Study of interior nudging for regional climate modeling using annual simulations: investigation of mean error and added value

Interior nudging for regional climate modeling applications is a controversial topic within the regional climate modeling community. The controversy stems from the need to correctly resolve the physics at time scales appropriate for regional climate modeling vs. potential limitations of the numerical conditions prescribed at the lateral boundaries. Despite the need to improve both the physics and fix potential limitations that exist at the lateral boundaries, there is a need to ensure that the length scales resolved by the driving data are retained.

This study is used to help determine the advantages and limitations of interior nudging schemes for regional climate modeling. Using the WRF model driven with NCEP-DOE AMIP-II Reanalysis for an annual simulation, we test two nudging techniques: the more traditional grid point nudging and spectral nudging vs. a no nudge simulation. Additional sensitivity simulations are performed with the interior nudging schemes. These simulations only use the coarser reanalysis information for interior nudging. Thus, one must carefully understand that the interior nudging doesn't suppress variability at the length scales resolved by the regional climate model. We examine the power spectra to determine if the nudging techniques are suppressing the length scales of interest. In addition, we use the NARR reanalysis data to determine the improvements in the mean error using the various nudging techniques.