首页|一种针对模式预报场的精细化插值新方法

一种针对模式预报场的精细化插值新方法

扫码查看
原文链接
  • 万方数据
  • 维普
随着智能网格预报业务的深入发展,为了不断提高网格预报产品的时空精度和准度,同时让网格和站点预报产品保持一致性,通过研究提出一种既考虑了静态真实地形又考虑了大气动态垂直变率的快速插值方法(Fast Refinement Interpolation,简称 FRI),并选择 2022年 1月 1日到 2022年 3月 31日的 ECMWF和 CMA-GFS、CMA-MESO三个模式的 3～36 h预报产品作为试验数据,开展了个例天气过程试验、FRI参数选取试验和长期插值检验对比.从长期插值检验和个例试验来看,FRI方法相对双线性插值方法具有明显优势,空间上,FRI方法能提高近地面 2m气温的空间插值精度,插值结果更加符合地形变化情况,时间上,FRI方法较双线性插值方法的 2m气温插值结果准确率明显提高,特别是在具有复杂下垫面的中西部地区.方法中的修正参数还能被作为检验模式三维预报场垂直变率的一种指标.FRI方法插值快速高效,具有明确的物理解释意义,它的提出为更精准的预报近地面气象要素提供了重要理论支撑.
A New Fast Refinement Interpolation Method for Model Forecasting
With the continuous development of grid forecasting operations with higher resolution and more accuracy,the spatial and temporal accuracies and precision of grid forecast products need to be continuously improved.To maintain compatibility between the grid and site forecasts,FRI(fast refinement interpolation)method was explored,which considers the static real-terrain and dynamic atmospheric vertical variation process.In this study,3-36-h forecast products for three models(ECMWF,CMA-GFS,and CMA-MESO)from 1 January 2022 to 31 March 2022,were selected as the experimental data.Further,the results of individual weather process tests,FRI parameter selection tests,and long-term interpolation tests were compared.The results of long-term interpolation tests and individual case trials showed that the FRI method has a clear advantage over the bilinear interpolation method.From a spatial perspective,the FRI method can improve the spatial interpolation accuracy of 2-m temperature near the ground,and the results are more consistent with topographic variations.From a temporal perspective,the FRI method significantly improves the accuracy of the 2-m temperature interpolation results compared with the bilinear interpolation method,especially in the western region with complex subsurface.Additionally,the loss parameter in the proposed method can be used as an indicator to check the three-dimensional atmospheric vertical variability of the model product.The FRI method is fast and efficient and offers a clear physical meaning,which provides important theoretical support for the more accurate reflection of near-surface meteorological forecast information.

Spatial interpolationInterpolation for the forecast field2-m temperatureRefined forecastingGlobal and regional weather models

曾晓青、曹勇、王玉、郭云谦

展开 >

国家气象中心,北京 100081

空间插值 预报场插值 2m气温 精细化预报 全球和区域天气数值模式

国家重点研究发展计划项目国家重点研究发展计划项目新疆维吾尔自治区重点研发专项中国气象局重点创新团队项目

2021YFC30009032022YFC30029042022294426CMA2022ZD04

2024

10.3878/j.issn.1006-9895.2302.22104

大气科学
中国科学院大气物理研究所

大气科学

CSTPCD北大核心
影响因子：2.11
ISSN：1006-9895
年,卷(期)：2024.48(5)