بررسی موردی عملکرد مدل Wrf جهت پیش‌یابی میدان باد تراز 10 متر و دمای تراز دومتر با استفاده از داده‌های ماهواره‌ای و ایستگاه‌های همدیدی در منطقه دریای عمان و دریای عرب

در پژوهش حاضر شبیه‌سازی پیش‌یابی و ارزیابی میدان‌ باد تراز 10 متر و همچنین دمای ‏تراز دومتر با مدل میان‌مقیاس ‎wrf‎‎‏ نسخه 1. 7. 3 در منطقه دریای عمان و دریای عرب انجام شده است. تمامی شبیه‌سازی‌ها برای دو پیکربندی متفاوت از دیدگاه پارامترسازی فیزیکی که به نام‌های phys1 و phys2 معرفی شده‌اند‏، انجام شده است. ارزیابی نتایج شبیه‌سازی با استفاده از داده‌های ایستگاه‌های همدیدی در دسترس و همچنین داده‌های دو سنجنده ماهواره‌ای ‎ascat‎‏ و ‎oscat‎‏ انجام شده است. برای بررسی عملکرد مدل عددی از سنجه‌های آماری مانند ضریب همبستگی پیرسون، ریشه میانگین مربعات خطاها و نمودار آماری تیلور استفاده شده است. نتایج نشان می‌دهند که در بخش ارزیابی‎ با داده‌های ماهواره‌ای باد تراز 10 متر نتایج عملکرد مدل‏، قابل قبول و مناسب بوده است. در این بخش شبیه‌سازی‌های برای مناطق دور از ساحل که به‌کمک داده‌های ماهواره‌ای ارزیابی شده‌اند‏، نسبت به داده‌های ایستگاه‌های همدیدی ساحلی و نزدیک ساحل، به مقادیر مشاهدات نزدیک‌ترند. نتایج حاصل از ارزیابی دمای تراز دو متر نیز گواه عملکرد مناسب مدل در این بخش است.

A Case study of WRF model performance to hindcast of 10m wind and 2m temperature against the satellite and synoptic stations data over the Gulf of Oman and The Arabian Sea

Reliable and sufficient information of 10m wind and temperature fields over open seas and near coastlines is a necessary and important data that has impact on many marine activities. Assimilation of NWP models can be used to assess an estimation of these fields. This study reports the performance of the weather research and forecasting (WRF) model to hindcast 10mwind and temperature fields that were evaluated under two different physical options of planetary boundary layer (PBL) and surface layer (SL) for an area over the Gulf of Oman and the Arabian Sea. The case study includes 16 simulations of 8 different days from WRF model version 3.7.1. The WRF model is configured with two nests. Parent nest has 0.3 degree and the inner nest has 0.1 degree horizontal grid resolution. The grid spacing of the inner domain is almost 11km. The LatLon (latitudelongitude) method is used as the map projection method. For all domains and all runs 39 terrain following vertical levels are set. The validation of the simulated fields is done considering two observational datasets (the weather stations for 10m wind and 2m temperature and satellite instruments just for 10m wind). Nearsurface observations of 2m temperature and 10m wind speed and direction are collected from 55 weather stations, located within the chosen area. The measurements from satellite instruments have become an important source of data in the regions that insitu observations are sparse like seas and oceans, hence observations from two different scatterometers (ASCAT and OSCAT) are also used to evaluate 10m wind simulations. Moreover, in order to better understand the model performance for different choices of the physical schemes, sensitivity of the model has been investigated. There is plenty of choices for the combination of parameterization schemes available for WRF model; for the current study two configurations are taken from other’s previous published research works. The physical parameterization that used in this study are Revised MM5 and MoninObukhov for surface layer and MRF and MYJ for planetary boundary layer. These choices are used to create two different configurations called Phys1 and Phys2. Comparison between winds from satellite scatterometer and simulated winds show very little difference and hence good agreement with observations. Acceptable accuracy has been obtained from statistical analyses. These analyses demonstrate that the maximum average RMSE of wind field is 2.39 m/s, based on results of comparing with ASCAT data and it is 2.37 m/s, based on results of comparing with OSCAT data. The analyses also show that simulation of wind fields have better results over offshore regions than coastlines weather stations. The outcome shows that the simulated 10m ‎wind‎ present acceptable general skills over the sea. The validation of 2m temperature presents that the model has a proper estimation about temperature field over the coasts and near coastal station within the simulation domain. The maximum average RMSE of temperature field is 2.6 degrees of centigrade. Finally, without any justification to run WRF for longer periods from a quantitative and qualitative assessment of the results, it can be concluded that for the WRF model has an acceptable performance to simulate 10m wind and 2m temperature over the Gulf of Oman and the Arabian Sea. It should be noted that to verify these results for longer periods more similar experiments must be tested.