Forcing datasets (also called as analysis) of coarse resolution from global models already might have errors and the errors might further increase when downscaled for use in WRF due to interpolation of values. Since in nudging wed try to nudge our model outputs towards the analysis, doesn't that put the nudging technique at a disadvantage?
Global model reanalyses have an assimilation of observations in them. Therefore, nudging with such data limits the errors in WRF simulation. Also, users have a choice to not nudge with in the planetary boundary layer. Moreover, nudging does not necessarily replaces WRF meteorology by global data, instead it checks the consistencies with in a factor defined by the user.
Hi,
Even though observations are used, the corrections are done only at the grid points.
I think, while downscaling, errors might creep into the LBC for the nested domain because the global model would not have taken into account the small scale variability of a nested domain like steep or complex terrain, diverse landuse or other topographic features. But in nudging, irrespective of its strength or vertical range of application we are comparing to those interpolated values only, which might have errors.
In principle, robust nudging can PUSH the model toward a relatively smooth state and this may constrain its (model) ability to simulate extremes efficiently.
Hi B. H. Vaid,
Indeed that is possible. The strength of the nudging can be adjusted as per requirement.
However my question is from a conceptual point of view, with the practicality of the process involved with the nudging procedure.
I suggest why don't you give a try to understand the process involved by changing the strength (i.e., with varying spectral nudging strength; say slight to robust) of nudging in the model run. It may assist you to differentiate the varied processes involved and explore the effect of nudging as well.
Best regards,
Vaid, B. H.
Spectral nudging should be avoided altogether because nudging is creating spurious wavenumbers to correct internal wave drift of the regional climate model from reanalysis and GCMs. It prevents the interior of the mesoscale model diverging from the large-scale circulation. Spectral nudging suppresses internal model variability and gives better results for large scales but at the expense of a reduced variability at smaller scales. It is able to nearly cancel out the differences independent of the domain size, i.e., differences for the simulations with spectral nudging are of the same order for the small and large domain. That means the nudged result is almost similar to the boundary values. So, what is the need for downscaling? That's why I suggest nudging should be avoided in regional climate modeling.
Hi Gemechu Fanta Garuma ,
Thanks for your inputs.
Let consider that I can only use nudging approach at the moment.
I read that divergence from the boundary conditions increase with the domain size. In that case would it be a good strategy to use higher strength spectral nudging only in the largest domain, and use very low/zero strength nudging in the interior domains?
Can you suggest other alternatives to nudging?
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Column