基于伴随敏感性的风廓线雷达和地基微波辐射计观测对模式预报的影响评估研究

唐兆康; 鲍艳松; 顾英杰; 范水勇; 齐亚杰; 崔伟; 陈强

doi:10.3878/j.issn.1006-9895.2107.20222

基于伴随敏感性的风廓线雷达和地基微波辐射计观测对模式预报的影响评估研究

doi: 10.3878/j.issn.1006-9895.2107.20222

唐兆康^{1, 5,},
鲍艳松^{1, 2, ,},
顾英杰¹,
范水勇³,
齐亚杰³,
崔伟⁴,
陈强⁴

1.
南京信息工程大学气象灾害预报预警与评估协同创新中心, 南京210044
2.
南京信息工程大学中国气象局气溶胶—云—降水重点开放实验室, 南京210044
3.
北京城市气象研究院, 北京100089
4.
上海卫星工程研究所, 上海201109
5.
南京南瑞水利水电科技有限公司，南京 211000

基金项目: 国家重点研发计划项目2017YFC1501704，上海航天科技创新基金资助项目 SAST2019-046，国家自然科学基金项目41975046

详细信息

作者简介:
唐兆康，男，1995年出生，硕士研究生，主要从事数值模式和资料同化研究。E-mail: tzk_nuist@foxmail.com

通讯作者:
鲍艳松，E-mail: ysbao@nuist.edu.cn

中图分类号: P435
计量
- 文章访问数: 405
- HTML全文浏览量: 60
- PDF下载量: 171
- 被引次数: 0
出版历程
- 收稿日期: 2020-11-02
- 录用日期: 2021-09-02
- 网络出版日期: 2021-09-09
- 刊出日期: 2022-07-19

Using the Adjoint-Based Forecast Sensitivity Method to Evaluate the Observations of Wind Profile Radar and Microwave Radiometer Impacts on a Model Forecast

Zhaokang TANG^{1, 5
,},
Yansong BAO^{1, 2
, ,},
Yingjie GU¹,
Shuiyong FAN³,
Yajie QI³,
Wei CUI⁴,
Qiang CHEN⁴

1.
Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044
2.
Key Laboratory for Aerosol-Cloud-Precipitation of China Meteorological Administration, Nanjing University of Information Science & Technology, Nanjing 210044
3.
Beijing Institute of Urban Meteorology, Beijing 100089
4.
Shanghai Institute of Satellite Engineering, Shanghai 201109
5.
Nanjing NARI Water Resources and Hydropower'Technology Company Limited, Nanjing 211000

Funds: National Key Research and Development Program of China (Grant 2017YFC1501704), Shanghai Aerospace Science and Technology Innovation Foundation (Grant SAST2019-046), National Natural Science Foundation of China (Grant 41975046)

摘要

摘要: 同化大量观测资料可以有效地改进模式预报结果，但不同观测对预报的影响有着显著差异，合理评估观测对预报的贡献是数值模式中最具挑战性的诊断之一。本文采用基于伴随的预报对观测的敏感性（Forecast Sensitivity to Observation，简称FSO）方法，构建WRFDA（Weather Research and Forecasting model’s Data Assimilation）框架下的WRFDA-FSO系统。基于2019年9月超大城市项目在北京地区获取的风廓线雷达（Wind Profile Radar，简称WPR）和地基微波辐射计（Microwave Radiometer，简称MWR）观测数据，利用WRFDA-FSO系统，开展观测对WRF模式12 h预报的影响试验，并分析风温湿观测对预报的贡献。结果表明：（1）同化的观测资料（MWR、WPR、Sound、Synop和Geoamv）均减小了WRF模式12 h预报误差，对预报为正贡献，其中MWR观测对预报的影响最大，WPR风场观测对预报的改进效果优于Sound的风场观测。（2）WPR的U、V观测和MWR的T、Q观测中，V观测和T观测对预报的正贡献值更高，对预报的改进效果更优。（3）WPR和MWR多数高度层的观测均减小了预报误差，对预报为正贡献，其中MWR的T观测对预报的正贡献主要位于近地面800 hPa以下。
- 数值模式 /
- 资料同化 /
- 预报对观测的敏感性 /
- 影响试验
Abstract: Many assimilated observations can effectively improve the results of a model forecast. However, there are significant differences in the effects of various observations on the forecast. It is one of the most challenging diagnostics in numerical models to reasonably evaluate the observation contribution to the forecast. In this paper, the weather research and forecasting model’s data assimilation (WRFDA) and forecast sensitivity to observation (FSO) system was constructed in WRFDA by the method of adjoint-based FSO. Based on wind profile radar (WPR) and ground-based microwave radiometer (MWR) data obtained by the mega city project in Beijing in September 2019, the experiments on the impact of observations on the 12 h forecast of the WRF model are carried out using the WRFDA-FSO system. The contribution of wind, temperature, and humidity observations to the forecast is analyzed. The results show the following: (1) In general, the assimilated observations (MWR, WPR, Sound, Synop, and Geoamv) reduce the 12 h forecast error of the WRF model and make a positive contribution to the forecast. Among these, MWR observations have the greatest impact on the forecast, and the improvement of WPR observations on the forecast is better than that of wind field observations of sound. (2) Among the U and V observations of WPR and temperature and specific humidity observations of MWR, the positive contribution value of V observations and temperature observations to the forecast is higher, and the effect of improving the forecast is better. (3) The WPR and MWR observations, at most levels, reduce the forecast error and are a positive contribution to the forecast. The positive contribution of temperature observations is mainly below 800 hPa near the ground.
- Numerical model /
- Data assimilation /
- Forecast sensitivity to observation /
- Impact experiment

HTML全文

图 1 2019年9月（a）00时（协调世界时，下同）和（b）12时观测对预报的影响（$ \Delta e $，单位：10³ J kg⁻¹）及其近似估计（$ \delta e $，单位：10³ J kg⁻¹）的时间序列

Figure 1. Time series of the impact of observation on the forecast ($ \Delta e $, units: 10³ J kg⁻¹) and its approximate estimation ($ \delta e $, units: 10³ J kg⁻¹) at (a) 0000 UTC and (b) 1200 UTC in September 2019

下载: 全尺寸图片幻灯片

图 2 2019年9月（a）00时和（b）12时不同观测的站点观测对预报的贡献及其位置分布（红色表示该站点观测对预报负贡献，蓝色表示该站点观测对预报正贡献；图例中“/”前后的数字分别表示该观测的正贡献站点数和总站点数）

Figure 2. Contribution of the station observation with various observations to the forecast and its location distribution at (a) 0000 UTC and (b) 1200 UTC in September 2019 (red indicates a negative contribution of the station observation to the forecast, and the blue indicates a positive contribution of the station observation to the forecast; the numbers before and after “/” in the legend represent the number of positive contribution stations and total stations of the observation respectively)

下载: 全尺寸图片幻灯片

图 3 2019年9月00时和12时的（a）不同观测对预报误差的贡献（单位：10⁴ J kg⁻¹）和（b）不同观测的有益观测百分比

Figure 3. (a) Contribution (units: 10⁴ J kg⁻¹) of various observations to the forecast error and (b) percentage of beneficial observations of various observations at 0000 UTC and 1200 UTC in September 2019

下载: 全尺寸图片幻灯片

图 4 2019年9月00时和12时北京地区7站点（HD、YQ、HR、MY、PG、DX和XYL）的（a）WPR观测和（b）MWR观测对预报误差的平均贡献（单位：10² J kg⁻¹）。HD：海淀，YQ：延庆，HR：怀柔，MY：密云，PG：平谷，DX：大兴，XYL：霞云岭

Figure 4. Average contribution (units: 10² J kg⁻¹) of (a) WPR observations and (b) MWR observations to the forecast error at seven stations, which are Haidian (HD), Yanqing (YQ), Huairou (HR), Miyun (MY), Pinggu (PG), Daxing (DX), and Xiayunling (XYL) in Beijing at 0000 UTC and 1200 UTC in September 2019

下载: 全尺寸图片幻灯片

图 5 2019年9月00时观测误差σ≤3且新息增量|δy|＞4的WPR（a）U观测和（b）V观测对应的对预报误差的贡献（单位：10² J kg⁻¹）与新息增量（单位：m s⁻¹）的散点图（黑色表示该观测对预报负贡献，灰色表示该观测对预报正贡献）

Figure 5. Scatter plot of forecast error contribution (units: 10² J kg⁻¹) and innovation increment (units: m s⁻¹) corresponding to (a) U observations and (b) V observations of WPR with observation error less than or equal to 3 and innovation increment greater than 4 (σ≤ 3 and |δy|＞4) at 0000 UTC in September 2019 (black indicates a negative contribution of the observation to the forecast, and gray indicates a positive contribution of the observation to the forecast)

下载: 全尺寸图片幻灯片

图 6 2019年9月00时MWR的T观测对应的（a）对预报误差的贡献（单位：10³ J kg⁻¹）、（b）分析增量（单位：K）、（c）新息增量（单位：K）以及（d）观测误差（单位：K）随高度的分布

Figure 6. The distribution of (a) a forecast error contribution (units: 10³ J kg⁻¹), (b) an analysis increment (units: K), (c) an innovation increment (units: K), and (d) an observation error (units: K) with a height corresponding to temperature observations of MWR at 0000 UTC in September 2019

下载: 全尺寸图片幻灯片

表 1 剔除近似结果不准确的时次前后各观测资料对预报误差的贡献统计（单位：10⁴ J kg⁻¹）

Table 1. Statistics of contributing various observations to the forecast error before and after eliminating inaccurate approximate results (units: 10⁴J kg⁻¹)

观测资料	00时对预报误差的贡献		12时对预报误差的贡献
观测资料	剔除前	剔除后	剔除前	剔除后
MWR	−117.06	−128.38	11.89	−62.56
WPR	5.88	−12.75	−125.16	−26.30
Sound	−26.07	−45.37	−82.87	−25.41
Synop	−14.90	−23.14	−26.71	−42.30
Geoamv	−0.77	−0.58	0.27	−0.30

下载: 导出CSV

表 2 2019年9月00时和12时廓线观测对预报误差的平均贡献（单位：J kg⁻¹）统计

Table 2. Statistics of the average contribution of profile observations to the forecast error (units: J kg⁻¹) at 0000 UTC and 1200 UTC in September 2019

观测	时次	观测对预报误差平均贡献/J kg⁻¹
观测	时次	U	V	T	Q
WPR	00时	15.09	−80.49	−	−
WPR	12时	−94.35	−127.96	−	−
MWR	00时	−	−	−468.49	−72.34
MWR	12时	−	−	−286.58	−17.00
Sound	00时	−8.73	−38.06	−75.78	−88.89
Sound	12时	−30.75	−44.42	−30.36	−8.55

下载: 导出CSV

表 3 2019年9月00时和12时不同高度层WPR观测对预报误差的平均贡献（单位：J kg⁻¹）统计

Table 3. Statistics of the average contribution of WPR observations at various altitudes to the forecast error (units: J kg⁻¹) at 0000 UTC and 1200 UTC in September 2019

高度层	00时对预报误差的平均贡献/J kg⁻¹		12时对预报误差的平均贡献/J kg⁻¹
高度层	U	V	U	V
<500 m	−129.32	−289.43	105.27	−497.88
500～1000 m	23.99	−60.47	−248.26	−300.04
1000～1500 m	216.62	−37.83	−80.44	−188.43
1500～2000 m	180.83	−135.84	−153.71	−144.17
2000～3000 m	−49.26	−103.49	−99.29	−102.72
3000～4000 m	−156.17	−38.77	−64.03	−20.35
≥4000 m	−61.11	−21.52	−33.31	12.6

下载: 导出CSV

表 4 2019年9月00时和12时不同高度层MWR观测对预报误差的平均贡献（单位：J kg⁻¹）统计

Table 4. Statistics of the average contribution of MWR observations at various altitudes to the forecast error (units: J kg⁻¹) at 0000 UTC and 1200 UTC in September 2019

高度层	00时对预报误差的平均贡献/J kg⁻¹		12时对预报误差的平均贡献/J kg⁻¹
高度层	T	Q	T	Q
>900 hPa	−1949.85	−77.67	−919.90	−51.68
800～900 hPa	−580.04	−132.18	−419.22	23.19
700～800 hPa	356.33	−1.47	−163.18	−90.48
600～700 hPa	144.00	−41.48	28.37	90.92
500～600 hPa	−5.21	55.16	10.35	−13.02
400～500 hPa	−7.08	−97.08	2.40	−21.19
≤400 hPa	6.33	−130.80	4.46	−39.21

下载: 导出CSV

表 5 2019年9月00时WPR观测对预报误差的贡献（单位：J kg⁻¹）分类统计

Table 5. Classified statistics of the contribution of WPR observations to the forecast error (units: J kg⁻¹) at 0000 UTC in September 2019

	U观测
	负贡献(×10³J/kg) / 观测数		正贡献(×10³J/kg) / 观测数
	\|δy\|＞4	\|δy\|≤4	\|δy\|＞4	\|δy\|≤4
σ≤3	148.75 / 37	157.99 / 730	−76.35 / 74	−195.55 / 969
σ＞3	0.12 / 4	0.92 / 36	−3.16 / 17	−2.43 / 80

	V观测
	负贡献(×10³J/kg) / 观测数		正贡献(×10³J/kg) / 观测数
	\|δy\|＞4	\|δy\|≤4	\|δy\|＞4	\|δy\|≤4
σ≤3	61.40 / 28	100.50 / 697	−35.13 / 46	−281.52 / 1039
σ＞3	0.14 / 3	0.87 / 49	−1.89 / 10	−1.98 / 75

下载: 导出CSV

表 6 2019年9月00时 MWR观测对预报误差的贡献（单位：J kg⁻¹）分类统计

Table 6. Classified statistics of the contribution of MWR observations to the forecast error (units: J kg⁻¹) at 0000 UTC in September 2019

	T观测
	负贡献(×10³J/kg) / 观测数		正贡献(×10³J/kg) / 观测数
	\|δy\|＞2	\|δy\|≤2	\|δy\|＞2	\|δy\|≤2
σ≤1	50.81 / 39	309.20 / 197	−308.19/ 41	−932.91 / 285
σ＞1	101.56 / 418	71.54 / 558	−287.15/ 305	−121.72 / 541

	Q观测
	负贡献(×10³J/kg) / 观测数		正贡献(×10³J/kg) / 观测数
	\|δy\|＞2	\|δy\|≤2	\|δy\|＞2	\|δy\|≤2
σ≤1	19.76 / 10	89.65 / 474	−4.79 / 2	−235.19 / 788
σ＞1	73.84 / 45	55.41 / 384	−49.06 / 85	−116.53 / 516

下载: 导出CSV

参考文献(36)

[1]	Auligné T. 2010. Forecast sensitivity to observations & observation impact [EB/OL]. http://www2.mmm.ucar.edu/wrf/users/wrfda/Tutorials/2010_Aug/docs/WRFDA_sensitivity.pdf [2020-09-25]
[2]	Baker N L, Daley R. 2000. Observation and background adjoint sensitivity in the adaptive observation-targeting problem [J]. Quart. J. Roy. Meteor. Soc., 126(565): 1431−1454. doi: 10.1002/qj.49712656511
[3]	Chen F, Dudhia J. 2001. Coupling an advanced land surface-hydrology model with the Penn state–NCAR MM5 modeling system. Part I: Model implementation and sensitivity [J]. Mon. Wea. Rev., 129(4): 569−585. doi: 10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2
[4]	成巍, 王斌, 徐幼平. 2012. 基于局地和非局地观测算子的GPS掩星资料后向映射四维变分同化研究 [J]. 中国科学:数学, 42(5): 377−387. doi: 10.1360/012012-17 Cheng Wei, Wang Bin, Xu Youping. 2012. Assimilation of GPS radio occultation data with the local and non-local operators using Backward-4DVar approach [J]. Scientia Sinica:Mathematica (in Chinese), 42(5): 377−387. doi: 10.1360/012012-17
[5]	丁锦锋. 2015. 中国AMDAR观测评估与应用研究 [D]. 南京大学博士学位论文. Ding Jinfeng. 2015. Evaluation of Chinese AMDAR weather reports and applied research [D]. Ph. D. dissertation (in Chinese), Nanjing University.
[6]	Errico R M. 2007. Interpretations of an adjoint-derived observational impact measure [J]. Tellus A:Dynamic Meteorology and Oceanography, 59(2): 273−276. doi: 10.1111/j.1600-0870.2006.00217.x
[7]	Gelaro R, Zhu Y Q. 2009. Examination of observation impacts derived from observing system experiments (OSEs) and adjoint models [J]. Tellus A:Dynamic Meteorology and Oceanography, 61(2): 179−193. doi: 10.1111/j.1600-0870.2008.00388.x
[8]	Gelaro R, Zhu Y Q, Errico R M. 2007. Examination of various-order adjoint-based approximations of observation impact [J]. Meteor. Z., 16(6): 685−692. doi: 10.1127/0941-2948/2007/0248
[9]	Gelaro R, Langland R H, Pellerin S, et al. 2010. The THORPEX observation impact intercomparison experiment [J]. Mon. Wea. Rev., 138(11): 4009−4025. doi: 10.1175/2010MWR3393.1
[10]	韩峰, 储可宽, 刘浩铄, 等. 2018. 一次过度预报的温带气旋的观测资料影响性研究 [J]. 气象科学, 38(5): 637−647. doi: 10.3969/2017jms.0091 Han Feng, Chu Kekuan, Liu Haoshuo, et al. 2018. Observation impact on an over-forecasted extratropical cyclone [J]. Journal of the Meteorological Sciences (in Chinese), 38(5): 637−647. doi: 10.3969/2017jms.0091
[11]	Iacono M J, Delamere J S, Mlawer E J, et al. 2008. Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models [J]. J. Geophys. Res. Atmos., 113(D13): D13103. doi: 10.1029/2008JD009944
[12]	Jung B J, Kim H M, Kim Y H, et al. 2010. Observation system experiments for typhoon Jangmi (200815) observed during T-PARC [J]. Asia-Pac. J. Atmos. Sci., 46(3): 305−316. doi: 10.1007/s13143-010-1007-y
[13]	Jung B J, Kim H, Zhang F Q, et al. 2012. Effect of targeted dropsonde observations and best track data on the track forecasts of typhoon Sinlaku (2008) using an ensemble Kalman filter [J]. Tellus A:Dynamic Meteorology and Oceanography, 64(1): 14984. doi: 10.3402/tellusa.v64i0.14984
[14]	Jung B J, Kim H M, Auligné T, et al. 2013. Adjoint-derived observation impact using WRF in the western North Pacific [J]. Mon. Wea. Rev., 141(11): 4080−4097. doi: 10.1175/MWR-D-12-00197.1
[15]	Kain J S. 2004. The Kain-Fritsch convective parameterization: An update [J]. J. Appl. Meteor. Climatol., 43(1): 170−181. doi: 10.1175/1520-0450(2004)043<0170:tkcpau>2.0.co;2
[16]	Kim S M, Kim H M. 2014. Sampling error of observation impact statistics [J]. Tellus A:Dynamic Meteorology and Oceanography, 66(1): 25435. doi: 10.3402/tellusa.v66.25435
[17]	Kim M, Kim H M, Kim J, et al. 2017. Effect of enhanced satellite-derived atmospheric motion vectors on numerical weather prediction in East Asia using an adjoint-based observation impact method [J]. Wea. Forecasting, 32(2): 579−594. doi: 10.1175/WAF-D-16-0061.1
[18]	Kunii M, Miyoshi T, Kalnay E. 2012. Estimating the impact of real observations in regional numerical weather prediction using an ensemble Kalman filter [J]. Mon. Wea. Rev., 140(6): 1975−1987. doi: 10.1175/MWR-D-11-00205.1
[19]	Langland R H, Baker N L. 2004. Estimation of observation impact using the NRL atmospheric variational data assimilation adjoint system [J]. Tellus A:Dynamic Meteorology and Oceanography, 56(3): 189−201. doi: 10.3402/tellusa.v56i3.14413
[20]	Lanzante J R. 1996. Resistant, robust and non-parametric techniques for the analysis of climate data: Theory and examples, including applications to historical radiosonde station data [J]. Int. J. Climatol., 16(11): 1197−1226. doi: 10.1002/(SICI)1097-0088(199611)16:11<1197::AID-JOC89>3.0.CO;2-L
[21]	李佳, 陈葆德, 张旭, 等. 2017. 2016年6月23日江苏阜宁龙卷的高分辨快速更新同化预报与分析 [J]. 大气科学, 41(6): 1221−1233. doi: 10.3878/j.issn.1006-9895.1707.17144 Li Jia, Chen Baode, Zhang Xu, et al. 2017. High-resolution rapid refresh analysis and prediction of the tornado occurring in Funing on 23 June 2016 [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 41(6): 1221−1233. doi: 10.3878/j.issn.1006-9895.1707.17144
[22]	Liu J J, Kalnay E. 2008. Estimating observation impact without adjoint model in an ensemble Kalman filter [J]. Quart. J. Roy. Meteor. Soc., 134(634): 1327−1335. doi: 10.1002/qj.280
[23]	刘志明, 张其松, 晏明. 2002. 静止气象卫星云导风的结果分析 [J]. 吉林气象(3): 15−18.
[24]	Lorenc A C, Marriott R T. 2014. Forecast sensitivity to observations in the Met Office Global numerical weather prediction system [J]. Quart. J. Roy. Meteor. Soc., 140(678): 209−224. doi: 10.1002/qj.2122
[25]	陆续, 马旭林, 王旭光. 2015. 三维变分同化机载雷达资料对飓风预报的影响研究——2012年Isaac试验 [J]. 大气科学, 39(6): 1111−1122. doi: 10.3878/j.issn.1006-9895.1503.14262 Lu Xu, Ma Xulin, Wang Xuguang. 2015. A study of the impact of airborne radar data assimilated by 3DVar on the prediction of hurricane–Isaac 2012 [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 39(6): 1111−1122. doi: 10.3878/j.issn.1006-9895.1503.14262
[26]	Pleim J E. 2007. A combined local and nonlocal closure model for the atmospheric boundary layer. Part I: Model description and testing [J]. J. Appl. Meteor. Climatol., 46(9): 1383−1395. doi: 10.1175/JAM2539.1
[27]	Rabier F, Klinker E, Courtier P, et al. 1996. Sensitivity of forecast errors to initial conditions [J]. Quart. J. Roy. Meteor. Soc., 122(529): 121−150. doi: 10.1002/qj.49712252906
[28]	Thompson G, Field P R, Rasmussen R M, et al. 2008. Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: Implementation of a new snow parameterization [J]. Mon. Wea. Rev., 136(12): 5095−5115. doi: 10.1175/2008mwr2387.1
[29]	Trémolet Y. 2007. First-order and higher-order approximations of observation impact [J]. Meteor. Z., 16(6): 693−694. doi: 10.1127/0941-2948/2007/0258
[30]	王曼, 段旭, 李华宏, 等. 2015. 青藏高原东侧常规观测资料对WRF模式预报误差的贡献分析 [J]. 大气科学学报, 38(3): 379−387. doi: 10.13878/j.cnki.dqkxxb.20130520003 Wang Man, Duan Xu, Li Huahong, et al. 2015. Evaluation of conventional observations contribution on WRF model forecast error in the eastern of Tibetan Plateau [J]. Transactions of Atmospheric Sciences (in Chinese), 38(3): 379−387. doi: 10.13878/j.cnki.dqkxxb.20130520003
[31]	王业桂, 张斌, 蔡其发, 等. 2018. 不同卫星微波遥感资料同化对台风路径模拟的影响 [J]. 大气科学, 42(2): 398−410. doi: 10.3878/j.issn.1006-9895.1709.17150 Wang Yegui, Zhang Bin, Cai Qifa, et al. 2018. Effects of assimilating microwave remote sensing data of different satellite on the simulation of typhoon track [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 42(2): 398−410. doi: 10.3878/j.issn.1006-9895.1709.17150
[32]	王丹, 阮征, 王改利, 等. 2019. 风廓线雷达资料在GRAPES-Meso模式中的同化应用研究 [J]. 大气科学, 43(3): 634−654. doi: 10.3878/j.issn.1006-9895.1810.18125 Wang Dan, Ruan Zheng, Wang Gaili, et al. 2019. A study on assimilation of wind profiling radar data in GRAPES-Meso model [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 43(3): 634−654. doi: 10.3878/j.issn.1006-9895.1810.18125
[33]	Wu T C, Velden C S, Majumdar S J, et al. 2015. Understanding the influence of assimilating subsets of enhanced atmospheric motion vectors on numerical analyses and forecasts of tropical cyclone track and intensity with an ensemble Kalman filter [J]. Mon. Wea. Rev., 143(7): 2506−2531. doi: 10.1175/MWR-D-14-00220.1
[34]	Yang E G, Kim H M, Kim J, et al. 2014. Effect of observation network design on meteorological forecasts of Asian dust events [J]. Mon. Wea. Rev., 142(12): 4679−4695. doi: 10.1175/MWR-D-14-00080.1
[35]	张斌, 张立凤, 熊春晖. 2014. ATOVS资料同化方案对暴雨模拟效果的影响 [J]. 大气科学, 38(5): 1017−1027. doi: 10.3878/j.issn.1006-9895.1401.13215 Zhang Bin, Zhang Lifeng, Xiong Chunhui. 2014. Effects of ATOVS data assimilation schemes on the simulation of heavy rain [J]. Chinese Journal of Atmospheric Sciences (in Chinese), 38(5): 1017−1027. doi: 10.3878/j.issn.1006-9895.1401.13215
[36]	Zhu Y Q, Gelaro R. 2008. Observation sensitivity calculations using the adjoint of the Gridpoint Statistical Interpolation (GSI) analysis system [J]. Mon. Wea. Rev., 136(1): 335−351. doi: 10.1175/MWR3525.1