روشي جديد جهت پيش‌بيني پراكنش مكاني دما و بارش در حوضه آبريز رودخانه قره‌سو (اردبيل)

A new method for the forecasting of Spatial Distribution of Precipitation and Temperature in Gharehsoo River Watershed

استفاده از پراكنش‌هاي مكاني بارش و دما نقش مهمي در افزايش دقت خروجي مد‌ل‌هاي هيدرولوژيكي دارند. هدف از اين مقاله تهيه پراكنش‌هاي مكاني دما و بارش در آينده در حوضه آبريز رودخانه قره‌سو است. حوضه آبريز مورد مطالعه در شمال غرب كشور و در استان اردبيل قرار دارد. اين حوضه آبريز از نظر توليد محصولات كشاورزي در ايران داراي اهميت بسيار است. در تهيه پراكنش‌هاي مكاني بارش و دما از روش‌هاي درونيابي شامل روش‌هاي وزني عكس فاصله، توابع پايه شعاعي(RBF)، مكاني چند جمله‌اي و كريجينگ از نرم‌افزار ArcGIS استفاده شده است. بدين منظور ابتدا داده‌هاي ماهانه بارندگي و دما در حوضه آبريز رودخانه قره‌سو با استفاده از 10 ايستگاه هواشناسي در سال 2004 تهيه شد، سپس به منظور انتخاب روش مناسب براي تهيه پراكنش‌هاي مكاني بارش و دماي حوضه آبريز كارايي روش‌هاي زمين آمار مورد بررسي قرار گرفت. با محاسبه شاخص‌هاي ميانگين خطا و ريشه ميانگين مربعات خطا و مقايسه، روش وزني عكس فاصله مناسب‌ترين روش براي تهيه پراكنش مكاني دما و روش RBF براي تهيه پراكنش‌هاي مكاني بارش در اين حوضه شناخته شده است. در صورتيكه با كمك روشي بتوان پراكنش‌هاي مكاني بارش و دما در آينده را تهيه كرد، مي‌توان پيش‌بيني‌هاي دبي را با استفاده از مدل‌هاي هيدرولوژيكي انجام داد. در اين مقاله الگوريتم روشي بيان شده كه مي‌توان به كمك آن پراكنش‌هاي مكاني بارش و دما در آينده را تهيه كرد. براي پيش‌بيني پراكنش‌هاي مكاني دما و بارش در آينده نياز به يك مدل پيش‌بيني‌كننده متغيرهاي آب و هوايي است كه در اين مقاله از داده‌هاي مدل اقليمي منطقه‌اي PRECIS استفاده شده است. خروجي داده‌هاي مدل PRECIS با قدرت تفكيك 50×50 كيلومتر بر اساس سناريوي B2 از سري سناريوي SERS و براي سال‌هاي 2071 تا 2100 است. نتايج پراكنش‌هاي مكاني دما در حوضه نشان مي‌دهد كه دما در تمامي حوضه آبريز رودخانه قره‌سو نسبت به دوره پايه بين 2 تا 5 درجه سانتيگراد افزايش مي‌يابد و همچنين نتايج پراكنش‌هاي مكاني بارش در حوضه به دليل افزايش و كاهش در ماه‌هاي مختلف سال روند خاصي را تسبت به دوره پايه نشان نمي‌دهد.

Introduction: Precipitation and temperature patterns have especial role in the accuracy of hydrologic models. The future patterns of rainfall and temperature can lead to better hydrological predictions. Hence, according to their importance, we try to derive the future rain and temperature patterns of the Gharehsoo River’s watershed. This watershed has been placed in the northwest of Iran in Ardebil province and it has high amount of agriculture productions. Interpolation schemes are utilized in this study to determine the rain and temperature patterns. The utilized software package is ArcGIS software. These interpolation techniques are included of Inverse Distance Weighting (IDW), Global polynomial, Local polynomial, Radial Basis Functions (RBF), ordinary Kriging and simple Kriging. Firstly, we gather the monthly temperature and precipitation data of 10 synoptic stations in 2004. Then, the interpolation schemes are evaluated in order to determine the best temperature and precipitation patterns. The evaluation criteria in this study were Root Mean Square Error (RMSE) and Mean Error (ME). The results of evaluation of different interpolation schemes demonstrated that IDW and RBF method are the best schemes for the spatial modeling of temperature and precipitation patterns, respectively. Using these patterns, it is straightforward to predict runoff using hydrological models. In this paper, a new algorithm is proposed for the prediction of temperature and precipitation patterns for future (2100). To predict temperature and precipitation pattern, it is necessary to utilize of a predictor  model  to predict the amount of precipitation and temperature. Then the amount of precipitation and temperature are converted to spatial pattern of precipitation and temperature using the developed algorithm in this study. PRECIS model that is a regional climate model was utilized as predictor model in this study.   Materials and methods: a) case study: The studied area (Gharehsoo river watershed) is located in the Northwest of Iran, between longitudes coordinates 47°45’ and 48°42’ E, and between latitude coordinates 37°46’ and 38°34’ N. The Gharehsoo river watershed area is approximately 4100 km2 and plays significant agricultural role in Iran. the mountainous areas have been located in the western and southeastern parts of watershed. Furthermore, there are many pasture and agriculture lands in this watershed. Watershed elevation varies from 1170 m in Gharehsoo river outflow to 4732 m in Sabalan mountainous. The precipitation in the watershed is highly related to the topography and wind in the watershed.. The sea fronts and orographic conditions are the main factors for precipitation in the western and eastern parts of watershed. In the winter, the cold front of Mediterranean Sea, coupled with the local effects of Sabalan Mountains lead to orographic rainfalls. In summer, weather conditions are predominant of Caspian Sea front is the major factor for precipitation in the eastern part of catchment. Autumn and spring rainfalls are the results of interaction between AfricanMediterranean and Caspian Sea fronts. b) Data: Temperature and precipitation data are two basic climatologically variables, measured at meteorological stations. Monthly precipitation (mm) and temperature data for 2004 was provided through Iran Meteorological Organization. The number of stations in the watershed and near to watershed was 11 stations. c) PRECIS Model PRECIS (Providing Regional Climates for Impacts Studies) is a regional modeling system that can be run over any area of the globe on a relatively inexpensive, fast PC to provide regional climate information for impacts studies. The idea of constructing a flexible regional modeling system originated from the growing demand of many countries for regionalscale climate projections. Only a few modeling centers in the world have been developed RCMs (Regional Climate Models) and utilize them to generate projections over specific areas, because it needs high amount of computational effort and time. The Hadley Centre has configured the thirdgeneration of Hadley Centre RCM, named PRECIS that is easy to set up. The past (19611990) and future climate SRES B2 scenario (20712100) were simulated by PRECIS model at a spatial resolution of 50×50 km for Iran.   Results and discussion: It’s necessary to have a series of precipitation and temperature patterns to produce monthly patterns for future. These series of maps are generated using the precipitation and temperature patterns of 2004. The hyetograph maps are calculated by the ration of total volume of precipitation in January and the area of watershed. The calculated total volume of precipitation in January using the precipitation pattern map was about 490 million m3. The ration of volume and the area of watershed was about 0.117 m. This number shows the average precipitation of January. Similarly, these operations can be performed for the other months of 2004. The unit hyetograph and thermograph maps are generated by dividing the precipitation and temperature patterns in 2004 to their corresponding monthly precipitation and temperature values. The precipitation and temperature data were extracted from the PRECIS model for 2100. The monthly temperature data of 2100 shows an increase of temperature about 2 to 5 degrees in future, but there is no specific trend in precipitation data. If the amount of the monthly temperature and precipitation of 2100 are divided by these amounts in 2004, the amount of B parameters are calculated for precipitation and temperature in different months. Finally, the precipitation and temperature patterns will be obtained by the product value of B parameters and unit hyetograph or thermograph maps in each month, respectively.   Conclusion: A new method was developed for reasonable prediction of spatial patterns of precipitation and temperature. This new method uses of the results of a Regional Climate Model (e.g. PRECIS model) coupled with the appropriate spatial modeling techniques (interpolation techniques). The derived precipitation and temperature patterns in 2100 in Gharehsoo River watershed show a reasonable similarity with the topography and the climate of the region, Hence This method can be introduced as an appropriate method for the hydrological forecasts and water resource management.