Impacts of parameterized orographic drag on the Northern Hemisphere winter circulation

TitleImpacts of parameterized orographic drag on the Northern Hemisphere winter circulation
Publication TypeMiscellaneous
Year of Publication2015
AuthorsSandu, I, Bechtold, P, Beljaars, A
Secondary TitleTechnical Memorandum
Number763
Abstract

<p>
&nbsp;</p>

<p>
A recent intercomparison exercise proposed by the Working Group for Numerical Experimentation (WGNE) revealed that the parameterized, or unresolved, surface stress in weather forecast models is highly model dependent, especially over orography. Models of comparable resolution differ over land by as much as 20% in zonal mean total subgrid surface stress (τ<sub>tot</sub>). The way τ<sub>tot</sub> is partitioned between the different parameterizations is also model dependent. In this study, we simulated in a particular model an increase in τ<sub>tot</sub> which is comparable with the spread found in the WGNE intercomparison. This increase was simulated in two ways, namely by increasing independently the contributions to τ<sub>tot</sub> of the turbulent orographic form drag scheme (TOFD) and of the orographic low level blocking scheme (BLOCK).We demonstrate that such an increase in ttot significantly affects the representation of the Northern Hemisphere winter circulation, both in ten-day weather forecasts and in seasonal integrations. However, the magnitude of the changes in circulation strongly depends on which of the schemes is modified. In short and medium range forecasts, the response is stronger when the TOFD stress is increased, while on seasonal timescales the effects are of comparable magnitude, although different in detail. At these time scales, the BLOCK scheme affects the lower stratosphere through changes in the resolved planetary waves which are associated with surface impacts, while the TOFD effects are mostly limited to the lower troposphere. These results suggest that the partitioning of τ<sub>tot</sub> between the two schemes plays an important role at all timescales.</p>

<p>
&nbsp;</p>