Evaluation and correction of FY-4A cloud cover fraction product in Ningxia

doi:10.11755/j.issn.1006-7639-2025-05-0799

Abstract

Abstract:

The accuracy evaluation of satellite-derived cloud cover is foundation of operational applications, and it is of great significance for effectively leveraging the high spatiotemporal resolution advantages of satellite remote sensing and compensating for the scarcity of ground-based observations. This study focuses on the Ningxia region, a typical arid and semiarid region, and utilizes ground observations of total cloud cover amount at five time points (08:00, 11:00, 14:00, 17:00, and 20:00, Beijing time, the same below) in 2019 to validate and evaluate the Fengyun-4A ( referred to as “FY-4A”) satellite cloud cover fraction product, then uses the normalized mixed correction method to correct the monthly cloud cover fraction products and analyze the daytime spatiotemporal distribution characteristics of cloud cover fraction in Ningxia. The results indicate that: 1) The monthly and daily variation trends of the FY-4A cloud cover fraction product in Ningxia region are highly consistent with those of ground manual observations of total cloud amount. The correlation coefficients between FY-4A satellite-derived cloud cover fraction and ground observed total cloud amount range from 0.7 to 0.9 across the whole region and at individual meteorological stations, but the derived cloud cover fractions are generally lower than the observed total cloud amounts. 2) In terms of the consistency rate between the total cloud amount and cloud cover fraction at five different times of a day, the consistency rates at 11:00, 14:00, and 17:00 are higher than those at 08:00 and 20:00. For cloud cover categories test, the consistency rates are in the order of clear sky > overcast > partly cloudy > mostly cloudy. 3) The normalized mixed correction method effectively reduces the bias between the FY-4A cloud cover fraction product and the manual total cloud amount observations. After correction, the monthly average values of the cloud cover fraction product are in good agreement with the cloud amounts observed at meteorological stations. 4) The annual average daytime cloud coverage in Ningxia ranges from 30% to 70%. Higher cloud cover fractions are found along the Helan Mountains, in the Yellow River irrigation area (from Shapotou District, Zhongwei, to Yinchuan), and in Haiyuan County and Guyuan City. The lowest cloud cover occurs in the eastern parts of Huinong District, Dawukou District, and Pingluo County in Shizuishan City. Cloud cover fraction is generally higher in May and June, and lowest in December.

Key words: FY-4A cloud cover fraction products, validation and evaluation, correction, spatiotemporal distribution

摘要：

开展卫星反演云量的准确率评估是业务应用的基础，对于有效发挥卫星遥感高时空分辨率的优势以及弥补地面观测资料的不足具有重要意义。以处于典型干旱半干旱区域的宁夏为研究区，利用2019年08：00、11：00、14：00、17：00、20：00（北京时，下同）5个时次的地面观测总云量资料对风云四号A星（简称“FY-4A”）云覆盖率产品进行检验和评估，并利用归一化混合订正法对逐月云覆盖率产品进行订正，在此基础上分析宁夏日间云覆盖率的时空分布特征。结果表明：1）宁夏区域FY-4A云覆盖率产品与地面人工观测总云量的逐月、逐日变化趋势具有较高的一致性，全区平均及各气象站FY-4A云覆盖率与人工观测总云量的相关系数普遍为0.7~0.9，但总体上低于总云量值。2）从5个时次总云量和云覆盖率一致率来看，11：00、14：00、17：00的一致率高于08：00和20：00；从云量分级检验来看，一致率依次为晴天>阴天>少云>多云。3）归一化混合订正方法能有效降低FY-4A云覆盖率产品与地面人工观测总云量的偏差，订正后各月的平均值与气象站云量基本吻合。4）宁夏年平均日间云覆盖率为30%~70%，贺兰山沿山、引黄灌区（中卫市沙坡头区—银川一带）、中卫市海原县及固原市云覆盖率较高，石嘴山市惠农区、大武口区东部、平罗县东部云覆盖率最低；5—6月全区云覆盖率普遍较高，12月云覆盖率最低。

关键词: FY-4A云覆盖率产品, 检验评估, 订正, 时空分布

CLC Number:

P426

YUAN Ruirui, WANG Kun, GAO Ruina. Evaluation and correction of FY-4A cloud cover fraction product in Ningxia[J]. Journal of Arid Meteorology, 2025, 43(5): 799-809.

袁瑞瑞, 王坤, 高睿娜. FY-4A云覆盖率产品在宁夏的检验评估及订正[J]. 干旱气象, 2025, 43(5): 799-809.

Figures/Tables 13

Fig.1 The study area and 12 meteorological stations for cloud cover observation

Fig.2 Overall technical process flowchart

Tab.1 Correlation coefficients between FY-4A cloud cover fraction and total cloud amount from meteorological stations for different window sizes

时间	窗口
时间	1×1	3×3	5×5
1月	0.75	0.77	0.76
7月	0.80	0.84	0.83

Fig.3 Monthly variation of total cloud amount and FY-4A cloud cover fraction in Ningxia and five representative stations in 2019

Tab.2 Monthly average values evaluation of total cloud amount and FY-4A cloud cover fraction in Ningxia and 12 meteorological stations in 2019

气象站	相关系数（R）	平均绝对误差（MAE）/%	均方根误差（RMSE）/%
全区平均	0.97**	9.32	10.00
惠农	0.94**	10.58	11.29
陶乐	0.94**	16.48	17.12
银川	0.97**	11.24	11.61
吴忠	0.94**	5.50	6.44
盐池	0.92**	15.45	16.29
同心	0.95**	6.07	7.02
中卫	0.81**	28.01	29.11
中宁	0.88**	21.31	22.20
海原	0.91**	9.41	11.03
固原	0.95**	13.29	14.11
西吉	0.91**	6.09	7.19
六盘山	0.91**	5.29	7.90

Fig.4 Daily variation of total cloud amount and FY-4A cloud cover fraction in Ningxia and five representative stations in 2019

Tab.3 Daily average values evaluation of total cloud amount and FY-4A cloud cover fraction in Ningxia and 12 meteorological stations in 2019

气象站	相关系数（R）	平均绝对误差（MAE）/%	均方根误差（RMSE）/%
全区平均	0.94**	6.00	6.60
惠农	0.89**	10.44	11.96
陶乐	0.89**	16.54	17.10
银川	0.87**	11.25	12.66
吴忠	0.90**	4.99	6.41
盐池	0.85**	15.27	16.40
同心	0.79**	6.37	8.05
中卫	0.69**	27.74	28.98
中宁	0.82**	21.10	21.92
海原	0.77**	9.88	10.87
固原	0.75**	13.31	14.51
西吉	0.82**	5.85	6.75
六盘山	0.72**	5.64	7.54

Fig.5 The box plot of consistency rate of total cloud amount and FY-4A cloud cover fraction at different time in 12 meteorological stations in Ningxia in 2019 （Red solid line represents the median， the green triangle represents the mean， black hollow circles represent outliers）

Fig.6 Distribution of FY-4A mean cloud cover fraction under different actual cloud amounts （The red dotted line represents the ideal trend line， where the satellite cloud cover product is equal to the mean of the total cloud cover observed on the ground）

Tab.4

晴天	少云	多云	阴天
80.5	58.5	22.4	69.3

Fig.7 Comparison of errors between manually observed cloud amount and FY-4A cloud cover fraction for monthly mean values before and after correction in Ningxia and five representative stations

Fig.8 Distribution of annual average FY-4A cloud cover fraction after correctionin in 2019 in Ningxia

Fig.9 Distribution of monthly FY-4A cloud cover fraction after correction in 2019 in Ningxia

References 20

[1]	曹芸, 何永健, 邱新法, 等, 2012. 基于地面观测资料的MODIS云量产品订正[J]. 遥感学报, 16(2):325-342.
[2]	曹越前, 张武, 药静宇, 等, 2015. 半干旱区云量变化特征及基与太阳辐射关系的研究[J]. 干旱气象, 33(4):684-693. DOI
[3]	常倬林, 崔洋, 张武, 等, 2015. 基于CERES的宁夏空中云水资源特征及其增雨潜力研究[J]. 干旱区地理, 38(6):1 112-1 120.
[4]	韩永清, 丛春华, 2015. FY-2E总云量产品在华北和黄淮区域的分析检验[J]. 气象与环境学报, 31(5):153-158.
[5]	范思睿, 王维佳, 林丹, 2020. 基于ISCCP云资料的中国地区不同类型云的时空分布[J]. 干旱气象, 38(2):213-225.
[6]	李峰, 2023. 六盘山夏季东西两侧降水资源及其动力特征研究[D]. 南京: 南京信息工程大学.
[7]	李帅, 陈勇航, 侯小刚, 等, 2021. FY-2F云量产品在新疆区域的评估及检验[J]. 干旱区研究, 38(4):1 031-1 039.
[8]	李娅, 郭建侠, 曹云昌, 等, 2018. FY-2G云量产品与地面观测云量对比分析[J]. 高原气象, 37(2):514-523. DOI
[9]	刘炼烨, 兰明才, 朱歆炜, 等, 2017. FY-2G卫星两种云订正产品在湖南的对比分析[J]. 暴雨灾害, 36(2):164-170.
[10]	王丽, 杜明哲, 李飞, 等, 2018. FY-2G总云量产品在河南省的订正检验[J]. 河南科学, 36 (3):385-389.
[11]	王帅辉, 韩志刚, 姚志刚, 2010. 基于CloudSat和ISCCP资料的中国及周边地区云量分布的对比分析[J]. 大气科学, 34(4):767-779.
[12]	魏栋, 沙宏娥, 秦豪君, 等, 2023. 西北中东部ECMWF-HR云量预报产品与FY-2G云量产品的对比[J]. 干旱气象, 41(3):483-490. DOI
[13]	韦翔鸿, 邱新法, 王娟, 2016. 基于地面观测数据的FY-2D云量订正研究[J]. 科技通报, 32(4):27-31.
[14]	吴薇, 黄晓龙, 郭旭, 等, 2022. 总云量实况分析产品在四川省的检验分析[J]. 气象与环境学报, 38(6):65-72.
[15]	吴伟, 王式功, 2011. 中国北方云量变化趋势及其与区域气候的关系[J]. 高原气象, 30(3): 651-658.
[16]	张德杰, 师春香, 张涛, 等, 2022. 多种资料的总云量产品在中国区域的对比分析[J]. 高原气象, 41(3):803-813. DOI
[17]	张小军, 马学谦, 田建兵, 2017. 1961—2015年青海省总云量时空变化特征及影响因子[J]. 干旱气象, 35(4):626-633.
[18]	SEPHENS G L, KUMMEROW C D, 2007. The remote sensing of clouds and precipitation from space: A review[J]. Journal of the Atmospheric Sciences, 64(11): 3 742-3 765. DOI URL
[19]	WEBB M J, SENIOR C A, SEXTON D M H, et al, 2006. On the contribution of local feedback mechanisms to the range of climate sensitivity in two GCM ensembles[J]. Climate Dynamics, 27(1): 17-38. DOI URL
[20]	YU H, KAUFMAN Y J, CHIN M, et al, 2006. A review of measurement-based assessments of the aerosol direct radiative effect and forcing[J]. Atmospheric Chemistry and Physics, 6(3): 613-666.