基于2017年5月7日广州特大暴雨分析影响半径对集合卡尔曼滤波方法同化效果的影响
|
摘要
集合卡尔曼滤波(EnKF)目前在资料同化的科研和业务中已得到广泛应用,可为集合预报提供较好的初始场,其影响半径的选取对同化结果影响显著。2017年5月7日在珠三角(珠江三角洲)一带出现极强降水,尤以广州的花都、黄埔、增城区为盛,甚至出现了极为罕见的特大暴雨。以本次极强降水过程为例,分析影响半径对EnKF同化效果的影响。结果发现利用EnKF方法同化观测站的10 m风和2 m温度资料后,可以较好地模拟出此次强降水过程,但仍存在着位置偏南,强度偏大,局地虚报和过报的现象。当水平影响半径取值过大时,大量虚假信息引入,产生过犹不及的效果,使得强降水过程南移较快,最终导致降水落区显著偏南偏东。且水平影响半径对模拟效果极为重要,因此取值要适当。
The Ensemble Kalman Filtering(EnKF) has been widely used in data assimilation. It can provide a better initial field for ensemble prediction, and the influence radius has a significant impact on the assimilation results. On May 7 th, 2017, there was a very heavy precipitation in the Pearl River Delta,especially in Huadu, Huangpu, and Zengcheng, where there was even a very rare torrential rain. Taking the process of this extremely heavy rainfall as example, this paper analyzed the effect of the influence radius on the assimilation with EnKF. The results show that the heavy rainfall process can be simulated well by assimilating observation with the EnKF. However, it still existed that the precipitation was located more to the south. Moreover, the rainfall intensity was too large, with false or over-predicted amounts in some areas.When the value of influence radius was too large, a large number of false information may cause the rapid movement of heavy rainfall to the south. Furthermore, it may result in more southward and eastward location of the precipitation area. The horizontal influence radius was very important to the simulation effect, so the value should be appropriate.
The Ensemble Kalman Filtering(EnKF) has been widely used in data assimilation. It can provide a better initial field for ensemble prediction, and the influence radius has a significant impact on the assimilation results. On May 7 th, 2017, there was a very heavy precipitation in the Pearl River Delta,especially in Huadu, Huangpu, and Zengcheng, where there was even a very rare torrential rain. Taking the process of this extremely heavy rainfall as example, this paper analyzed the effect of the influence radius on the assimilation with EnKF. The results show that the heavy rainfall process can be simulated well by assimilating observation with the EnKF. However, it still existed that the precipitation was located more to the south. Moreover, the rainfall intensity was too large, with false or over-predicted amounts in some areas.When the value of influence radius was too large, a large number of false information may cause the rapid movement of heavy rainfall to the south. Furthermore, it may result in more southward and eastward location of the precipitation area. The horizontal influence radius was very important to the simulation effect, so the value should be appropriate.
引文
[1] HOUTEKAMER P L, LEFAIVRE L, DEROME J, et al. A system simulation approach to ensemble prediction[J]. Mon Wea Rev, 1996,124(6):1 225-1 242.
[2] ANDERSON J L. An ensemble adjustment Kalman Filter for data assimilation[J]. Mon Wea Rev, 2001,129(12):2 884-2 903.
[3] EVENSEN G. Sequential data assimilation with a nonlinear quasi-geostrophic model using Monte Carlo methods to forecast errorstatistics[J]. Geophys Res, 1994,99(5):10 143-10 162.
[4] EVENSEN G, LEEUWEN P J. Assimilation of GEOSAT altimeter data for the Aghulas current using the ensemble Kalman filter with aquasigeostrophic model[J]. Mon Wea Rev, 1996,124(1):85-96.
[5]张爱忠,齐琳琳,纪飞,等.资料同化方法研究进展[J].气象科技,2005,33(5):385-394.
[6]杨毅,弓中强,王金艳,等.时间扩展取样集合卡尔曼滤波同化模拟探空试验研究[J].气象学报,2012, 70(1):101-108.
[7] HOUTEKAMER P L, MITCHELL H L. Data assimilation using an ensemble Kalman filter technique[J]. Mon Wea Rev, 1998, 126(3):796-811.
[8] WHITAKER J S, HAMILL T M. Ensemble data assimilation without perturbed observations[J]. Mon Wea Rev, 2002, 130(12):1 913-1 924.
[9] ZHANG S, LI H, ZHANG W, et al. Estimating the soil moisture profile by assimilating near surface observations with the EnsembleKalman Filter(En KF)[J]. Adv Atmos Sci, 2005, 22(6):936-945.
[10]朱江,汪萍.集合卡尔曼平滑和集合卡尔曼滤波在污染源反演中的应用[J].大气科学, 2006, 30(6):871-882.
[11]刘寅,何光鑫,刘佳伟,等.雷达资料的3DVAR和EnSRF直接同化方法对一次暴雨预报的影响分析[J].热带气象学报,2014,30(5):894-904.
[12]高士博,闵锦忠,黄丹莲. EnKF局地化算法对雷达资料同化的影响研究[J].大气科学学报, 2016, 39(5):633-642.
[13]庄照荣,薛纪善,李兴良. GRAPES集合卡尔曼滤波资料同化系统Ⅱ:区域分析及集合预报[J].气象学报,2011,69(6):860-871.
[14] MITCHELL H L, HOUTEKAMER P L, PELLERIN G. Ensemble size, balance, and model-error representation in an ensemble Kalmanfilter[J]. Mon Wea Rev, 2002, 130(11):2 791-2 808.
[15] WU X, LI W, HAN G, et al. An adaptive compensatory approach of the fixed localization in the EnKF[J]. Mon Wea Rev, 2015, 143(11):4 714-4 735.
[16] THOMPSON T E, WICKER L J, WANG X. Impact from a volumetric radar-sampling operator for radial velocity observations withinEnKF supercell assimilation[J]. J Atmos Oceanic Technol, 2012, 29(10):1 417-1 427.
[17] HAMRUD M, BONAVITA M, ISAKSEN L. EnKF and Hybrid Gain Ensemble Data Assimilation. Part I:EnKF Implementation[J].Mon Wea Rev, 2015, 143(12):4 847-4 864.
[18] HOUTEKAMER P L, MITCHELL H L. A sequential ensemble Kalman filter for atmospheric data assimilation[J]. Mon Wea Rev, 2001,129(1):123-137.
[19] SKAMAROCK W C, KLEMP J B, JIMY D H, et al. A description of the advanced research WRF version 3[R]. NCAR Tech, Note, 2008:113.
[20] BARKER D M, HUANG W, GUO Y R, et al. A three-dimensional variational data assimilation system for MM5:Implementation andinitial results[J]. Mon Wea Rev, 2004, 132(4):897-914.
[21] ZHU L, WAN Q, SHEN X, et al. Prediction and predictability of high-impact western Pacific landfalling Tropical Cyclone Vicente(2012)through convection-permitting ensemble assimilation of Doppler radar velocity[J]. Mon Wea Rev, 2016,144(1):21-43.
[22]朱磊,万齐林,刘靓珂,等.雷达径向风资料EnKF同化应用对Vicente(2012)登陆台风强度变化过程预报的影响试验研究[J].热带气象学报,2017,33(3):345-356.
[23] MENG Z,ZHANG F. Test of an ensemble Kalman filter for mesoscale and regional-scale data assimilation. Part III:Comparison with3DVAR in a real-data case study[J]. Mon Wea Rev, 2008, 136(2):522-540.
[24] MENG Z, ZHANG F. Test of an ensemble Kalman filter for mesoscale and regional-scale data assimilation. Part IV:Performance over awarm-season month of June 2003[J]. Mon Wea Rev, 2008, 136(10):3 671-3 682.
[2] ANDERSON J L. An ensemble adjustment Kalman Filter for data assimilation[J]. Mon Wea Rev, 2001,129(12):2 884-2 903.
[3] EVENSEN G. Sequential data assimilation with a nonlinear quasi-geostrophic model using Monte Carlo methods to forecast errorstatistics[J]. Geophys Res, 1994,99(5):10 143-10 162.
[4] EVENSEN G, LEEUWEN P J. Assimilation of GEOSAT altimeter data for the Aghulas current using the ensemble Kalman filter with aquasigeostrophic model[J]. Mon Wea Rev, 1996,124(1):85-96.
[5]张爱忠,齐琳琳,纪飞,等.资料同化方法研究进展[J].气象科技,2005,33(5):385-394.
[6]杨毅,弓中强,王金艳,等.时间扩展取样集合卡尔曼滤波同化模拟探空试验研究[J].气象学报,2012, 70(1):101-108.
[7] HOUTEKAMER P L, MITCHELL H L. Data assimilation using an ensemble Kalman filter technique[J]. Mon Wea Rev, 1998, 126(3):796-811.
[8] WHITAKER J S, HAMILL T M. Ensemble data assimilation without perturbed observations[J]. Mon Wea Rev, 2002, 130(12):1 913-1 924.
[9] ZHANG S, LI H, ZHANG W, et al. Estimating the soil moisture profile by assimilating near surface observations with the EnsembleKalman Filter(En KF)[J]. Adv Atmos Sci, 2005, 22(6):936-945.
[10]朱江,汪萍.集合卡尔曼平滑和集合卡尔曼滤波在污染源反演中的应用[J].大气科学, 2006, 30(6):871-882.
[11]刘寅,何光鑫,刘佳伟,等.雷达资料的3DVAR和EnSRF直接同化方法对一次暴雨预报的影响分析[J].热带气象学报,2014,30(5):894-904.
[12]高士博,闵锦忠,黄丹莲. EnKF局地化算法对雷达资料同化的影响研究[J].大气科学学报, 2016, 39(5):633-642.
[13]庄照荣,薛纪善,李兴良. GRAPES集合卡尔曼滤波资料同化系统Ⅱ:区域分析及集合预报[J].气象学报,2011,69(6):860-871.
[14] MITCHELL H L, HOUTEKAMER P L, PELLERIN G. Ensemble size, balance, and model-error representation in an ensemble Kalmanfilter[J]. Mon Wea Rev, 2002, 130(11):2 791-2 808.
[15] WU X, LI W, HAN G, et al. An adaptive compensatory approach of the fixed localization in the EnKF[J]. Mon Wea Rev, 2015, 143(11):4 714-4 735.
[16] THOMPSON T E, WICKER L J, WANG X. Impact from a volumetric radar-sampling operator for radial velocity observations withinEnKF supercell assimilation[J]. J Atmos Oceanic Technol, 2012, 29(10):1 417-1 427.
[17] HAMRUD M, BONAVITA M, ISAKSEN L. EnKF and Hybrid Gain Ensemble Data Assimilation. Part I:EnKF Implementation[J].Mon Wea Rev, 2015, 143(12):4 847-4 864.
[18] HOUTEKAMER P L, MITCHELL H L. A sequential ensemble Kalman filter for atmospheric data assimilation[J]. Mon Wea Rev, 2001,129(1):123-137.
[19] SKAMAROCK W C, KLEMP J B, JIMY D H, et al. A description of the advanced research WRF version 3[R]. NCAR Tech, Note, 2008:113.
[20] BARKER D M, HUANG W, GUO Y R, et al. A three-dimensional variational data assimilation system for MM5:Implementation andinitial results[J]. Mon Wea Rev, 2004, 132(4):897-914.
[21] ZHU L, WAN Q, SHEN X, et al. Prediction and predictability of high-impact western Pacific landfalling Tropical Cyclone Vicente(2012)through convection-permitting ensemble assimilation of Doppler radar velocity[J]. Mon Wea Rev, 2016,144(1):21-43.
[22]朱磊,万齐林,刘靓珂,等.雷达径向风资料EnKF同化应用对Vicente(2012)登陆台风强度变化过程预报的影响试验研究[J].热带气象学报,2017,33(3):345-356.
[23] MENG Z,ZHANG F. Test of an ensemble Kalman filter for mesoscale and regional-scale data assimilation. Part III:Comparison with3DVAR in a real-data case study[J]. Mon Wea Rev, 2008, 136(2):522-540.
[24] MENG Z, ZHANG F. Test of an ensemble Kalman filter for mesoscale and regional-scale data assimilation. Part IV:Performance over awarm-season month of June 2003[J]. Mon Wea Rev, 2008, 136(10):3 671-3 682.