, Land sat data for intended years were corrected geometrically andradiometric ally. Next, different land use classes were defined and training samplesobtained via field visits. The obtained results show that, over time period of 1987-2000, theextent of low-density rangeland and farmland in study region had been increased by 89.09and 321.08 ha, respectively, while good rangeland and fair rangeland faced a decliningtrend of 358.29 ha and 48.89 ha. Also, during time period of 2000-2006, the extent of poorrangeland and farmland within study region has increased by 64.98 and 727.12 ha,respectively, while good rangeland and fair rangeland faced a declining trend of 144.01 haand 648.1 ha. Accuracy of vegetation maps resulting from satellite data classification usingalgorithm of artificial fuzzy ARTMAP neural network was 90.97% and 94% for TM(1987) images and ETM+ (2000,2006) respectively which indicates high accuracy ofARTMAP algorithms for classifying satellite. Therefore, this study proves high efficiencyand potential of artificial fuzzy ARTMAP neural network for classification of remotesensing images. تفاصيل المقالة

Using satellite imagery for the study of Earth's resources is attended by manyresearchers. In fact, the various phenomena have different spectral response inelectromagnetic radiation. One major application of satellite data is the classification ofland cover. In recent years, a number of classification algorithms have been developed forclassification of remote sensing data. One of the most notable is the decision tree. The aimof this study was to compare three types of decision trees split algorithm for land coverclassification in Doviraj catchment in Ilam province, Iran. For this, propose, first, thegeometric and radiometric corrections were performed on the 2007 ETM+ data. Field dataas training sites were collected in the various classes of land use. The results of imageclassification accuracy assessment showed that the Gini split classification. With kappavalue 89.98 and the entire accuracy 91.17% was significantly higher, then categorization ofbranching and the branching ratio and Entropy with kappa values of 88.45 and 90.65 andthe entire accuracy of 86.21 and 86.15%, respectively. تفاصيل المقالة