Revisions during 2002

22 January 2002 Introduction of Cycle 24r3. This version includes the following important changes that together will affect all components of the system (Data assimilation, atmospheric and oceanic waves forecasts, EPS):

1. More data are activated (SeaWinds data from QUIKSCAT, less thinning of aircraft observations, more intelligent thinning and better scan correction of ATOVS radiances);
   
2. Pre-processing (bias correction) of SSMI data and redundancy checks for SYNOP, SHIP, DRIBU, AIREP, TEMP and PILOT observation are refined;
3. 4D-var analysis algorithms are upgraded:
   • pre-conditioning is added to the minimisation, resulting in a 40% reduction of the number of adiabatic iterations;
   • correlation functions slightly revised (compacted);
   • the radiative transfer model used to assimilate satellite radiances has been completely re-written;
   • the observation timeslot has been reduced from 1 hour to 30 minutes;
     
4. Model changes include a new finite-element vertical discretization, small changes in the convective precipitation scheme and supersaturation checks, and an improved temporal scheme for oceanic waves generation;
   
5. Initial EPS perturbations in the tropics are included. The perturbations are generated for a maximum of four target areas by Gaussian sampling of the 5 leading diabatic singular vectors for each area. The Caribbean (0°-25°N and 100°-60°W) is always target area, as is every tropical storm of category larger than 1 between 25°N and 25°S. (In the event that these criteria produce more than four target areas, the closest areas are merged.)

9 April 2002 Introduction of Cycle 25r1. This version includes the following changes:

1. A revised shortwave radiation scheme with variable effective radius of liquid cloud water;
2. Retuning of the land surface (TESSEL) parametrization to reduce winter/spring warm biases in low-level temperatures;
3. Improved physics for the oceanic wave model;
4. Improved wind-gust post-processing;
5. The activation of new data streams in the assimilation: water-vapour radiances from Meteosat-7, SBUV and GOME ozone data, and European wind profilers;
6. A bug in the convective momentum transfer was also fixed;