| dc.description.abstract | General circulation models (GCMs) are unable to represent local subgrid-scale features because of their coarse resolution. Thus, the nested regional climate modeling technique, also refer to as dynamical downscaling, which GCMs are used to provide the initial (ICs) and lateral meterological and surface boundary condition (LBCs), are developed to produce the regional-scale features. However, the simulation of Regional Climate Models (RCMs) tends to drive away from the driving field and the skill will damp rapidly with time. Through the spectral nudging, the RCMs can alleviate the problems.
Here, we use the Weather Research and Forecasting (WRF) model over East Asia to dynamically downscale the 0.5-degree Climate Forecast System Reanalysis (CFSR). We perform the three phase of experiments for the winter in 2009. In phase one, we compare the different downscaling skill with a grid spacing at 15km for climate, including the long- and short-term integration and long-term with spectral nudging; besides, we add the specific humidity in spectral nudging to improve the result of precipitation. In phase two, we discuss two sensitivity test of spectral nudging. In the last phase, we compare the dynamically downscaling between different information of LBCs and spectral nudging with a grid spacing at 5km for climate.
Compare to ERA interim, the downscaling simulation using the spectral nudging with additionally adding moisture not only can keep the large-scale pattern, but significantly improve the realistic result of precipitation in phase one. In phase two, though spectral nudging with more wavenumber can generate the more reasonable pattern, it constrains the feature at small scale and the precipitation will be underestimated because of immature system. Through the spectral nudging, the large-scale pattern can be handled, and the result of cold and warm start show the similar result. In the phase three, simulation with GCMs for information of LBCs and nudging produce the highest skill. The spectral nudging can generate realistic regional-scale climate information. In particular, spectral nudging which are configured to nudge moisture can revise the error to generate reasonable precipitation pattern that are not resolved by simply updating the LBCs as done traditionally. | en_US |