The orographic drag scheme, developed by NOAA's Global Systems Laboratory, Physical Sciences Laboratory and Environmental Modeling Center is a set of subgrid-scale orographic drag parameterizations that calculate momentum tendencies due to the effects of unresolved topography. The drag forces they represent are those due to: 1) meso-scale gravity (mountain) waves that propagate vertically and break in the free atmosphere of the troposphere, stratosphere and above; 2) low-level flow blocking; 3) turbulent orographic form drag (TOFD), which is generated by turbulent pressure perturbations that are correlated with the terrain slope. The distinction between the meso-scale and turbulent-scale gravity waves are that the former are generated by topography with horizontal scales on the order of 5 km and greater, which can support vertical propagation through the typical static stabilities found in the free atmosphere, while the latter are generated by topography with smaller horizontal scales down to about 1 km.
The GWD and flow-blocking scheme is based on Kim and Doyle (2005) [108] and Choi and Hong (2015)[38] and its code originated from the NCAR Weather Research and Forecasting (WRF) model that was implemented by S. Hong. The TOFD scheme is adapted from Beljaars et al. (2004)[14].
All orographic GWD schemes require static input data files that contain statistical information about the subgrid terrain within each model grid cell, such as the standard deviation of the subgrid topography, which comes from the high-resolution USGS 30-second GMETED2010 dataset. These data files augment the usual "oro_data.tile*.nc" files, which contain orographic height data and GWD static data for the orographic drag parameterization. The static data files for the meso-scale GWD and blocking schemes are named "oro_data_ls.tile*.nc". The source topography for the datasets are calculated from a 2.5-minute lat-lon grid to filter out small-scale topographic variations. The static data files for the small-scale GWD and TOFD schemes are named "oro_data_ss.tile*.nc". The data is from the 30-second topographic dataset, but band-passed filtered from ~20km down to ~2km as per Beljaars et al.(2004) [14].
The meso-scale GWD and blocking schemes are explicitly tapered off from horizontal grid resolutions starting at ~13 km down to 3 km resolution, at and below which the scheme is not active.