بارش رگباری از پدیدههای بحرانسازی است که وقوع ناگهانی و پر شدت آن موجب بروز خسارات فراوان بر انسان و محیط پیرامونی آن میشود. شناخت و آگاهی از بارش رگباری و چگونگی وقوع، شدت و تداوم آن در فصول مختلف کمک بسزایی در مقابلهی صحیح با آن مینماید. با توجه به رفتار متفاوت بارش و تغییرات سریع در فواصل مکانی کم و در طولانیمدت، ارائه مدلهای مناسب ، متناسب با اقلیم آن منطقه، جهت پیشبینی احتمالی آن ضروری است. دراینپژوهشبه تحلیل مکانی بارش رگباری استان مازندران پرداخته شد و براساس دادههای استخراجشده گراف بارانسنجی در 12 ایستگاه سینوپتیک استان مازندران، برابر میزان بارش بالای 10 میلیمتر در دوره 5ساله، از سال 2006 تا2010 مورد بررسی قرار گرفت. برای پهنهبندی محدوده موردمطالعه از روش IDWبا سه توان 1،2،3 و روش کریجینگ با مدلهای کروی، دایرهای، نمایی و گوسین استفاده شده است. ارزیابی و تعیین بهترین مدل و صحت سنجی نقشههای تولیدشده انجام شد. همچنین جهت مقایسه آماری مدلها از مقدار ریشه مربع خطاها RMS ،MAE،RMSE و ضریب همبستگی آنها استفاده شده، که بهترین مدل برای پهنهبندی مدل IDW با دو توان 1،3و کریجینگ معمولی دایرهای بود. استخراج نقشه بهینه بهوسیله رگرسیون چند متغیره بر اساس مدل روش همزمان و روش پسرونده انجام شد و شش متغیر که در ایجاد بارش بیشترین تأثیر را دارند، شامل عرض و طول جغرافیای، تعداد روز بارش، ارتفاع، رطوبت نسبی و دمای نقطه شبنم مورد بررسی قرار گرفت. نتایج نشان میدهد، میزان همبستگی این شش متغیر در فصل بهار97/0، تابستان99/0، پاییز98/0، زمستان99/0ودربارش سالانه99/0 است که نشاندهنده رابطه قوی بین این شش متغیر در بارش رگباری استان مازندران میباشد.
عنوان مقاله [English]
Spatial Analysis of Shower in Mazandaran Province in GIS Environment
Precipitation is an atmospheric factor, its quantity and distribution vary considerably in different parts of the planet, and is one of the most influential climatic elements that has always been influenced by the climate. Its amount changes in time and place continuously.Knowing the temporal and spatial distribution of rainfall is a useful tool for understanding how non-uniform distribution of water resources and vegetation in each region takes place.Precipitation occurs when the wet weather and the climb factor exist both in the region.In other words, the wet air must rise to a certain height so that it can reach the saturation point due to the subsequent cooling down, and in the next, the cloud produces precipitation.The absence of any of these two factors prevents the occurrence of precipitation.
Rainfall variation is considered as a key factor in the structure and functioning of ecosystems, but its impact on scale and magnitude is much less than its spatial variation.The climatic element, especially precipitation, has significant changes in time periods.Therefore, the recognition of the element of precipitation as one of the two elements of the climate and its changes in different times and places allows the optimal utilization of the natural environment.The amountand spatial distribution of rainfall is a fundamental factor for decision making, design and evaluation of hydrological models as well as water management and planning.Temporal spatial variations have diverse and varied impacts on the management and planning of water resources along a water basin.Climate change is one of the factors affecting the change of water resources.Precipitation, as a highly variable element, has always been a concern for climatologists and waterologistsas a fundamental factor in the blue balance. The extreme variability of rainfall along the time-space has a variety of study approaches.The purpose of this research is to identify the conditions of rainfall in Mazandaran province. Therefore, the location of rainfallin this province was investigated.In this regard, identification of the effective factors of the occurrence of these rainfall in different seasons and their role in the province has been addressed and its results will be available as a scientific and practical solution.
Materials and Methods
In this study, for the purpose of identifying the rainfall in the province of Mazandaran, five years of rainfall from 2006 to 2010 have been used from a total of 12 synoptic stations.Using extracted data from precipitation graphs, rainfall of more than 10 mm was extracted in the studied area.Then the data were categorized into four parts: spring, summer, autumn, and winter of the year. To create the database, they entered the SPSS and ARC GIS10 software.In the spatial analysis of the data, the semi-modification of these models has been used, which was calculated using ARC GIS10 software.The methods used in the zoning of Kriging and IDW models for fitting include: IDW with three potentials of 1,2,3, and the Kriging method with spherical, circular, exponential, Gaussian, and spherical models, which is performed with conventional Kriging technique.Also, for statistical comparison of models, root mean square error of RMSE, MAE, RMSE and their correlation coefficient were used.Then, optimal mapping based on multivariate regression was fitted based on the simulation method and the recursive method of six variables in rainfall generation including latitude and longitude, number of rainfall days, elevation, relative humidity and dew point temperature. The effects of these factors on rainfall in the province will be evaluated in different seasons and annually.
The results of the spring survey show that there were 5 stations out of 12 stations without rainfall.These stations are located in the plain and in the mountain range of the region.The analysis showed that the correlation coefficient between variables is R^2= 967, which indicates a strong relationship between the set of independent variables and the dependent variable.85.8% of rainfall in the spring season in Mazandaran province depend on these variables. In the summer, only 2 stations in the province did not experience rainfall ranges, both of which were at high altitudes and include the station Alasht and Kyasar.Variables show a very strong relationship in the summer with a correlation coefficientof R^2=0.995 which is 0.9. 9%of rainfall in Mazandaran province depends on these six variables.The fall season is one of the high seasons in the province of Mazandaran. Only one station (Siahbisheh) has been registered from 12 storm rainfall stations.Estimates show that the six variables analyzed in this chapter with a correlation coefficient of R^2 = 0.983 represent a strong correlation.The results of the winter season show that all stations in Mazandaran province have rainfall, although it includes fewer days than theautumn season.All stations experience at least one day at Alasht Station for up to 7 days in Ramsar.The results of the analysis show that in winter, the correlation coefficient is R^2 = 0.996.
For zoning of the study area, the IDW method with three potentials of 1, 2, 3 and the Kriging method have been used with spherical, circular, exponential and Gaussian models. The evaluation and determination of the best model and verification of the produced maps was carried out. Also, for statistical comparison of the models, the root mean square errors of RMS, MAE, RMSE and their correlation coefficient were used, which, the best model for zoning was the IDW model with two potentials of 1,3 and ordinary circular kriging. Optimal mapping was done by multivariate regression based on the model of synchronous and retrograde method, and six variables that have the greatest effect on rainfall, including latitude and longitude, rainfall days, elevation, relative humidity and dew point temperature were studied.The results show that the correlation values of these six variables are 0.97 in spring, 0.99 in summer, 0.98 in autumn, 0.99 in winter and 0.99 in annual rainfall which indicates a strong relationship between these six variables in the rainfall ofMazandaran province.