Global Data Processing System
Progress Report for 1999
HONG KONG, CHINA
1. Summary of highlights
A new computer, CRAY SV1-1A with 16 CPUs and a peak
performance of 19.2 Gflop/s, was commissioned in 1999 at the Hong Kong Observatory (HKO).
The Regional Spectral Model (RSM) became operational on the new computer by the end of the
year with an outer horizontal domain of 60km nesting into a 20km inner domain. Rainfall
information from radar and rain gauges was routinely incorporated into the RSM through a
physical initialization process.
2. Equipment in use at the centre
Current systems at the HKO with their major characteristics are listed
below:
Machine |
Quantity |
Peak performance |
No. of
CPU |
Memory |
Year of Installation |
SGI CRAY SV1-1A |
1 |
19.2 Gflop/s |
16 |
8 GB |
1999 |
SGI Origin 2000 |
1 |
4.0 Gflop/s |
8 |
1.5 GB |
1998 |
SUN E450 |
2 |
2.4 Gflop/s |
4 |
1 GB |
1998 |
SGI O2 |
2 |
0.4 Gflop/s |
1 |
1 GB |
1998 |
IBM RS/6000 590 |
3 |
0.3 Gflop/s |
1 |
128 MB |
1994 |
The CRAY SV1-1A is used to run the analysis and forecast system for
both the inner and outer domains of the RSM.
The SGI Origin 2000 is also deployed to support the operation of the
RSM. It is mainly used for data pre-processing and product generation based on the output
of the RSM.
One of the SUN E450 is used for provision of NWP products to support
aviation operations and for development of the RSM. The other is for running the rainstorm
nowcasting system to support the Central Forecasting Office (CFO). Two O2 are used as
graphics display workstations for visualization of nowcasting products and staging the web
server for RSM products.
One of the IBM 590 servers is used for running the operational
database, generating diagnostic products, plotting weather charts and producing graphics
displays to support the operations of the CFO and Airport Meteorological Office (AMO). The
other 590 servers are configured to take up the operational tasks in case the first one
breaks down. They are also used for the processing of climatological data and research.
3. Data and products from GTS in use
The number of reports received from GTS circuits on a typical day
in 1999 are given below:
SYNOP/SHIP 20 000
TEMP/PILOT 4 000
AIREP 3 000
SATEM/SATOB 2 000
About 900 GRID bulletins and 20 T6 fax charts are received each day on
the GTS.
On days with tropical cyclones in the region, additional GTS bulletins
(e.g. RADOB) and radio facsimile charts are also received.
4. Data input system
Automated.
5. Quality control system
For quality control of incoming data, adherence to prescribed
coding formats, internal consistency, and physical and climatological limits are checked
automatically.
External consistency check against the first guess field provided by
the forecast from the RSM is also carried out automatically.
Quality control of outgoing observational data originating from Hong
Kong is implemented to ensure conformity to WMO coding formats and to enforce checking
against internal consistency, time consistency as well as physical and climatological
limits.
6. Monitoring of observing system
Monitoring on the territorial level.
7. Forecasting system
For short-range (0-72 hours) forecasting, the HKO commenced
operation of the RSM by the end of 1999, replacing the Observatory Limited Area Model
(OLAM), the one degree resolution model adapted from the Very Fine Mesh Model (VFM65) of
Japan Meteorological Agency (JMA).
7.1 System run schedule
The outer 60km-RSM is run four times a day to produce 48-hour forecasts
for the area 9S - 59N, 65-152 E based on 00, 06, 12 and 18 UTC analysis data. The inner
20km-RSM is run 8 times a day for 24-hour forecasts for the area 1035 N, 100-128E
based on 00, 03, 06, 09, 12, 15, 18 and 21 UTC analyses.
7.2 Medium-range forecasting system (4-10 days)
Not implemented.
7.3 Short-range forecasting system (0-72 hours)
7.3.1 Data assimilation, objective analysis and initialization
Meteorological data assimilated by the analysis scheme of the RSM are
as follows:
(A) From GTS
SYNOP, SHIP surface data and ship data
TEMP, PILOT radiosonde and pilot data
AIREP aircraft data
SATEM satellite thickness data
TOVS, ATOVS virtual temperature profiles
SATOB satellite wind data
(B) From RSMC Data Serving System (DSS) of JMA
GMS digital data - total cloud amount, mean cloud top temperature and
its standard deviation for moisture bogus
GMS cloud motion vectors during tropical cyclone situations
(C) From NCEP data server
Daily sea surface temperature analysis at 1 degree resolution
(D) Locally generated data
Tropical cyclone bogus data during tropical cyclone situations
Three-dimensional multivariate optimal interpolation is performed four
times a day based on 00, 06, 12 and 18 UTC data for the 60km outer domain. Data cut-off
time is about three hours after the observation time. For the inner domain, the same
objective analysis scheme is performed 8 times a day based on 00, 03, 06, 09, 12, 15, 18,
and 21 UTC. Data cut-off time is about 2 hours. All analyses are applied to 36 vertical
levels.
The horizontal domains of both inner and outer models compose of 151 x
145 model grids in Mercator projection. The first guess fields of the models are provided
by their respective latest forecasts.
Hourly rainfall information derived from real-time calibration of radar
reflectivity with rain gauge data as well as from the GMS digital cloud data, are
incorporated into the model through a physical initialization process. In this process,
the moisture of the initial field (between the lifting condensation level and the cloud
top inferred from the cloud top temperature) at the point where rain is observed is
adjusted to allow precipitation process to be switched on. The heating rate of the
precipitation process is also adjusted to correspond to the rainfall amount observed. The
rainfall information in the hour preceding analysis time is used in the outer model. For
the inner model, pre-runs for 3 hours preceding analysis time are performed to incorporate
the rainfall information.
Non-linear normal mode initialization is performed before the forecast
model is run.
7.3.2 Model
The RSM is adapted from JMA with the following characteristics:
Basic equation |
Primitive hydrostatic equations |
Vertical |
Sigma-P hybrid coordinate, model top at
10hPa. |
Forecast parameters |
Ln(surface pressure), horizontal wind
components, virtual temperature, specific humidity. |
Initialization |
Non-linear normal mode initialization |
Physical processes
Radiation scheme
short wave |
Sugi et al.(1990)
Calculated every hour |
Long wave |
Calculated every hour |
Moisture processes |
|
Cumulus convection |
Arakawa-Schubert (1974) |
Mid-level convection |
Moist convective adjustment proposed by
Benwell and Bushby (1970) and Gadd and Keers (1970) |
Large-scale condensation |
Included |
Grid-scale evaporation and
Condensation |
Included |
Planetary boundary layer |
Scheme proposed by Troen and Mahrt (1986)
in which non-local specification of turbulent diffusion and counter-gradient transport in
unstable boundary layer are considered. |
Surface |
4-layer soil model |
|
Daily sea-surface temperature analysis
(fixed in forecast) |
|
Climatological snow and sea ice
distribution |
|
Climatological evaporation rate, roughness
length and albedo |
Numerical methods
Horizontal |
Double Fourier |
Vertical |
Finite difference |
Time |
Euler semi-implicit time integration, time
step of 280 seconds. |
Topography |
Envelope topography, derived from
30-second latitude/longitude resolution grid point topography data |
Horizontal diffusion |
Linear, second-order Laplacian |
Boundary conditions |
For the outer model, 6-hourly boundary
data including mean sea level pressure, wind components, temperature and dew point
depression at 15 pressure levels (1000, 925, 850, 700, 500, 400, 300, 250, 200, 100, 70,
50, 30, 20, 10 hPa ) and the surface, are provided by the Global Spectral Model of JMA.
For the inner model, hourly boundary data are provided by the outer
60km model. |
Further details on the formulation of the RSM are given in JMA(1997).
7.3.3 Numerical weather prediction products
For the outer 60km domain, the RSM produces primarily 3-hourly
numerical products which include pressure/ geopotential heights, wind, temperature, dew
point depression at 15 pressure levels ( 1000, 925, 850, 700, 500, 400, 300, 250, 200,
100, 70, 50, 30, 20, 10 hPa ) and the surface. Accumulated rainfall at the surface is also
produced. For the inner 20km domain, the same forecast elements as the above are produced
but at hourly intervals.
7.3.4 Operational techniques for application of NWP products
Post-processing of RSM prognostic data are performed and over 400 GIF
products are generated per model run which are made available to the forecasters for
reference through a web-based display.
These products include forecast time cross-section and tephigrams as
well as forecast rainfall distribution charts for Hong Kong. Local text forecasts based on
RSM prognostic data with warnings of thunderstorms and rainstorms, are generated
automatically. Pre-defined 3-D products of forecast temperature and cloud cover using
VIS5D are also produced.
Kalman filtering technique is employed to provide daily minimum and
maximum temperature forecasts in Hong Kong based on 60 km RSM prognostic data.
A weather map algorithm is employed in the RSM to produce hourly
weather map (fine area, cloudy area, cumulated rainfall contours) based on RSM prognostic
data. The actual IR satellite imagery for the same time, when available, is also
superimposed on the forecast weather map to facilitate visual verification of the weather
map product by forecasters.
8. Plan for the future
The verification of the RSM forecasts, particularly the prediction of
rainfall, will be focused. The feasibility of operating a local model at resolution of
below 20km is being studied.
9. References
Arakawa, A. and W.H. Schubert. 1974: Interaction of a Cumulus Cloud
Ensemble with the Large-Scale Environment, Part I. J. Atmos. Sci., 31: 674-701.
Benwell, G.R.R. and F.H. Bushby, 1970: A case study of frontal behavior
using a 10-level primitive equation model. Quart. Jour. Roy. Meteor. Soc., 96,287-296.
Gadd, A.J. and Keers, J.F. 1970 : Surface exchanges of sensible and
latent heat in a 10-level model atmosphere. Quart. Jour. Roy. Meteor. Soc., 96, 297-308.
Japan Meteorological Agency (JMA) : Progress report on numerical
weather prediction 1997. Appendix. 126 pp.
Sugi, M., K. Kuma, K. Tada, K. Tamiya, N. Hasegawa, T. Iwasaki, S.
Yamada and T. Kitade, 1990: Description and performance of the JMA operational global
spectral model (JMA-GSM88). Geophys. Mag., 43, 105-130.
Troen, I., and L. Mahrt, 1986: A simple model of the atmospheric
boundary layer: Sensitivity to surface evaporation. Boundary Layer Meteor., 37, 129-148.
|