Cycle 29r1 summary of changes

5 April 2005Intoduction of Cycle 29r1. This version includes the following changes:

  • a new moist boundary layer scheme that generates more stratocumulus clouds in subtropical highs, and has shown to be better at generating low level clouds in some anticyclonic conditions such as those found in December over continental Europe;
  • a wavelet formulation introduced in the representation of background error statistics (Jb), the tuning of which has been based on Data Assimilation ensemble runs with a special treatment for ozone statistics
  • a revised use of surface pressure observations: airport hourly reports (METAR) are active (about 25,000 per day), all surface pressure data are subject to an adaptive bias correction scheme and get a higher weight in the assimilation (reduced observation error assumption); PAOB observations are not used any more;
  • MODIS winds from AQUA are activated (previously only winds from TERRA were used) and their observation error was reduced; 10 AIRS channels have been blacklisted due to their negative impact on stratospheric humidity;
  • a bug in the first time step of the semi-Lagrangian treatment of the physics was fixed; snow cover tile coupling has been revised; a new dissipation source function has been introduced for the oceanic wave modelisation;
  • several technical implementations paving the way for the assimilation of rain, of GPS data and for the variational bias correction of satellite radiances have been introduced but not activated yet
  • The impact of data assimilation changes on the scaling of EPS initial perturbations was tested and found negligible, therefore scaling parameters were not changed. The use of Nonlinear Normal Mode Initialisation for the EPS Control and EPS initial perturbations has been discontinued.
  • The overall impact of these changes has been found to be meteorologically small during the test period (July 2004 to implementation of the new cycle), although some clear beneficial impact could be seen occasionally for cloud cover.
  • The change to the new cycle is transparent for users apart from a change in the atmospheric model identification number in GRIB headers, which will be set to 125.
  • Implementation of a revised formulation for wave dissipation in the ocean wave model.
  • The wave model identification numbers will also change and will be:
    • 116 for the Global Wave Model and
    • 216 for the European Waters Wave Model


Cycle changes

Please note

Prior to 2006 information on cycle upgrades has been migrated from the old website to the new website "as is". The information has not been reviewed during the migration.