صحت‏سنجی بارش ‏های برآوردی از رادار و کالیبراسیون ضرایب برآورد رادار کرمانشاه

میزان بارش اندازه‏گیری‏شده توسط رادار با میزان بارش دریافت‏شده در سطح زمین تفاوت دارد. در نتیجه، برای تعیین میزان بارش با استفاده از داده‏های رادار، باید داده‏های رادار را بر اساس داده‏های زمینی تصحیح کرد. پیوند میان بارش و بازتابندگی رادار نمایی است از z=arb . اگر مقادیر ضرایب این مدل درست انتخاب نشود، برآورد مقدار بارش توسط رادار با اشتباه همراه می‏شود. در این پژوهش بارش‏های 26 تا 27 آبان 1394 و 10 تا 12 آذر 1395 در ده ایستگاه‏ استان کرمانشاه بررسی شد. در بارش اول زاویه ارتفاع بهینه پرتو انتخاب و رابطه مربوط به آن استخراج شد و ضرایب تصحیح به‏دست آمد. با استفاده از این رابطه، مقدار بارش برآوردی رادار از 31درصد به 96درصد افزایش یافت و میانگین مجموع بارش برآوردی رادار از 8.9 به 32.4 میلی‏متر رسید که از میانگین واقعی فقط 1 میلی‏متر کمتر است. در بارش دوم، با استفاده از داده‏های بارش رادار، فقط یک معادله استخراج شد و ضرایب تصحیح رادار به‏دست آمد. نتایج برآورد بارش رادار به این روش نیز مورد قبول بود و میانگین مجموع بارش برآوردی رادار از 9.6 به 23.5 میلی‏متر افزایش یافت که 4 میلی‏متر از مقدار واقعی کمتر بود.

Validation of radar rainfall estimated and Calibration of Kermanshah Radar Coefficients

Introduction Kermanshah Province is one of the western mountainous provinces located in the middle Zagros mountains.The rainfall in this province is similar to other mountainous regions orographic and rugby. Forecasting rainfall in terms of severity, amount and continuity in the usual way is often not precisely possible and requires very expert forecaster and familiarity with local conditions.This leads to severe, destructive and sometimes catastrophic floods in the province. Today, weather radars can be predicted by experts as a valuable tool for forecasting rainfall, provided that they are calibrated in accordance with local conditions and calibrated over time with climate change. If the calibration of weather radar is good, radar can estimate the amount of rain over the vast areas with good accuracy. Relationship between rain and reflectivity radar exponential Z=aRb, where a and b are coefficients radar. And the amount of R depends on factors such as: the type of rain, rain season, latitude and topography. Variation range of a is few tens to few hundred, and the range of variation b is 1 to 3. Drop size and distribution of rainfall, the coefficient of this. Materials and methods The purpose of this study is to measure the Kermanshah weather radar with the conditions of the province, so that the specialists predict that using this radar can predict the characteristics of rainfall before its occurrence, provide warnings to the people and authorities and the damage to the dwellings Historic buildings, administrative centers, urban and rural facilities, farms, humans and animals to the extent possible In this study, rain from 17 to 18 November 2015 and 30 November to 2 December 2016 at the stations, Kermanshah, Eslamabad, Sarpol, Ghasre Shirin, Harsin, Javanroud, Tazabad, Songhor, Ravansar, Ghilan Gharb and Soumar at distance of 30 to 100 kilometers from Kermanshah’s radar are investigated. Results and discussion Linear regression method using hourly rain data and the amount of reflectivity, linear equations and the coefficient R2, reflectivity, respectively. For rainfall of 17 to 18 November 2015 for all stations different angle radar beam and the reflectance was measured. And the amount of rain gauges in the same period was obtained. The radar beam elevation angle optimized for each station, a separate line equation was obtained. Using the coefficients of the equation and radar, rain intensity level was determined Surface Rainfall Intensity (SRI) and the total rainfall was an hour and finally total rain was estimated for each station. Radar estimated rainfall amounts for the first part was more than the amount measured by rain gauges. Because the error was found in all stations, it is assumed that the size and distribution of rainfall in the first part of rainfall vary with the next step. So to solve this problem, rain at all stations was divided into two parts. For the first part of the conflict between rainfall radar and rain gauge rainfall, there was a common linear equations and coefficients a and b were obtained again for all stations. The amount of rainfall was estimated again. A result showed that the amount of rainfall was estimated in this way is better than before. Use this relationship to estimate the amount of rain 31 percent to 96 percent and the average total rainfall radar estimated that 8.9 to 32.4 millimeter Increased than before calibration. and only was 1 millimeter less than of the actual average obtain by gauge. Since acquiring the optimum beam angle the radar for any location, timeconsuming and difficult. So when the time is not enough, it is better for each rainfall radar obtained a relationship. So in the following for rain 30 November to 2 December 2016, using rainfall data of all stations, for correction factors radar, only one equation was obtained. and The results of this way was good, and the mean rain radar to estimate rainfall radar calibration 9.6 to 23.5 millimeter increased and only 4 millimeter less than of the actual rain that obtain by gauge. Conclusion The results showed that the coefficients radar for any location and at any time is different. So to increase the accuracy of radar rainfall it is better for each separate equation obtained. However, to obtain a separate equation for each location is the best. Finally, it is suggested that the radar equation coefficients for each region of Iran, which is covered by the weather radar, are calculated, so that they can accurately predict the precipitation and give warnings to the different centers.