Sensitivity of medium‐range weather forecasts to the use of an envelope orography

• Published in: Quarterly Journal of the Royal Meteorological Society
• Main Authors: Jarraud, M., Simmons, A. J., Kanamitsu, M.
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 1988
• Subjects: Antarctic; Antarctic Peninsula; Antarc*
• Online Access: http://dx.doi.org/10.1002/qj.49711448208; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1002%2Fqj.49711448208; https://rmets.onlinelibrary.wiley.com/doi/pdf/10.1002/qj.49711448208

Abstract The performances of grid‐square mean and enhanced ‘envelope’ orographies are compared in a set of 10‐day forecasts performed using the ECMWF spectral model at horizontal resolutions T21, T42, T63 and T106. Twenty‐four cases chosen at monthly intervals from May 1983 to April 1985 are examined, and for diagnosis of results for the northern hemisphere these are divided into 12 ‘winter’ (November to April) and 12 ‘summer’ (May to October) situations. The winter results in many respects confirm those of previous studies. Apart from T21 resolution, use of the envelope orography in this season is clearly beneficial in general, and synoptic assessment illustrates how mountain barrier effects are better represented by the envelope. Conversely, in summer the envelope has a detrimental effect (according to some, though not all measures) at T42 and T63 resolutions, though not at T106. This appears to be at least in part due to the more northerly position of the (northern hemisphere) jet in summer, which interacts with mountains that are less of a barrier, with T106 giving a better separation of the localized orographic features of these regions. Some problems are seen at all seasons with the envelope representation of the Asian relief. Results for the southern hemisphere are, not surprisingly, generally less sensitive to the representation of mountains, the principal impact coming from the southern Andes and Antarctic Peninsula to the south. A small beneficial impact of the envelope orography is found in the tropics. In individual cases, the use of envelope rather than mean orographv gives rise to local forecast differences near mountains which tend to propagate and amplify (principally on synoptic scales), following the upper‐level flow. Much of the hemisphere can be influenced in 7–10 days. Ensemble‐ and time‐averaged errors are generally reduced by use of the envelope, though not dramatically, and not at all locations.