干旱气象

• 论文 • 上一篇    下一篇


基于垂直温度梯度对WRF-RUC模式山东2 m温度订正研究

夏凡李昌义刘诗军   

  1. (山东省气象科学研究所,山东济南250031)
  • 出版日期:2018-03-05 发布日期:2018-03-05
  • 作者简介:夏凡(1986—),男,工程师,从事数值天气预报与模式数据质控工作. E-mail:xf0718@126.com。
  • 基金资助:

    山东省气象局青年基金项目(2017SDQN08)、山东省气象科学研究所数值天气预报应用技术开放研究基金(SDQXKF2015M04)和山东省气象局面上项目(2016sdqxm08)共同资助

Research on Revision of 2 m Temperature Forecasted by WRF-RUC Model in Shandong Based on Vertical Temperature Gradient


XIA Fan, LI Changyi, LIU Shijun   

  1. (Shandong Meteorological Scientific Institution, Ji’nan 250031, China)
  • Online:2018-03-05 Published:2018-03-05

摘要:

为了减小观测站实际地形高度与对应模式地形高度的差异给2 m预报温度带来的误差,基于垂直温度梯度设计两种订正方案:第1种方案假定整层大气温度垂直梯度为-0.006 5 ℃·m-1;第2种方案计算了距离观测站实际地形高度最近的两层模式温度间的垂直梯度。利用两种方案对天气与预报快速跟新循环(简称WRF-RUC)模式2016年10—12月山东逐日2 m预报温度进行订正并对比订正前后的预报效果,对高度差异最大的泰山站2 m预报温度的订正效果进行研究。结果表明:(1)山东大部分观测站的实际地形高度与对应模式地形高度差异较小,导致两种方案订正效果不明显;(2)对泰山站2 m预报温度的订正,第1种方案在预报前期有正效果,在预报后期出现负效果;第2种订正方案在整个预报时段呈现正效果;(3)WRF-RUC模式2 m预报温度存在负的系统偏差,导致第1种方案在预报后期对泰山站2 m预报温度的订正存在负效果,第2种方案有效克服了逆温影响,对泰山站2 m预报温度的订正效果更优。

关键词: WRF-RUC, 2 m温度, 地形高度, 垂直温度梯度, 泰山

Abstract:

Based on vertical temperature gradient, two schemes were designed to reduce the forecast error originated from the difference between observed terrain height and model terrain height. The first  scheme was  on basis of a hypothesis that the vertical temperature gradient of the whole atmospheric layer is -0.006 5 ℃·m-1, while the vertical temperature gradient was calculated between the two model layers which were closer to the observed terrain height, and it was utilized by the second method. These two schemes were used to make revision on day-to-day 2 m forecast temperature of WRF (weather research and forecast)-RUC (rapid update cycle) model during the period from October to December in 2016 in Shandong. The revised and unrevised 2 m forecast temperature of WRF-RUC model were compared and some research work was focused on 2 m forecast temperature at Mount Tai station. The results were as follows: (1)The difference between observed terrain height and model terrain height in most of observing stations in Shandong was so small that revision effect of two schemes was unnoticeable. For stations with observed terrain height being lower than or equal to model terrain height, the RMSE calculated by the second scheme was significantly greater than that calculated by the unrevised and the first scheme, so the revision effect of the second method was negative. (2)The revision effect on Mount Tai 2 m forecast temperature of the first schemes was positive in former forecast period, while it was negative in later forecast period, and that of the second method was positive in the whole forecast period. The relative error of 2 m forecast temperature at Mount Tai station revised by the second method maintained between -2 and 2 ℃. (3)The negative systematic deviation of WRF-RUC on 2 m temperature caused negative revision effect in later forecast period by using the first scheme, the second scheme was more skillful on Mount Tai 2 m forecast temperature revision because it diminished the influence of temperature inversion.

Key words: WRF-RUC, 2 m temperature, terrain height, vertical temperature gradient, Mount Tai