كليدواژه :
برآورد بارش , تحليل خوشه‌اي , ميانيابي , زمينآمار , مدلسازي روابط مكاني
چكيده فارسي :
بيش از نيمي از جمعيت جهان در مناطقي زندگي مي‌كنند كه بحران آب و بارش در آنجا جدي است. به خاطر مقابله با اين بحرانها، محققان در علم اقليم شناسي به اطلاعات بارندگي، تحليل الگوها و مدلسازي روابط فضايي، داده كاوي مكاني و همچنين برآورد و تخمين بارش به جهت مديريت و مقابله با اين شرايط به شدت نيازمندند. از طرفي ديگر، تنوع مكاني الگوهاي بارش، حاصل عوامل متنوعي همچون موقعيت جغرافيايي، ارتفاع، ويژگيهاي توپوگرافي منطقه مانند شيب و... است كه بررسي اين پديده را دشوارتر نموده و موجب شده كه در مطالعات گذشته به طور جامع به آن پرداخته نشود يا تنها تاثير برخي عوامل همانند ارتفاع بررسيگردد. از اينرو در اين مقاله قصد داريم تا با كمك ابزارها و تكنيكهاي مختلف علم زمينآمار و با درنظرگرفتن عوامل محيطي و مكاني مختلف موثر بر بارش، راهكار مكان مند جامعي براي توصيف اين عارضه جغرافيايي توسعه دهيم. از اينرو، حوضه آبريز درياچه اروميه به دليل شرايط بحراني كه در ساليان اخير با آن مواجه بوده است، به عنوان منطقه مطالعه موردي انتخاب گرديد. ابتدا بارش ابتدا با كمك انواع روشهاي سنتي و زمينآمار، ميانيابي شده و كريجينگ معمولي با كمك اعتبارسنجي تقاطعي و RMS برابر با 4.15 به عنوان بهترين روش انتخاب شد. سپس با كمك روش هاي مختلف آمار فضايي، ازجمله تحليل خوش هاي، نواحي جنوب و جنو ب غربي اين درياچه به عنوان قسمتهاي داغ و پربارش شناسايي شدند. در پايان نيز، از جهت مدل سازي روابط مكاني، رگرسيون عمومي بر بارش برازش داده شد و متغير عرض جغرافيايي به عنوان تاثيرگزارترين متغير وابسته بر بارش شناسايي گرديد
چكيده لاتين :
More than half of the world's population lives in areas where the water crisis and rainfall are serious. To cope with these crises, climatology researchers require rainfall data, pattern analysis, and rainfall estimation and management in order to manage and cope with these conditions. Iran is located in the middle belt in dry belt, which is characterized by low rainfall and high evapotranspiration. The average rainfall in the country is 250 mm and is subject to severe spatial and temporal changes. Variety of spatial factors such as position, elevation, topographic characteristics such as slope and aspect are the most effective factors in the spatial variation of rainfall. Each of these characteristics is able to determine the pattern of precipitation behavior.
Therefore, in this paper, the aim is to develop a comprehensive mechanism for describing this geographical problem with the help of various earth sciences tools and techniques and considering the various environmental and spatial factors affecting rainfall. The Basin of Urmia Lake is one of the most important and most valuable aquatic ecosystems in Iran and the world. The ecosystem of this lake is a typical example of a closed basin that all runoff drain in this basin. The catchment area of Lake Urmia is selected as a case study due to the critical situation that has been facing in recent years. At first, the synoptic data of 21 stations of the Meteorological and Adventure Organization of the Ministry of Energy are used. This data is collected during the period of 63 years of statistical period from 1951 to 2014, and then the annual precipitation rates of the stations are calculated as the dependent variable based on these statistics. In addition data, longitude, latitude, height and slope of each station as well as the average annual and average annual wind speed were also extracted as independent variables.
First, initial statistical tests (rainfall data normalization at stations, data normalization, trend review and deletion) were performed. Then, a combination of traditional and statistical methods are reviewed and examined. As a result, the ordinary Kriging method was selected with RMS equal to 4.15. Then, with the help of different analytical and spatial methods, including cluster analysis, the southern and southwestern regions of this lake as hot and high-frequency parts as well as low-low cold spots in the northern and central parts of the basin Lake Urmia and two spots in the Sarab and Salmas areas with low concentration of rainfall were identified in this area. At the end, in order to model spatial relationships, general regression was fitted to rainfall, and the latitude is obtained as the most effective dependent variable. In addition, longitude and wind speed are detected as the least effective variables on precipitation in the lake of Urmia. The results of this paper have shown that land survey methods are more accurate than traditional methods for locating Lake Urmia.
