基于敏感性因素分析的数值天气预报修正方法
摘要:模型输出统计作为数值天气预报(numerical weather prediction，NWP)修正常用的方法，可有效修正系统误差，节省计算资源和时间。为了提高NWP的预测精度，从影响NWP精度的因素出发，分析不同因素对NWP误差的影响模式，得到对NWP风速误差影响最大的因素。然后以该敏感因素为划分依据建立分段修正模型提高NWP的准确性。为防止研究结果依赖于某种特定建模方法，采用4种常用的统计方法建立模型。算例结果表明：NWP风速误差对风向最为敏感，以风向为划分依据建立的修正模型精度最高。以线性回归方法为例，以风向为划分依据比以风速和气压为划分依据的NWP误差分别降低了3.6%和5.3%。
Abstract：As a common method of numerical weather prediction (NWP) correction, model output statistics can effectively correct system errors and save computing resources and time. In order to improve the prediction accuracy of NWP, this paper analyzes the influence modes of different factors on NWP error, Starting from the factors that affect the accuracy of NWP, and obtains the factors that have the greatest impact on NWP wind speed error. Then based on the sensitive factors, a piecewise correction model is established to improve the accuracy of NWP. In order to prevent the research results from relying on a specific modeling method, four commonly used statistical methods are used to establish the model. The results of numerical examples show that the NWP wind speed error is the most sensitive to the wind direction, and the modified model based on the wind direction has the highest accuracy. Taking the linear regression method as an example, the NWP error based on wind direction is 3.6% and 5.3% lower than that based on wind speed and air pressure.
标题：基于敏感性因素分析的数值天气预报修正方法
title:Correction Method of Numerical Weather Prediction Based on Sensitivity Factor Analysis
作者：丛智慧,刘倩,刘永前,韩爽
authors:CONG Zhihui,LIU Qian,LIU Yongqian,HAN Shuang
关键词：数值天气预报(NWP)修正,风速误差,敏感性分析,风向,
keywords:numerical weather prediction (NWP) correction,wind speed error,sensitivity analysis,wind direction,
发表日期：2020-11-25
- 文件大小:
- 2.88 MB
- 下载次数:
- 60
-
高速下载
|
|
