CMA-MESO模式对塔里木盆地暖季降水的预报性能评估
牟欢 1赵丽 2杨霞 1陈春艳 1李如琦 1孙鸣婧1
作者信息
- 1. 新疆维吾尔自治区气象台，新疆 乌鲁木齐 830002
- 2. 新疆信息工程学校，新疆 乌鲁木齐 830013
- 折叠
摘要
塔里木盆地地形复杂，暖季降水预报的不确定性显著。为了深入认识中国气象局区域高分辨率数值模式（CMA-MESO）在塔里木盆地的降水预报性能，基于2023—2024年暖季CMA-MESO预报数据与地面观测资料，采用多种降水检验指标，系统评估了该模式对塔里木盆地暖季降水的预报偏差。结果表明：CMA-MESO总体上能较好地模拟盆地暖季降水的空间分布特征。随着海拔升高，TS评分上升，而PC评分下降。降水频率预报偏差幅度在亚高海拔区最为显著，正偏差中心带集中分布在昆仑山西段迎风坡，24 h和1 h降水频率预报最大偏差分别为44.1%和11.6%。TS评分的日变化显著，并呈昼高夜低的特征，这种昼夜偏差在低海拔区最为显著。从降水强度的变化看，CMA-MESO模式对24 h和1 h有效降水的TS评分分别为34.4%和8.8%，但随着降水强度的增加，预报性能迅速下降。值得注意的是，1 h降水预报的概率密度偏差主要集中在0.1~2 mm，针对性订正该区间预报可更有效地降低整体降水频率偏差。研究结果可为CMA-MESO在干旱区复杂地形下的降水预报改进及防灾减灾业务提供科学依据。
Abstract
The complex terrain of the Tarim Basin and active local convective weather during the warm season significantly increase uncertainty in numerical precipitation forecasting, making it difficult to meet the demand for refined precipitation forecasts. To better understand the forecast bias characteristics of the China Meteorological Administration Mesoscale (CMA-MESO) numerical weather prediction model in this region and enhance itsforecast performance, this study systematically evaluated the CMA-MESO models capability in forecasting 24 hour accumulated precipitation and 1 hour accumulated precipitation during the warm season (from May to September) over the Tarim Basin. The evaluation used forecast data from the CMA-MESO model and surface observational data from the 2023-2024 warm seasons, employing multiple operational precipitation verification metrics. The results showed that the CMA-MESO model could generally simulate the spatial distribution characteristics of precipitation in the basin during the warm season. As the altitude increased, the Threat Score (TS) rose, while the Probability of Correct detection (PC) decreased. The forecast bias in precipitation frequency was most significant in the moderate-elevation zones, with a positive bias center zone concentrated on the windward slope of the western Kunlun Mountains. The maximum forecast bias in 24 hour accumulated precipitation frequency and 1 hour accumulated precipitation frequency was 44.1% and 11.6%, respectively. The diurnal variation of the TS was significant, characterized by higher values during the day and lower values at night; this diurnal bias was most pronounced in the low-altitude areas. In terms of precipitation intensity, the TS for effective 24 hour accumulated precipitation and 1 hour accumulated precipitation in the CMA-MESO model were 34.4% and 8.8%, respectively. However, the forecast performance declined rapidly with increasing precipitation intensity. Notably, the overestimation in the frequency of 1 hour accumulated precipitation was primarily confined to the 0.1~2 mm range, and targeted corrections to forecasts in this interval could more effectively reduce the overall precipitation frequency bias. The results of this study provide a scientific basis for improving the precipitation forecasts of the CMA-MESO model in complex terrain in arid areas and support disaster prevention and mitigation operations.关键词
塔里木盆地/暖季/CMA-MESO/降水/预报评估Key words
Tarim Basin/warm season/CMA-MESO/rainfall/forecast evaluation引用本文复制引用
牟欢,赵丽,杨霞,陈春艳,李如琦,孙鸣婧.CMA-MESO模式对塔里木盆地暖季降水的预报性能评估[EB/OL].(2026-07-23)[2026-07-28].https://chinaxiv.org/abs/202607.00193.学科分类
大气科学（气象学）/灾害、灾害防治
