干旱气象 ›› 2026, Vol. 44 ›› Issue (3): 490-498.DOI: 10.11755/j.issn.1006-7639-2026-03-0490

• 技术报告 • 上一篇    下一篇

新疆地形约束大气加权平均温度模型研究

陈志强1,2(), 刘琦2,3,4, 王洋2,3,4, 拜杰2,3,4, 刘杰1,2,3,4()   

  1. 1 新疆农业大学交通与物流工程学院新疆 乌鲁木齐 830052
    2 新疆交通规划勘察设计研究院有限公司科技研发中心新疆 乌鲁木齐 830006
    3 新疆高寒高海拔山区交通基础设施安全与健康重点实验室新疆 乌鲁木齐 830006
    4 新疆交通投资(集团)有限责任公司新疆 乌鲁木齐 830006
  • 收稿日期:2025-07-08 修回日期:2025-12-25 出版日期:2026-06-30 发布日期:2026-07-16
  • 通讯作者: 刘杰(1986—),男,甘肃酒泉人,博士,正高级工程师,主要从事地质灾害防治研究。E-mail: hfutliujie@163.com
  • 作者简介:陈志强(2002—),男,河南信阳人,硕士生,主要从事地质灾害防治研究。E-mail: 577062482@qq.com
  • 基金资助:
    新疆维吾尔自治区“天山英才-青年拔尖人才”计划项目(2023TSYCCX08);新疆交通投资(集团)有限责任公司科技创新团队项目(XJJTZKX-FWCG-202312-0476);新疆交通投资(集团)有限责任公司科技创新团队项目(XJJTZKX-FWCG-202312-0477)

Research on a terrain-constrained weighted mean temperature model for Xinjiang

CHEN Zhiqiang1,2(), LIU Qi2,3,4, WANG Yang2,3,4, BAI Jie2,3,4, LIU Jie1,2,3,4()   

  1. 1 School of Transportation and Logistics EngineeringXinjiang Agricultural UniversityUrumqi 830052, China
    2 Xinjiang Transportation PlanningSurvey and Design Institute Co., Ltd, Technology Research and Development CenterUrumqi 830006, China
    3 Xinjiang Key Laboratory for Safety and Health of Transportation Infrastructure in Alpine and High-altitude Mountainous AreasUrumqi 830006, China
    4 Xinjiang Transportation Investment (GROUP) Co.LtdUrumqi 830006, China
  • Received:2025-07-08 Revised:2025-12-25 Online:2026-06-30 Published:2026-07-16

摘要:

研究新疆复杂地形区大气加权平均温度(Tm)模型对提高干旱半干旱地区全球导航卫星系统(Global Navigation Satellite System,GNSS)反演大气可降水量(Precipitable Water Vapour,PWV)精度具有重要意义。基于新疆9个无线电探空站2021—2024年观测数据,按照地貌特征将研究区划分为山脉区、混合区和盆地区,采用最小二乘回归方法,引入地表气温、高程、年及半年周期项,以及地表气温与高程耦合的地形约束因子,构建了适用于不同地貌分区的大气加权平均温度模型(Terrain-Constrained Temperature Model,TC-TM)。结果表明,TC-TM在新疆地区的偏差、均方根误差和平均绝对误差平均值分别为-0.28、3.34、2.64 K,整体精度优于Bevis、GPT3_1和莫智翔模型;其中,地形约束因子的引入有效增强了模型对复杂地形区温度垂直结构差异的表征能力。在GNSS-PWV反演应用中,TC-TM对应的PWV反演RMSE和相对误差全区均值分别为0.15 mm和1.18%,均优于对比模型。基于地形约束的分区建模策略能够有效提高复杂地形区Tm估算精度及GNSS-PWV反演质量。

关键词: 大气加权平均温度, 复杂地形, 地形约束, 全球导航卫星系统反演大气可降水量

Abstract:

Investigating the weighted mean temperature of the atmosphere (Tm) over Xinjiang is of great significance for improving the accuracy of global navigation satellite system (GNSS)-derived precipitable water vapor (PWV) in arid and semi-arid regions. Based on observational data from nine radiosonde stations in Xinjiang during 2021-2024, the study area was divided into mountain, mixed, and basin regions according to geomorphological characteristics. The paper used the least-squares regression method to develop a terrain-constrained temperature model (TC-TM) by incorporating surface temperature, elevation, annual and semiannual periodic terms, as well as a terrain-constrained factor coupling surface temperature and elevation. The results show that the mean Bias, root mean square error (RMSE), and mean absolute error of the TC-TM model over Xinjiang were -0.28, 3.34, and 2.64 K, respectively, outperforming the Bevis, GPT3_1, and MZX models. The introduction of the terrain-constrained factor effectively enhanced the model’s capability to characterize differences in vertical temperature structures over complex terrain. In GNSS-PWV retrieval applications, the mean PWV retrieval RMSE and relative error of the TC-TM model over the entire region were 0.15 mm and 1.18%, respectively, both lower than those of the comparison models. The results demonstrate that the terrain-constrained regional modeling strategy can effectively improve Tm estimation accuracy and GNSS-PWV retrieval performance in complex terrain regions.

Key words: weighted mean temperature (Tm), complex terrain, terrain-constrained, GNSS-PWV

中图分类号: