The NCEP Regional Spectral Model: An Update

• Published in: Bulletin of the American Meteorological Society 1997-10, Vol.78 (10), p.2125-2143
• Main Authors: Juang, Hann-Ming Henry, Hong, Song-You, Kanamitsu, Masao
• Format: Article
• Language: English
• Subjects: Climate models; Earth, ocean, space; Environmental modeling; Exact sciences and technology; External geophysics; Forecasting models; Geophysics. Techniques, methods, instrumentation and models; Hydrostatics; Meteorology; Modeling; Perturbation methods; Physics; Spectral methods; Water pressure; Weather forecasting
• ISSN: 0003-0007; 1520-0477
• DOI: 10.1175/1520-0477(1997)078<2125:tnrsma>2.0.co;2

The National Centers for Environmental Prediction (NCEP)/Environmental Modeling Center regional spectral model (RSM) has been improved in several aspects since Juang and Kanamitsu. The major improvements of RSM are its efficiency and functionality. The change of the map factor in the semi-implicit scheme from a mean value to maximal value over the regional domain, the relaxation of the lateral boundary from explicit method to implicit method (or simple blending), and the local diffusion over areas of strong wind allowed the doubling of the model computational time step. The model physics was upgraded with the improvements in the operational global spectral model (GSM) and with an additional explicit cloud scheme. An option to run in either hydrostatic or nonhydrostatic mode has been introduced. Another option to run on a CRAY machine or on a workstation has been fully tested. The nesting process has been changed to provide the capability of nesting into a coarse-resolution RSM, besides the GSM, in a one-way fashion. Thus, multinesting becomes possible, even with different map projections. Regional data assimilation with a gridpoint version of statistical interpolation and the three-dimensional variational method on sigma surfaces has been incorporated. All the output has been encoded in GRIB format, so it can be read on different machines. The authors have tested the improved functionalities of the RSM over a broad range of applications, at resolutions between 80 and 10 km. The daily routine experimental forecasts over North America have acceptable performance. Because the perturbation method, used in the RSM, results in smaller computational error than the full field method, and because the consistency between the GSM and RSM allows for a better treatment of the lateral boundary, the RSM could be used to enhance the reanalysis and regional climate simulations that have long-range integrations. The RSM is also used in the regional ensemble experiments at NCEP. The model was also applied in case studies, such as the case of PYREX in the regional COMPARE project. Several institutions both in the United States and overseas started using the RSM, mostly for regional short-range forecast and climate modeling studies. The RSM has been scheduled to implement into operations at NCEP to possibly enhance the guidance on aviation and on daily weather forecast over Hawaii. The current version of the RSM is available to any institution requesting from the director of N