Annual WWW Technical Progress Report
On the Global Data Processing System 1999
Meteorological Service Singapore
1. SUMMARY OF HIGHLIGHTS
All critical operational systems were converted to be Y2K ready in
March 1999. Contingency procedures were also developed to address unforeseen Y2K-related
disruptions. As part of Y2K contingency, critical aviation weather information are also
placed on the Internet for access during the rollover period.
In the second half of 1999, the data communication links which
facilities the exchange of GTS data and numerical products between the meteorological
services of Indonesia, Melbourne and Singapore, were upgraded to higher speeds.
In December 1999, the NEC SX3 was replaced by a NEC SX-4 supercomputer.
New networking and graphical visualisation equipment has been acquired
to replace the existing SUN server and workstations. The new equipment include Silicon
Graphics Origin 2000 server, Onyx2 graphic system and Octane visual workstations. The new
network supports Gigabit Ethernet, 10/100BaseT Ethernet and FDDI.
2. EQUIPMENT IN USE
The equipment for the meteorological data processing system include
a NEC SX4 supercomputer (16GB main memory, 128 GB disk storage, 16 Gflops), 2 Fujitsu
M1600 mainframes, a number of mini-computers (Fujitsu DS90s, SUN server, VAX),
workstations (SUN Sparcs, Silicon Graphics) and a network of PCs.
3. DATA AND PRODUCTS FROM GTS IN USE
Data
The following reports were received daily. Figures denote the daily
average number for the respective reports.
Products
GRIB ECMWF (from GTS)
GRIB EGRR (from SADIS system and point-to-point computer link)
GRIB KWBC (from SADIS system)
GRIB JMA (from Distributed Data Base server of RSMC Tokyo)
4. DATA INPUT SYSTEM
Automated.
5. QUALITY CONTROL SYSTEM
Quality control checks are performed during the decoding, pre-analysis
and analysis stages. These include climatological check, vertical and horizontal
consistency check and gross error check.
6. MONITORING OF THE OBSERVING SYSTEM
Monitoring is carried out at the national level.
7. FORECASTING SYSTEM
7.1 SYSTEM RUN SCHEDULE
Operational NWP runs are carried out twice a day for the 0000 UTC
analysis.
|
Early Run |
Final Run |
Pre-processing : |
0330 UTC |
0530 UTC |
Analysis : |
0340 UTC |
0540 UTC |
GSM 3-day forecast : |
0400 UTC |
0600 UTC |
LAM 2-day forecast : |
0440 UTC |
0640 UTC |
GSM 4-10 day forecast : |
- |
0700 UTC |
7.2 MEDIUM-RANGE FORECASTING SYSTEM (4-10 days)
7.2.1 Data assimilation, objective analysis and initialisation
Global Analysis Model (GAM) : no change
7.2.2 Model
Global Spectral Model (GSM) : no change
7.2.3 Numerical weather prediction products
No change.
7.2.4 Operational techniques for application of NWP products
The lowest level wind forecasts from the UKMO and JMA models are used
to drive the sea-state model, while selected FLM forecasts are used as input fields for
the air dispersion model.
7.3 SHORT-RANGE FORECASTING SYSTEM (0-72 hrs)
7.3.1 Data assimilation, objective analysis and initialisation
No change.
7.3.2 Model
FLM-12 (40.0E-180.0E; 50.0N-45.0S) and VFM-13 (92.4E-125.8E;
20.6N-12.2S)
7.3.3 Numerical weather prediction products
No change.
7.3.4 Operational techniques for application of NWP products
NWP products such as winds, relative humidity and precipitation are
used as guidance tools in operational forecasting.
7.4 SPECIALISED FORECASTS
7.4.2 Model
Atmospheric dispersion forecasts
A long-range Lagrangian dispersion model and a forward trajectory
model (adapted from the JMA).
Wave forecasts
The third generation WAM (Wave Modeling) model is run for two
domains covering the globe (coarse run) and the Asia-Pacific region 60N-20S; 60E-180E
(nested run).
7.4.3 Numerical weather prediction products
Air dispersion models :
3-dimensional forward trajectories starting at 500m, 1500m and
3000m above the ground; time-integrated air concentrations within 500m layer above the
ground; surface deposition (dry) of airborne pollutants.
7.4.4 Operational techniques for application of NWP products
The main application of the dispersion models is to generate
forecast products for environmental emergencies such as episodes of smoke haze and
volcanic ash, and nuclear accidents.
8. PLANS FOR THE FUTURE
The Graphics Visualisation System is being revamped to optimise the
resources offered with the new equipment.
The HYSPLIT-4 (Hybrid Single_Particle Lagrangian Integrated
Trajectories) dispersion model will be implemented with the
assistance of NOAA Air Resources Laboratory to provide improved dispersion predictions
during the regional smoke haze episodes.
Explore the adaptation and implementation of a more advanced high
resolution limited-area NWP model to replace the current models (FLM and VFM).
|