Date	Time	Title	Speaker	Affiliation
9 January	10am	The Value in Probabilities for Aerodrome Forecasts	Ross Keith	BoM, Townsville
13 February	10am	Application of Polarmetric Radar to Quantitative Rainfall Measurement in Sydney	Phillip Jordan	BoM Hydrology Section, Melbourne
20 February	No Seminar: AMOS Conference, Melbourne
27 February	10am	Characterization of cyclone evolution and monsoon variability	Jenni Evans	Penn State University, USA
6 March	10am	Evolution of the Global Observing System - A WMO Perspective	Paul Menzel	NOAA/NESDIS, USA
13 March	10am	Supergradient winds in tropical cyclones - the search for the smoking gun	Jeff Kepert	BMRC
19 March (Tues)	10am	Satellite imagery from NWP models	Hiromi Owada	JMA, Japan
20 March	10am	The Effect of Altimeter Sampling Patterns on Estimates of Wave Model Error Correlations	Diana Greenslade	BMRC
21 March (Thurs)	10am	Research Activities at the JMA/MRI and Climate Modelling in Japan	Hiroki Kondo	MRI/JMA, Tsukaba, Japan
22 March (Fri)	10am	Quantifying Predictability	Craig Bishop	Penn State University, USA
26 March (Tues)	10am	Tidal Effects of the Sun and Moon on Terrestrial Climate	Norman Treloar	Qld Centre for Climate App.
26 March (Tues)	2pm	Development of MPI Version of BMRC Global Model. Phases I & II: Eulerian Dynamics	Atsushi Kubota and Michael Naughton	Hiroshima City University and BMRC
27 March	10am	Patterns of Interannual Variability in the Wintertime Circulation: Potentially Predictable and Weather-noise components	Carsten Frederiksen	BMRC
3 April	10am	How well can we simulate the polar vortex?	Greg Roff	BMRC
10 April	10am	Public Affairs at the Bureau of Meteorology	Sarah Phillips	BoM, Public Affairs Unit
17 April	10am	Use of ATOVS radiances in a 50-level GASP Assimilation	Brett Harris	BMRC
24 April	10am	MIPs and FLOPS, The alphabet soup and Experimental Design	Bryant McAvaney	BMRC
1 May	10am	Describing subgrid-scale cloud variability in GCMs - How homogeneous are model clouds?	Christian Jakob	BMRC
8 May	9:30am	National fog and low cloud analyses	Gary Weymouth	BMRC
14 May (Tues)	10am	Development and assessment of the Forecasting Ocean Assimilation Model (FOAM) system	Mike Bell	UK Met Office
15 May	10am	Developments for a seasonal forecast system at NSIPP	Michele Rienecker	NASA Goddard Space Flight Center
16 May (Thurs)	10am	SSALTO/DUACS: satellite altimetry for operational oceanography	Pierre-Yves Le Traon	CLS, Toulouse
22 May	10am	Skill of radar-based nowcast algorithms	Beth Ebert	BMRC
24 May (Fri)	10am	Numerical Weather prediction and weather forecasting at high latitudes	Neil Adams	BoM and CRC, Tasmania
29 May	10am	The Aerosonde Grows Up	Greg Holland	Aerosonde Limited
6 June (Thurs)	10am	The conformal-cubic atmospheric model	John McGregor	CSIRO Atmospheric Research
19 June	10am	GCOS - The Global Climate Observing System	Mike Manton	BMRC
26 June	10am	Valuing the benefits of Met Services: An economic framework	Don Gunasekera	BMRC
2 July (Tues)	11am, 5th floor	Soaring the Polar Vortex to 100,000 feet	Elizabeth Carter and Edward Teets	Firnspiegel, CA USA, and NASA Dryden, USA.
3 July	10am	Does Climate Matter?	Mary Voice	ex NCC
10 July	10am	Statistical downscaling of rainfall: can we use an analogue approach?	Bertrand Timbal	BMRC
17 July	10am	Low altitude wind shear associated with convection and its impact on aviation	Rodney Potts	BMRC
19 July (Fri)	10am	Rossby Wave Hydraulics	Simon Clarke	Monash University
22 July (Mon)	10am, 5th floor	An open forum on the role and future work programs of the IPCC	Rajenda Pachauri	Chairman of the IPCC
24 July	10am	BAM3.0 20 years AMIP-type run surface flux simulation: mean, variability and their impact on the new coupled model POAMA	Guomin Wang	BMRC
1 August (Thurs)	10am	Ground Based Remote Sensing of Cirrus Clouds in the Tropical Western Pacific	Tom Ackerman	DOE's ARM
7 August	10am	Modelling of Ocean Storm Surges	Mikhail Entel	BMRC
12 August (Mon)	10am	The Synoptic Weather Patterns that Cause Major Flooding in Dryland Australian Rivers	Graham Grootemat	University of Wollongong
21 August	10am	Using MGA to search the meteorological literature	Andrew Hollis	National Meteorological Library
27 August (Tues)	10am	Spurious Power-law Relations among Rainfall and Radar Parameters	A.R. Jameson	RJH Scientific Inc., USA
28 August	10am	Using the high-resolution LAPS model to predict fog at Sydney and Perth airports	Xinmei Huang	BMRC
4 September	10am	Towards objective verification of NWP model cool-change timing	Graham Mills	BMRC
12 September (Thurs)	10:30am	Aviation turbulence research activities in the U.S.	Robert Sharman	NCAR, Boulder USA
13 September (Fri)	10 am	Numerical modelling of thunderstorm-generated turbulence	Todd Lane	NCAR, Boulder USA
18 September	10 am	Operational Ocean Prediction	Neville Smith	BMRC
25 September	10 am	Progress in the use of ARM data for model evaluation - A tale of clouds and probabilities	Christian Jakob	BMRC
2 October	10 am	BAMS - The BAM Single column model	Greg Roff	BMRC
9 October	10 am	Air-sea feedbacks throughout the Maritime Continent	Harry Hendon	BMRC
15 October (Tues)	10 am	Fictitious pressure gradients arising from reductions to sea level over central Australia	Bob Seaman	BMRC
23 October	10 am	An automated synoptic typing system using archived and real-time NWP model output	Robert Dahni	BoM COSB
30 October	10 am	How well is the Australian climate simulated by current AGCMs: Results from AMIP2 experiments	Huquiang Zhang	BMRC
6 November	10 am	Seasonal predicatbility as defined through large-scale coherence: An Indonesian case-study	John McBride	BMRC
11-13 November	No Seminar: BMRC Modelling Workshop, Melbourne
27 November	10 am	Comparison of Integrated Water Vapour Estimates from GPS, Radiosondes and NWP Models over Australia for Year 2000	Tomasz Glowacki	BMRC
3 December (Tues)	10 am	Simulated impact of greenhouse warming on hurricane intensities using the GFDL Hurricane Model	Tom Knutson	GFDL, USA
4 December	10 am	Evidence of large-scale, free Rossby waves	Rol Madden	NCAR, USA
11 December	2:30 pm	A Highly Resolved Regional Climate Model Developed at the IPRC	Yuqing Wang	IPRC, U. Hawaii
16 December (Mon)	10am	Recent Results from the next ECMWF Reanalysis ERA-40 -- Observing System Variability and Tropical Cyclones	Mike Fiorino	PCMDI, U.S.A.
18 December	10am	Simulations of global climate and vegetation change in the Hadley Centre GCM	Richard Betts	UKMO

The normal venue is the BMRC conference room (Floor 13).

`Traditionally,' seminars are at 10 am on Wednesdays with duration of 30 to 50 minutes + questions. Dates and times other than the usual Wednesday morning are shown in bold print.

Emphasis is on work in progress. Partly because of this, the schedule is susceptible to change.
To receive this information via email, simply send an email to majordomo@bom.gov.au containing the single line in the body of the message: subscribe bmrc_seminars

Videotapes of several seminars that have been given during the year are available for loan from the National Meteorological Library. These are indicated by a camera icon next to the seminar date. In addition, a list of seminars held in the library can be found on the catalogue by entering Series: BMRC, Format: Video. If you would like to have a talk videotaped please contact the seminar coordinator.

If you would like to know more details of coordinating seminars (if, for example, you are hosting a visitor who will be giving a seminar and the regular seminar coordinator is not available), have a look at the document, "Instructions for BMRC Seminar Coordinator"

For further details contact the seminar coordinator, Matthew Wheeler, on 03 9669 4068, m.wheeler@bom.gov.au

ABSTRACTS

An experiment involving Meteorologists in various offices of the BoM has been in progress for about two years. They have been logging their percentage confidence, at various lead times, that the weather at each airport will be below the alternate minimum. Results are shown by way of relative operating characteristics and reliability diagrams. It is shown that individual forecasters adopt varying decision thresholds, and that the difference in decision thresholds between individuals is more significant than differences in skill.

It is shown that forecasts at lead times less than around 3 hours fail to match persistence, so airlines would be better off using present weather for short domestic flights. An argument is mounted that providing TAFs as a probability of breaching the instrument approach minimum would provide a cost outcome superior to the current categorical system. An example for a particular flight will be presented. Furthermore it will be shown that even the moderately reliable TAFs as currently produced would, if given in probabilities, produce almost the same savings. This would be achieved without any increase in skill.

I will begin by briefly revisiting the earlier modelling work, and discussing the dynamics that lead to supergradient flow, as well as some of the other interesting phenomena predicted by the theory. These will be compared with analyses of data taken in several Australian and North Atlantic tropical cyclones. The techniques used in preparing the analyses will also be discussed. Storms to be discussed include Severe Tropical Cyclone Vance and Hurricanes Andrew, Bonnie, Danielle, Georges and Mitch. Data sources include conventional, wind profiling radar, reconnaissance aircraft and GPS dropsonde.

B Overall, the analyses will be seen to be in good agreement with the analytical and numerical predictions. This implies that some of the operationally used techniques for forecasting tropical cyclone winds, storm-surge and wind-waves, as well as for climatological "worst-case" damage estimation, may require modification.

In particular, the effect of the altimeter sampling pattern is investigated. This is achieved by first calculating "error correlations" from modelled wave fields, using the model climatology as the background field. For the purposes of this work, these are considered to be the "true" error correlations. The modelled wave fields are then sampled along simulated altimeter ground tracks, error correlations are re-calculated from this simulated altimeter data and the results are compared to the "true" error correlations.

It is found that there is considerable regional variability in the "true" error correlations. Although the altimeter sampling pattern generally underpredicts the correlation length scale, the regional variability is captured well by the simulated altimeter data. It is also found that over much of the globe, the "true" error correlations are considerably anisotropic. The ability of the simulated altimeter data to capture this anisotropy will be discussed.

Research activities at the MRI cover not only meteorological and climatological areas but also oceanographic, seismological, and geo-chemical ones. Brief descri ption will be made on these activities.

We have a super-computer system of HITACHI SR8000 with 36 nodes each of which can operates at 8GFLOPS at the most. Observation systems of the MRI include a meteorological observation tower of 213 m height, a system of a turntable and circular wind tunnel, a meteorological radar, a Doppler weather radar, a portable Doppler radar, a wind profiler, a lidar, etc. Besides the main building, there are several facilities for experiments.

The climate research department consists of 5 "laboratories." Their activities w ill be presented. Modelling research ranges from that for the long-range climate change to that for seasonal to inter-annual variations. Since our understanding in the indirect effect of the aerosol in relation to clouds is still quite uncertain, observation study in this area is also being undertaken.

Then, several ongoing major research activities in Japan on climate modeling, those for global warming projection in particular, will be presented.

1) Adaptive observation network design. (Bishop et al., 2001, MWR)
2) Pre-Emptive forecasts that use recent data to identify which linear combination of forecasts best fit recently obtained data.
3) Ensemble generation
4) Fixed observation network design (Bishop et al., 2002, JAS), and
5) Data assimilation (Etherton and Bishop, 2002, MWR).

The adaptive observation network design technique discussed in Bishop et al., (2001) is now used operationally by NCEP for their annual Winter Storms Reconnaissance program in which regular weather reconnaissance aircraft flights are made over the Pacific in order to reduce errors in forecasts of winter storms. Analysis of previous experiments shows that Bishop et al.'s technique can be used to make quantitatively reliable predictions of the reduction in forecast error variance due to supplemental observations.

The present study should be described as "work in progress". It suggests the presence of multi-decadal tidal components in the Southern Oscillation and the Inter-decadal Pacific Oscillation, and such components could be useful for decadal-scale forecasting.

Hypothetical sub-decadal (1 to 10 year) climate cycles are generated from these multi-decadal cycles. Evidence for the presence of these faster cycles in climate oscillations and in sea surface temperatures (SSTs) will be given. A phase approach to seasonal forecasting will be used to suggest seasonal tidal patterns in SSTs and rainfall.

Phases III and IV of this Project are now commencing using the same strategy to produce MPI versions of semi-Lagrangian dynamics and model physics, leading to full distributed memory version of the operational GASP model.

The methodology is tested on synthetic data, using Monte Carlo simulations, and by using the extratropical wintertime 500hPa geopotential height taken from the NCEP-NCAR reanalysis.

Results from the NCEP-NCAR reanalysis show that, for both hemispheres, the patterns of the potentially predictable component have features characteristic of many well-known teleconnection patterns. For example, in the Northern Hemisphere, these patterns have qualitative features of the North Atlantic Oscillation/Arctic Oscillation, Pacific North American, Western Pacific, East Atlantic, Tropical Northern Hemisphere, Eastern Atlantic/Western Russia, Scandinavian and Eurasian patterns. Similarly, in the Southern Hemisphere, these patterns have features qualitatively similar to the High Latitude Mode, Pacific South American, Wavenumber 3, Tropical Southern Hemisphere and Meridional Wavetrain patterns seen in other studies. The role of SST-forcing is also briefly discussed.

Several model simulations are available from the GRIPS study (BMRC, CNRM, CMAM, FUB, MPI, MRI, NASA, UKMO). The model intercomparison enables us to obtain an indication of the robustness of the simulations as well as the dependence of the results on model features. The comparison with observations uses large scale analysed datasets (UKMO and NMC) to validate the simulated character of the circulation throughout the vortex lifecycles.

I will also attempt to explain the interlinking role that diagnostic sub-projects must play in any successful MIP and the role that "sensitivity experiments" can play within a standardised experimental framework.

I will argue that in some cases the experimental design can be seriously flawed and will illustrate my concerns using two attempts at exploring model-model differences; resolution changes and land surface changes.

About 90% of fogs under clear skies are detected by NOAA16, and about 20% of fogs are obscured by higher cloud.

Daytime fog and low cloud detection is under development, taking advantage of synergy with the satellite-based solar exposure program. Additionally, a surface-based passive atmospheric infrared sounder should provide new data for fog research and forecasting.

An ocean assimilation that uses optimal interpolation methods is currently used to initialize the ocean model for coupled forecasts, with focus on the tropical Pacific. However, Ensemble Kalman filter methods are in test phase. Even though one of the strongest responses to the equatorial El Niņo signal lies over the continental U.S., predictability studies indicate that the key to summertime precipitation forecasts over the continental U.S. lies in the initialization of soil moisture. Hence, NSIPP places high emphasis on modeling land surface hydrology.

The presentation will highlight recent developments in our assimilation and modeling activities and our forecast system.

The nowcasting systems made skillful predictions of rain in the Sydney region for up to 3 hours. Almost all of the schemes consistently outperformed persistence, but had difficulty outperforming extrapolation. The ability of the algorithms to accurately predict both the location and intensity of the rain falls off quickly with time. Light rain was easier to forecast than heavier rain, and greater skill was shown for established convective systems. The locations of gust fronts and convergence lines were well forecast by one algorithm.

The aircraft has grown from a prototype to the sophisticated Mark 3 aircraft with advanced composites (and a dash of balsa wood); a small, fuel injected engine that carries the aircraft for up well over a thousand miles on a gallon of fuel; and a robust flight system that makes all flight decisions. The aircraft are backed up by a flight system that now routinely enables com mand and customer viewing to be exercised anywhere on earth that has reasonable communications.

Whilst we have been developing the aircraft, advances in instrumentation have lead to almost every desirable system moving down in size, weight and cost to enable Aerosonde deployment. In addition to the standard Vaisala meteorological instruments, we are now flying video and still cameras (vis and near IR), infrared surface temperature, icing sensors and electronic warfare systems. Funded instrumentation includes: synthetic aperture radar, laser range finder, cloud physics and atmospheric chemistry.

I will describe the current system and select some recent missions to show its capacity. I will also bring along an operational aircraft for show and tell.

Another attractive feature of the model is the use of reversible staggering for the winds, available because of the cyclic nature of the grid. Superior dispersion properties result from the reversible-staggering treatment.

Some results will be shown from simulations being undertaken for the AMIP atmospheric GCM intercomparison. The model is also being used in stretched-grid mode, using the Schmidt (1977) transformation. Examples will be provided of its application to regional climate modelling for various domains, and to numerical weather prediction for a variety of resolutions.

References

McGregor, J.L., 1996: Semi-Lagrangian advection on conformal-cubic grids. Mon. Wea. Rev., 124, 1311-1322.

Rancic, M., R.J. Purser, and F. Mesinger, 1996: A global shallow-water model using an expanded spherical cube: Gnomonic versus conformal coordinates. Quart. J. Roy. Meteor. Soc., 122, 959-982.

Schmidt, F., 1977: Variable fine mesh in spectral global model. Beitr. Phys. Atmos., 50, 211-217.

Over the last five years, the Atmospheric Observation Panel for Climate has developed a strategy to provide baseline systems to complement the comprehensive observing system, which serves the data assimilation and modelling community. The baseline systems aim to provide high-quality data to calibrate the more dense national and regional networks and to support the monitoring of basic global climate indicators.

The domestic and international operating environment for NMSs is also being influence d by several other factors. The rapid progress in the science and technology underpin ning meteorological service provision is likely to continue. There is an increasingly sophisticated public and private sector clientele who is continuing to demand better quality and more services with improved level of accuracy. There is a growing demand for more rigorous and more broadly based demonstration of the economic value of mete orological services. There is a growing involvement of financial markets in relation to the meteorological services through products such as weather derivatives. With inc reasing exposure to market forces, there is the growing challenge of identifying appr opriate policies and mechanisms for funding and pricing of meteorological services. T he increasing exposure to market forces of domestic meteorological service provision is affecting the continuing recognition of the need for free and unrestricted access to meteorological data and products provided by NMSs across the world.

Given this changing operating environment, there is a growing need for a more widely understood economic framework through which to assess the benefits and the costs of m eteorological service provision at the national level. The purpose of this seminar is to discuss some of the features of such a framework.

The word Perlan is an Icelandic word meaning pearl. Perlan is the name given to this project and is inspired by mother- of-pearl or nacreous clouds occasionally seen at high altitudes and high latitudes. The mother- of-pearl or Polar Stratospheric Clouds (PSCs) are present in the northern hemisphere when wave outbreaks are ongoing.

Strong stratospheric mountain waves have been identified in data from Sweden and the south island of New Zealand. In Sweden the northern mountains easily perturbed the low level flow over the mountains generating tropospheric waves with the smaller northern hemisphere Polar Vortex residing over this region at high altitudes. New Zealand is favorable, even though its located at lower latitudes because of the great size of the southern hemisphere Polar Vortex which can extend into the lower latitudes. It is known that these waves propagate into the middle and upper stratosphere when the outer region of the polar vortex lies above a strong tropospheric wind band, above mountainous terrain. Analysis shows that these waves will produce vertical wind speeds that will lift a specially designed sailplane potentially to 30 km (100,000 feet). The atmospheric conditions and their favored locations and seasons will be discussed along with the project details.

There are two phases of the Perlan Project (web site: http://www.weatherextreme.com/perlan). Phase one uses a modified production Glaser-Dirks Flugzeugbau GmbH DG505M sailplane to reach 62,000 feet (19 km) to demonstrate the feasibility of the project. This phase requires the use of pressure suits in an unpressurized cabin. This phase consists of flights that took place in the Sierra Nevada mountains of California from January April of 2002, and world record flight attempts that will take place in July and August of 2002 in New Zealand. Phase 2 of the project is to soar to 100,000 feet which will require a highly specialized sailplane with a pressurized cockpit designed for very high altitude atmospheric research.

At the start of a new century, are we climate-proofing ourselves and our societies or not? Clearly, the vigorous greenhouse debate that has been occurring over the past two decades, about possible extent and likely impacts of global warming, indicates that, in spite of modern technology and weather and climate preparedness, societies are not immune from the vagaries of the climate and the extremes of the weather.

There are those who claim that the value of the weather changes as the economy changes, and that weather as a resource, perhaps even more than weather as a hazard, is likely to vary in value and importance to those exposed to it. Some of the changing relationships through time show up in:

(a) Social trends, eg, areas with cloudless skies, in spite of the hazard of exposure to the sun, have become popular for tourism and settlement;

(b) Weather disaster statistics: impact of some events varies, and one can ask: how would the same weathers effects in the same place have been different at a different point in history;

(c) Changing climate impacts on food supply, where climate, transport and technology combine to alter profitability;

(d) Changing relationships between climate, extreme events and mortality.

Do extreme climatic (weather) events change the course of history, and if so, by how much? Hurricane Mitch did damage roughly equivalent to the GNP of the small country of Nicaragua. How much will it change the course of that nations history? Historians tell us this question is difficult to answer, because the impact depends on many other factors. Does the impact of Cyclone Tracy on Darwin provide a useful case study to elaborate on such questions?

The above findings may provide insight into why the newly developed coupled seasonal forecast model (where the same BAM model is coupled to ACOM2 ocean model) has exhibited high skill while there is significant model drift in SST. Further diagnostics from the forced ACOM2 stand-alone experiments have shown that the BAM model heat flux bias is likely the major contributor to the coupled model SST drift.

Each flood event (about 600) has been classified for each catchment selected and work to describe and compare the synoptic weather patterns responsible for flooding in dryland Australia has commenced. The results of this work so far will be presented and it is hoped that this will promote feedback on the classification system adopted.

In this talk I will describe my progress in developing such an objective system. Time series of hourly model forecasts of 10m wind, and screen-level temperature and dewpoint were extracted from each meso-LAPS forecast through the summer of 2001-2002 at station locations through South Australia and Victoria, and equivalent hourly time series of observations from the AWS at those stations. Criteria to define a "wind change", based on changes in wind direction, temperature rate of change, and dewpoint change were defined based on these data, with some physical basis. Matchups between wind changes identified from the observation and from the model time series have been made, and forecast timing errors documented, for 8 South Australian stations.

The issue of threshold selection, what constitutes a wind change, and model performance will be addressed, and I'll discuss what are the next steps to take with this project.

Two and three-dimensional cloud-resolving model simulations, with their configurations motivated by the turbulence encounter, will be presented. These calculations explicitly resolve both the convection and the turbulence-causing instabilities in the vicinity of the cloud. In particular, gravity waves are seen to break near the cloud top, causing vertical mixing and turbulence. In addition to this gravity wave breakdown, smaller-scale Kelvin-Helmholtz-like waves form on the cloud interface. A discussion of these out-of-cloud turbulence-generating mechanisms will be presented with application to aviation safety.

As with previous collaborations, the focus for model development will be in Hobart. This will allow the broad coastal and open ocean expertise of the Division to be brought into the model development, particularly with respect to coastal issues. This also aligns well with the new Division focus on Marine Environmental Prediction. The development of the analysis and assimilation system will be shared with BMRC focussing on the introduction of an enhanced (compared with the present coupled model) data assimilation system. BMRC has the lead on studies of ocean predictability and on testing and validating the ocean model and assimilation system. All data and product management will be via MARS, including recording and tracking quality control. Product delivery to RAN will be via AIFS. The lead for this component, and initial operational trials of the different components, will be with NMOC. In parallel, the Bureau is attaching high priority to the development of a Strategic Plan for the Ocean Services Program with a long-term view to on-going Bureau support for the operational and research components. There are several scientific issues with perhaps the most important being the quantification of ocean predictability and ocean forecast skill. The presentation will also provide information on several other related initiatives.

BAMS provides a simple method to isolate and study physics parameterization schemes and:

-can complete a 20 day run with a 1200 second timestep in 40 seconds on Gale (so is fast and cheap);

-has been organized so it can run with CCM3.6 or BAM3.1 physics packages and compare the results with observations;

-is able to display field development on the fly;

-enables the experimenter to use ``point-and-click'' methods to change initial conditions.

This seminar will provide a short introduction to BAMS and demonstrate its use in model development and further studies.

Using two years of observations over central Australia (20-27 south, 128-137 east), it is shown that the largest systematic fictitious increments to the pressure differences between station pairs are between 0.7 and 0.9 hPa in most months. The geostrophic wind equivalent of these largest increments is of order 5 m/s. The increments usually make the pressure gradients too strong, but produce gradient reversals on some individual days. There is also some correspondence between the spatial pattern of the fictitious increments and the stations whose sea level pressures appear consistently higher or lower than expected from neighbours.

The synoptic types are derived using MSLP analyses over the Australian region from the BoM's archived METANAL grids and the NCEP re-analysis dataset. Principal components were first computed to represent the features of the MSLP fields. Reducing the dimensionality of the dataset, the derived variables based on these principal components were then fed to a K-means unsupervised clustering scheme to derive the synoptic types.

Real-time NWP model output extracted from the operational database is then classified using these synoptic types. Synoptic type guidance for a variety of NWP models is generated operationally and stored in the BoM's Forecast Database (under development), a comprehensive real-time database which allows forecasters to flexibly display, assess, manipulate and store all types of observational data, NWP data, and processed data (analyses and forecasts).

The real-time synoptic type guidance is accessible to new software applications used by forecasters, making the applications "context-sensitive" and allowing automatic presentation of statistical summaries of similar situations. Alerts for likely significant weather appropriate to the synoptic type may also be generated.

The BoM's synoptic typing system is unique in that no other operational system exists (to our knowledge) that incorporates both the development of synoptic types and generation of synoptic type guidance from the automatic classification of real-time NWP model output. The system continues to be under development.

In addition, a version of the BMRC Atmospheric Model (BAM) AMIP2 integration is included in the analysis and its performance is compared with the 16 AMIP2 model simulations above. Using a simple land-surface scheme with bucket-type soil model, results from the BAM simulations have demonstrated similar features as seen from other bucket-type models in simulating soil moisture variations. A detailed analysis of soil moisture memory in the model has been undertaken to study two competing processes affecting soil moisture memory in the bucket scheme: one is the lack of canopy constraints in the release of soil water and the other is the lack of canopy interception and the uniformed water holding capacity used in runoff calculations. Analysing the model results over the northern Australian and Amazonian regions shows that rapid evaporation of soil moisture is the dominant reason why soil memory is short in the bucket scheme, even though the effect of surface runoff increases in the Amazonian region.

This talk will summarize recent advances in GPS meteorology, discuss factors impacting on GPS-IWV accuracy, and compare GPS-IWV validation results obtained in BMRC with those obtained elsewhere.

In this talk, we first examine two major observing system contamination events in the pioneering global reanalyses of NCEP (NCEP R1) and ECMWF (ERA-15). These events produced step changes in the temperature reanalyses that ruined their application to trend analysis problems such as "global warming." The second and longer ECMWF reanalysis is nearing completion of its production phase and while ERA-40 has avoided the major problems of NCEP R1 and ERA-15, large observing system variability is still evident and suggests that data assimilation schemes developed for NWP in the observing system of today may not be appropriate for analyzing the low-frequency components of the general circulation. Examples of the ERA-40 observing system variability include: 1) secular changes in the annual cycle of temperature and mass of the atmosphere between periods with and without satellite radiance observations; 2) large trends in precipitable water and precipitation in the 1990s; and 3) the impact of ozone assimilation in the stratosphere. Analysis of tropical cyclones in twice daily 10-day reforecast of the ERA-40 model (T159L60) showed that the system under performed vis-a-vis expectations from operational NWP models of comparable resolution. Despite the substantial impact of the observing system on the quality of the ERA-40 in the low frequencies, these reanalyses and, the enhanced collection of observations the preceded the project, will become a new benchmark for future projects.