Cornick Matthew, Hunt Brian, Ott Edward, Kurtuldu Huseyin, Schatz Michael F
University of Maryland, College Park, Maryland 20742, USA.
Chaos. 2009 Mar;19(1):013108. doi: 10.1063/1.3072780.
Data assimilation refers to the process of estimating a system's state from a time series of measurements (which may be noisy or incomplete) in conjunction with a model for the system's time evolution. Here we demonstrate the applicability of a recently developed data assimilation method, the local ensemble transform Kalman filter, to nonlinear, high-dimensional, spatiotemporally chaotic flows in Rayleigh-Bénard convection experiments. Using this technique we are able to extract the full temperature and velocity fields from a time series of shadowgraph measurements. In addition, we describe extensions of the algorithm for estimating model parameters. Our results suggest the potential usefulness of our data assimilation technique to a broad class of experimental situations exhibiting spatiotemporal chaos.
数据同化是指结合系统时间演化模型，从一系列测量值（可能存在噪声或不完整）的时间序列中估计系统状态的过程。在此，我们展示了一种最近开发的数据同化方法——局部集合变换卡尔曼滤波器，在瑞利 - 贝纳德对流实验中对非线性、高维、时空混沌流的适用性。使用该技术，我们能够从一系列阴影图测量值的时间序列中提取完整的温度和速度场。此外，我们描述了用于估计模型参数的算法扩展。我们的结果表明，我们的数据同化技术对于广泛的呈现时空混沌的实验情况具有潜在的实用性。
