大气科学学报2026，Vol.49Issue(2)：324-335,12.DOI:10.13878/j.cnki.dqkxxb.20240910002
四川省地基毫米波云雷达产品数据质量对比评估与分析
Comparative assessment and analysis of ground-based millimeter-wave cloud radar data quality in Sichuan
摘要
Abstract
Assessing data quality is essential for the effective application of new atmospheric observation instru-ments and is critical for ensuring the accuracy and reliability of meteorological datasets.With the rapid develop-ment of China's ground-based remote-sensing vertical observation network,millimeter-wave cloud radar(MMCR)has been widely deployed in operational meteorology.Although the co-location of MMCR and radio-sonde systems provide favorable conditions for data quality evaluation,comprehensive multi-stations using long-term datasets remain limited. This study evaluates the quality of MMCR products from seven stations in Sichuan Province using MMCR measurements,L-band radiosonde profiles,and surface precipitation data collected from 2023 to 2024.Cloud lay-ers retrieved from radiosonde profiles serve as the reference,with adjustments made to account for radiosonde horizontal drift and precipitation-induced signal attenuation.Overall,the MMCR demonstrates strong consistency with radiosonde observations in detecting cloud occurrence and vertical structure,with agreement rates exceeding 65%at all stations.In high-altitude plateau regions,the MMCR more effectively penetrates thin cloud layers,capturing finer vertical structures and yielding a notably higher frequency of multi-layer cloud detections compared with radiosonde retrievals. Analysis of cloud-height retrievals indicates that MMCR-derived cloud-base heights(CBH)exhibit higher accuracy than cloud-top heights(CTH)at all stations.This discrepancy is attributed to attenuation of millimeter-wave signals by atmospheric water vapor,which weakens returns from higher altitudes.Consequently,cloud-height error characteristics differ between plateau and basin regions.For example,large CTH errors occur at low-altitude basin stations(Wenjiang,Yibin,Dachuan),whereas the smallest mean CTH error(-0.27 km)is found at the high-altitude Ganzi station.When evaluated in 1 km altitude basins,the MMCR performs best for CBH is between 1 and 3 km and CTH between 8 and 9 km,with reduced accuracy when CBH exceeds 6 km or CTH is below 2 km.Environmental influences were further examined by calculating vertically integrated tempera-ture and relative humidity from radiosonde profiles up to the observed cloud heights and correlating these with cloud-height errors.Both temperature and humidity significantly affect MMCR retrieval accuracy.A strong negative correlation is found between temperature and cloud-height error,with warmer conditions corresponding to larger negative CBH and CTH biases.In addition,increasing relative humidity leads to increasingly negative CTH errors,especially when integrated relative humidity exceeds approximately 60%.Under such conditions,CTH errors in basin regions show a pronounced negative bias,suggesting that an RH(relative humidity)thresh-old near 60%may serve as a useful indicator of potential CTH underestimation. In summary,this study incorporates atmospheric temperature and humidity profiles to assess the applicability and performance of MMCR across diverse environments and cloud conditions.Nonetheless,uncertainties associated with radiosonde-based cloud retrieval algorithms may influence the evaluation.Future work will integrate additional observation sources—such as satellite and lidar measurements—to further validate MMCR cloud retrievals and enhance data quality assessments.关键词
毫米波云雷达/L波段探空/云垂直结构/数据质量评估Key words
millimeter-wave cloud radar/L-band radiosonde/cloud vertical structure/data quality assessment引用本文复制引用
蹇宛霖,景号然,胡春,季承荔,谢晓林,郭越凡,刘小波,杨珂珂..四川省地基毫米波云雷达产品数据质量对比评估与分析[J].大气科学学报,2026,49(2):324-335,12.基金项目
中国气象局青年创新团队项目(CMA2023QN11) （CMA2023QN11）
中国气象局创新发展专项(CXFZ2024J057) （CXFZ2024J057）
四川省气象局多波段气象雷达组网应用技术研究创新团队项目(SCQXQNCXTD202402) （SCQXQNCXTD202402）
2025年四川省气象局复盘总结专项(SCQXFPZJ2025-06) （SCQXFPZJ2025-06）
高原与盆地暴雨旱涝灾害四川省重点实验室2025年度科技发展基金项目(SCQXKJZD202502) （SCQXKJZD202502）
