Accounting for the effect of observation errors on verification of MOGREPS

摘要

A number of articles have looked at the effect of observation errors on verification statistics. In this article, those methods are brought together in order to assess the importance of observation errors on verification statistics for the Met Office short-range ensemble prediction system (MOGREPS). The results indicate that the effect of observation errors is substantial - reducing the apparent skill of the forecast system by around 1 day in forecast lead time. The effect of observation errors is typically largest at short lead times when forecast errors are smallest. (c) Crown Copyright 2008. Reproduced with the permission of the Controller of HMSO. Published by John Wiley & Sons, Ltd.