Sensitivity tests for an ensemble Kalman filter for aerosol assimilation
Sensitivity tests for an ensemble Kalman filter for aerosol assimilation
复制标题
DOI：
10.5194/acp-10-6583-2010
复制
发表时间：
2010-07
影响因子：
6.3
通讯作者：
N. Schutgens;T. Miyoshi;T. Takemura;T. Nakajima
中科院分区：
文献类型：
--
作者：
N. Schutgens;T. Miyoshi;T. Takemura;T. Nakajima
我们利用 AERONET 的 AOT（气溶胶光学厚度）和 AAE（气溶胶 Å Angström 指数）观测结果，对全球气溶胶同化系统进行了灵敏度测试。同化系统采用集成卡尔曼滤波器，需要调整三个数值参数：集成大小 nens、局部补丁大小 npatch 和膨胀因子。此外，还进行了实验来测试系统各种实现的影响。例如，我们使用不同的发射集合处方或不同的观测组合。各种实验相互比较，并与独立的 AERONET 和 MODIS/Aqua 观测进行比较。这种同化导致 AOT 和 AAE 领域建模的显着改进。此外，剩余的误差大多是随机的，而对于没有同化的实验来说，它们大多是系统性的。此外，这些结果不太取决于我们的参数或设计选择。看来，局部斑块大小的值迄今为止对同化的影响最大，对于 nens = 20 的系综大小，同化已经充分收敛。同化 AOT 和 AAE 显然优先于同化两个不同波长的 AOT。相比之下，初始条件或超出两种模式（粗略和精细）的气溶胶描述几乎没有影响。我们还讨论了使用集成扩展作为分析的 AOT 和 AAE 字段的误差估计。我们表明，发射集合的一个非常常见的规定（每个网格单元中的独立随机修改）可能难以在预测集合中生成足够的分布。
We present sensitivity tests for a global aerosol as- similation system utilizing AERONET observations of AOT (aerosol optical thickness) and AAE (aerosol ˚ Angstr¨ om ex- ponent). The assimilation system employs an ensemble Kalman filter which requires tuning of three numerical pa- rameters: ensemble size nens, local patch size npatch and in- flation factor . In addition, experiments are performed to test the impact of various implementations of the system. For instance, we use a different prescription of the emission en- semble or a different combination of observations. The various experiments are compared against one- another and against independent AERONET and MODIS/ Aqua observations. The assimilation leads to significant im- provements in modelled AOT and AAE fields. Moreover re- maining errors are mostly random while they are mostly sys- tematic for an experiment without assimilation. In addition, these results do not depend much on our parameter or design choices. It appears that the value of the local patch size has by far the biggest impact on the assimilation, which has sufficiently converged for an ensemble size of nens = 20. Assimilating AOT and AAE is clearly preferential to assimilating AOT at two different wavelengths. In contrast, initial conditions or a description of aerosol beyond two modes (coarse and fine) have only little effect. We also discuss the use of the ensemble spread as an error estimate of the analysed AOT and AAE fields. We show that a very common prescription of the emission ensemble (in- dependent random modification in each grid cell) can have trouble generating sufficient spread in the forecast ensem- ble.