Sergey Frolov;
Douglas R. Allen;
Craig H. Bishop;
Rolf Langland;
Karl W. Hoppel;
David D. Kuhl
NRL
The 8th EnKF Data Assimilation Workshop
First application of the local ensemble tangent linear model (LETLM) to a realistic model of the global atmosphere
Talk:
frolov2018_letlm.pptx
The local ensemble tangent linear model (LETLM) provides an attractive approach to develop easy-to-maintain, computationally scalable tangent linear model and the adjoint of a complex non-linear system. In this paper, we report on the first application of the LETLM in the context of a realistic model of the global atmosphere. This paper tests the ability of the LETLM to propagate an analysis increment over a 6-hour period in a low resolution model of the global atmosphere (T47 resolution equivalent to 2.5 degree equatorial grid spacing). We find that both the traditional TLM and the LETLM have significant skill over persistence everywhere in the atmosphere, except for temperature in the planetary boundary layer. We find that the LETLM was, on average, more accurate than the traditional TLM (by about 20% in the troposphere and 10% overall). Our sensitivity studies showed that the LETLM was most sensitive to the number of ensemble members, with the performance of the LETLM gradually improving with the ensemble size up to the maximum size attempted (400). Inclusion of physics in the LETLM ensemble lead to a significant improvement in representation of the boundary layer winds (up to 50%), in addition to winds and temperature in the free troposphere and in the upper stratosphere/lower mesosphere. Computationally, the LETLM was about an order of magnitude more expensive than a traditional TLM. However, significant cost savings can be achieved in the LETLM, since most of these computations can be pre-computed before the four-dimensional variational data assimilation is executed.
The local ensemble tangent linear model (LETLM) provides an attractive approach to develop easy-to-maintain, computationally scalable tangent linear model and the adjoint of a complex non-linear system. In this paper, we report on the first application of the LETLM in the context of a realistic model of the global atmosphere. This paper tests the ability of the LETLM to propagate an analysis increment over a 6-hour period in a low resolution model of the global atmosphere (T47 resolution equivalent to 2.5 degree equatorial grid spacing). We find that both the traditional TLM and the LETLM have significant skill over persistence everywhere in the atmosphere, except for temperature in the planetary boundary layer. We find that the LETLM was, on average, more accurate than the traditional TLM (by about 20% in the troposphere and 10% overall). Our sensitivity studies showed that the LETLM was most sensitive to the number of ensemble members, with the performance of the LETLM gradually improving with the ensemble size up to the maximum size attempted (400). Inclusion of physics in the LETLM ensemble lead to a significant improvement in representation of the boundary layer winds (up to 50%), in addition to winds and temperature in the free troposphere and in the upper stratosphere/lower mesosphere. Computationally, the LETLM was about an order of magnitude more expensive than a traditional TLM. However, significant cost savings can be achieved in the LETLM, since most of these computations can be pre-computed before the four-dimensional variational data assimilation is executed.
Thanks to McGill's Atmospheric and Oceanic Sciences department for hosting this web-site.