7 January 1992 In model cycle 41 the horizontal
diffusion was increased and the time-stepping for cumulus momentum transfer
was changed. The diffusion was modified again on 15 January 1992.
20 February 1992 PAOB surface data from the
Australian Bureau of Meteorology re-introduced for use in the analysis.
13 May 1992 Cycle 42
2 June 1992 A change to the determination of
sea ice in the SST analysis was introduced, leading to an improved description
of the ice edge.
9 June 1992 Two changes were introduced in the
analysis:
- humidity data from SYNOP observations were excluded from the analysis,
to reduce excessive convective precipitation in short and early medium-range
forecasts;
- a redundancy check on aircraft data was introduced so that new high
density data (ACARS) can be correctly used in the O1 scheme.
23 June 1992 The temperature data from the NOAA
satellites used in the analysis of the Northern Hemisphere are now derived
from cloud cleared radiance data received from NESDIS, using a 1D-variational
inversion technique.
17 August 1992 A
change was made to the forecast model (cycle 43), to include:
- introduction of vertically non-interpolating semi-Lagrangian scheme
and smaller time filter;
- suppression of inversion clouds at the lowest three model levels
rather than just the lowest level;
- introduction of a prognostic equation for sea-ice temperature to
replace the use of climatological values;
- other minor parametrization changes.
1 September 1992 Cycle 44
7 December 1992
A modification to the model post-processing was implemented (model cycle
45) to improve the extrapolation of temperature and MSL pressure under
model orography. The resulting extrapolated fields better continue the
free atmosphere distribution of fields and show much less dependence
on the overlying orography.