COMPARISON AND EVALUATION OF HRCLDAS-V1.0 AND ERA5 SEA-SURFACE WIND FIELDS
-
摘要: 基于2020年中国近海31个浮标的逐小时数据，使用统计分析方法对中国气象局高分辨率陆面数据同化系统(HRCLDAS-V1.0)和欧洲中期天气预报中心第5代全球大气再分析数据(ERA5)海面风场进行了系统的检验，检验结果表明：两者在我国近海均具有较高的可信度，风速平均绝对误差(MAE)分别为1.16 m/s和1.09 m/s，风向MAE分别为23 °和22 °。随着风力增大两者的风速准确度均有所降低，当风力等级≥10级时，前者准确度优于后者；对于风向而言，随着风力增大，两者准确度均升高。此外，选取2020年典型的两次冷空气过程和2008号台风“巴威”过程，检验两者在不同天气过程影响下的准确度，两类融合产品均能较好地再现冷空气过程引起的风向变化，而对不同强度的冷空气过程下的风速反映存在差异；对于台风引起的大风，在风速较低时两者风速均具有不错的表现，但HRCLDAS-V1.0对峰值强度的表现优于ERA5。Abstract: Based on hourly data from 31 buoys in the offshore areas of China in 2020, the sea-surface wind products available from the High Resolution China Meteorological Administration Land Surface Data Assimilation System (HRCLDAS-V1.0) and the Fifth Generation Global Atmospheric Reanalysis (ERA5) dataset were evaluated. The results showed that both HRCLDAS-V1.0 and ERA5 sea-surface wind had high credibility in the offshore areas of China. Their wind speed mean absolute error (MAE) were 1.16 m/ s and 1.09 m/s respectively and wind direction MAE were 23 ° and 22 ° respectively. As wind intensity increased, their wind speed bias all became larger. When wind intensity level was greater than or equal to 10, the HRCLDAS-V1.0 wind speed was better than the ERA5 wind speed. In terms of wind direction, the accuracy of both improved with the increase of wind intensity. In addition, two typical cold air events in 2020 and Typhoon Bavi (2008) were used to examine the accuracy of HRCLDAS-V1.0 and ERA5 seasurface wind. Both HRCLDAS-V1.0 and ERA5 could reproduce the wind direction change caused by cold air events. However, there were some differences in wind speed during cold air events of different intensities. The HRCLDAS-V1.0 and ERA5 wind speed all had good performance when the wind speed was low during Typhoon Bavi. However, the performance of HRCLDAS-V1.0 was better than that of ERA5 for wind at peak intensity.
-
Key words:
- sea-surface wind field /
- data fusion /
- HRCLDAS-V1.0 /
- wind field evaluation /
- buoy /
- ERA5
-
图 7 不同海域各级风速MAE(a)、风向MAE(b)和样本点个数(c)
不同颜色代表不同海域，圆点代表HRCLDAS-V1.0，三角代表ERA5，HRCLDAS-V1.0和ERA5在各海域的各级风样本点数量相同均用圆点表示。
图 9 冷空气过程下MF14005和MF05003浮标及HRCLDAS-V1.0和ERA5风速和风向的时间序列
左侧为风速，右侧为风向，黑色线代表浮标，红色线代表HRCLDAS-V1.0，蓝色线代表ERA5。
图 11 台风过程下MF07001和QF110浮标及HRCLDAS-V1.0和ERA5风速和风向的时间序列
黑色线代表浮标，红色线代表HRCLDAS-V1.0，蓝色线代表ERA5。
表 1 风力等级划分表
风力分级 风速/(m/s) 风力分级 风速/(m/s) 0 0.0~0.2 9 20.8~24.4 1 0.3~1.5 10 24.5~28.4 2 1.6~3.3 11 28.5~32.6 3 3.4~5.4 12 32.7~36.9 4 5.5~7.9 13 37.0~41.4 5 8.0~10.7 14 41.5~46.1 6 10.8~13.8 15 46.2~50.9 7 13.9~17.1 16 51.0~56.0 8 17.2~20.7 17 ≥56.1 
下载: 导出CSV
表 2 HRCLDAS-V1.0和ERA5风速检验结果
风速检验 样本数 ME/(m/s) MAE/(m/s) HRCLDAS-V1.0风速 127 620 -0.22 1.16 ERA5风速 127 620 -0.31 1.09
下载: 导出CSV
表 3 HRCLDAS-V1.0和ERA5风向检验结果
风向检验 样本数 ME/° MAE/° HRCLDAS-V1.0风向 127 620 -4 23 ERA5风向 127 620 -5 22
下载: 导出CSV
表 4 两次冷空气过程HRCLDAS-V1.0和ERA5风速和风向统计检验结果
风速和风向 MF14005 MF05003 MAE 最大偏差 MAE 最大偏差 HRCLDAS-V1.0风速 0.84 m/s 2.40 m/s 2.04 m/s 4.39 m/s ERA5风速 0.87 m/s 2.15 m/s 2.21 m/s 4.23 m/s HRCLDAS-V1.0风向 7 ° 17 ° 9 ° 20 ° ERA5风向 4 ° 11 ° 9 ° 21 °
下载: 导出CSV
表 5 台风过程HRCLDAS-V1.0和ERA5风速和风向的统计检验结果
风速和风向 MF07001 QF110 MAE COR 峰值误差 MAE COR 峰值误差 HRCLDAS-V1.0风速 2.12 m/s 0.92 -4.3 m/s 2.53 m/s 0.85 -2.6 m/s ERA5风速 3.46 m/s 0.72 -7.9 m/s 1.82 m/s 0.92 -3.9 m/s HRCLDAS-V1.0风向 8 ° - - 17 ° - - ERA5风向 11 ° - - 17 ° - -
下载: 导出CSV
-
[1] 张东翔. 多源卫星海面风场产品检验及融合研究[D]. 长沙: 国防科技大学, 2018. [2] 刘付前, 骆永军, 王超, 等. 应用卫星散射计资料研究中国海海面风场时空特征[J]. 环境保护与循环经济, 2009(10): 37-38. https://www.cnki.com.cn/Article/CJFDTOTAL-LNCX200910016.htm [3] 邓婉月, 项杰, 杜华栋. 南海海面风场融合研究[C]. 第32届中国气象学会年会S18气象卫星遥感新资料——新方法——新应用. 2015. [4] KAKO S, ISOBE A, KUBOTA M. High-resolution ASCAT wind vector data set gridded by applying an optimum interpolation method to the global ocean[J]. J Geophys Res Atmos, 2011, 116(D23107). doi: 10.1029/2010jd015484 [5] NOELIA OSES FERNÁNDEZ, AZPIROZ I, MARCHI S, et al. Analysis of Copernicus' ERA5 climate reanalysis data as a replacement for weather station temperature measurements in machine learning models for olive phenology phase prediction[J]. Sensors, 2020, 20(21): 6 381. doi: 10.3390/s20216381 [6] KOBAYASHI C, ENDO H, OTA Y, et al. Preliminary results of the JRA-55C, an atmospheric reanalysis assimilating conventional observations only[J]. Scientific Online Letters on the Atmosphere: SOLA, 2014, 10(1): 2 014-2 016. [7] KANAMITSU M, KISTLER R E, REYNOLDS R W. NCEP/NCAR reanalysis and the use of satellite data[J]. Advances in Space Research, 1997, 19(3): 481-489. doi: 10.1016/S0273-1177(97)00059-8 [8] YAN Q, ZHANG J, MENG J, et al. Use of an optimum interpolation method to construct a high-resolution global ocean surface vector wind dataset from active scatterometers and passive radiometers[J]. International Journal of Remote Sensing, 2017, 38(19-20): 5 569-5 591. [9] 项杰, 王慧鹏, 王春明, 等. 南海海面风场变分融合的初步研究[J]. 热带气象学报, 2015, 31(2): 153-153. https://www.cnki.com.cn/Article/CJFDTOTAL-RDQX201502002.htm [10] 姜祝辉, 游小宝, 肖义国. 高度计风速与辐射计风速的变分融合研究[J]. 物理学报, 2013, 62(12): 129202. https://www.cnki.com.cn/Article/CJFDTOTAL-WLXB201312076.htm [11] 姜祝辉, 黄思训, 杜华栋, 等. 利用变分结合正则化方法对高度计风速资料调整海面风场的研究[J]. 物理学报, 2010, 59(12): 8 968-8 977. https://www.cnki.com.cn/Article/CJFDTOTAL-WLXB201012094.htm [12] 师春香, 姜立鹏, 张涛. CMA陆面数据同化业务系统及产品介绍[C]. 第30届中国气象学会年会论文集, 南京, 2013: 1-4. [13] 韩帅, 师春香, 姜志伟, 等. CMA高分辨率陆面数据同化系统(HRCLDAS-V1.0)研发及进展[J]. 气象科技进展, 2018, 8(1): 102-108、116. https://www.cnki.com.cn/Article/CJFDTOTAL-QXKZ201801028.htm [14] JIANG Y, HAN S, SHI C, et al. Evaluation of HRCLDAS and ERA5 datasets for near-surface wind over Hainan Island and South China Sea [J]. Atmosphere, 2021, 12(6): 766. [15] XIE Y, KOCH S, MCGINLEY J, et al. A space-time multiscale analysis system: A sequential variational analysis approach[J]. Mon Wea Rev, 2011, 139(4): 1 224-1 240. [16] YANG K, HUANG G W, TAMAI N. A hybrid model for estimating global solar radiation[J]. Solar Energy, 2001, 70(1): 13-22. [17] 张涛, 苗春生, 王新. LAPS与STMAS地面气温融合效果对比试验[J]. 高原气象, 2014, 33(3): 743-752. https://www.cnki.com.cn/Article/CJFDTOTAL-GYQX201403016.htm -
点击查看大图
图(11) / 表(5)
计量
- 文章访问数: 75
- HTML全文浏览量: 12
- PDF下载量: 16
- 被引次数: 0
粤公网安备 4401069904700003号