COMPARISON RESEARCH OF ENKF AND PARTICLE FILTER IN A SIMPLE MODEL
-
摘要: 集合卡尔曼滤波和粒子滤波是大气海洋领域两种先进的数据同化方法。理论上讲，粒子滤波克服了集合卡尔曼滤波中先验分布的高斯假定。但现有的关于两种方法的比较研究不够全面和系统，基于简单的洛伦兹63模式，重点对基于确定性集合卡尔曼滤波和均权重粒子滤波的数据同化方法开展对比分析，通过对观测误差和模式误差的不同配置，设计了四组试验着重研究两种方法相同试验条件下的同化效果。试验结果表明：与采用最优膨胀系数的集合卡尔曼滤波的同化方法相比，均权重粒子滤波的均方根误差更加依赖于观测信息的质量，但最优膨胀因子的集合卡尔曼滤波的均方根误差低于粒子滤波同化方法。Abstract: The Ensemble Kalman Filter (EnKF) and Particle Filter (PF) are two kinds of advanced data assimilation methods in the atmospheric and oceanic research. Theoretically, the PF overcomes the Gaussian assumption of the priori distribution in EnKF. But up to now comparison and research on them are incomplete and unsystematic. Based on the Lorenz-63 model and different settings of the observational and model errors, this paper designed four sets of experiments to analyze and compare one kind of deterministic EnKF with the equal-weight PF. The results show that the RMSEs of the EnKF with the optimallyinflatedfactor are lower while those of the equal-weight PF are more dependent on the observation quality.
-
Key words:
- atmospheric environment /
- data assimilation /
- EnKF /
- equal-weight particle filter
-
表 1 四组试验设置
试验 σ ρ β 观测误差 试验1 9.95 28 8/3 无 试验2 9.95 28 8/3 有 试验3 9.95+白噪声 28+白噪声 8/3+白噪声 无 试验4 9.95+白噪声 28+白噪声 8/3+白噪声 有 
下载: 导出CSV
-
[1] ANDERSON J L. An ensemble adjustment Kalman filter for data assimilation[J]. Mon Wea Rev, 2001, 129(12): 2884-2903. [2] ANDERSON J L. A local least squares framework for ensemble filtering[J]. Mon Wea Rev, 2003, 131(4): 634-642. [3] 王晨稀.热带气旋集合预报研究进展[J].热带气象学报, 2013, 29(4): 698-704. [4] BOCQUET M, PIRES C A, WU L. Beyond Gaussian statistical modeling in geophysical data assimilation[J]. Mon Wea Rev, 2010, 138(8): 2997-3023. [5] 马旭林, 陆续, 于月明, 等.数值天气预报中集合-变分混合资料同化及其研究进展[J].热带气象学报, 2014, 30(6): 1188-1195. [6] 沈浙奇, 唐佑民, 高艳秋.集合资料同化方法的理论框架及其在海洋资料同化的研究展望[J].海洋学报, 2016, 38(3): 1-14. [7] VAN LEEUWEN P J. A variance-minimizing filter for large-scale applications[J]. Mon Wea Rev, 2003, 131(9): 2071-2084. [8] GORDON N J, SALMOND D J, SMITH A F M. Novel approach to nonlinear/non-Gaussian Bayesian state estimation[C]//IEE Proceedings F-Radar and Signal Processing. IET, 1993, 140(2): 107-113. [9] SNYDER C, BENGTSSON T, BICKEL P, et al. Obstacles to high-dimensional particle filtering[J]. Mon Wea Rev, 2008, 136(12): 4629-4640. [10] ZHU M, VAN LEEUWEN P J, AMEZCUA J. Implicit equal-weights particle filter[J]. Q J R Meteo Soc, 2016, 142(698): 1904-1919. [11] ANDERSON J L, ANDERSON S L. A Monte Carlo implementation of the nonlinear filtering problem to produce ensemble assimilations and forecasts[J]. Mon Wea Rev, 1999, 127(12): 2741-2758. [12] LORENZ E N. Deterministic nonperiodic flow[J]. J Atmos Sci, 1963, 20(2): 130-141. [13] HAN G J, ZHANG X F, ZHANG S, et al. Mitigation of coupled model biases induced by dynamical core misfitting through parameter optimization: simulation with a simple pycnocline prediction model[J]. Nonlinear Processes in Geophysics, 2014, 21(2): 357-366. [14] ZHANG S, HARRISON M J, ROSATI A, et al. System design and evaluation of coupled ensemble data assimilation for global oceanic climate studies[J]. Mon Wea Rev, 2007, 135(10): 3541-3564. [15] VAN LEEUWEN P J. Nonlinear data assimilation in geosciences: an extremely efficient particle filter[J]. Q J R Meteo Soc, 2010, 136(653): 1991-1999. [16] VAN LEEUWEN P J, ADES M. Efficient fully nonlinear data assimilation for geophysical fluid dynamics[J]. Comp Geos, 2013, 55: 16-27. [17] VAN LEEUWEN P J. Efficient nonlinear data-assimilation in geophysical fluid dynamics[J]. Comp Fluids, 2011, 46(1): 52-58. [18] ADES M, VAN LEEUWEN P J. An exploration of the equivalent weights particle filter[J]. Q J R Meteo Soc, 2013, 139(672): 820-840. [19] WU X, ZHANG S, LIU Z, et al. Impact of geographic-dependent parameter optimization on climate estimation and prediction: simulation with an intermediate coupled model[J]. Mon Wea Rev, 2012, 140(12): 3956-3971. [20] 王晨稀.基于随机全倾向扰动的台风路径集合预报试验[J].热带气象学报, 2015, 31(1): 32-42. -
点击查看大图
图(13) / 表(1)
计量
- 文章访问数: 1103
- HTML全文浏览量: 26
- PDF下载量: 532
- 被引次数: 0
粤公网安备 4401069904700003号