Figure 24. The inter-annual variability of the upper level atmosphere divided into weather noise, low frequency chaotic and external impact (SST forced) components. The highest variability related to weather noise, and to a lesser extent to low-frequency chaos, is found in the high latitudes, while for external forcing by sea surface temperatures variability is naturally highest above the tropical oceans. Taken together these analyses imply that the potential for long-range climate predictability may be largely confined to the tropics and sub-tropics.

World-wide media interest in the 1997-98 El Niño event had a profound impact on the public perception and interpretation of climate forecasts from the National Climate Centre (NCC). A study was carried out in BMRC to understand the relationship between the accuracy of the forecasts, the efficacy of the dissemination of NCC forecasts through the media, and the degree of difficulty people encountered in understanding and using the forecasts. The work, which has implications for many forecasting products involving probabilities, suggests that the dissemination process has a significant impact on the way in which forecasts are ultimately used by the public.

Research continued within BMRC on improving our understanding of tropical cyclones. These extreme weather events, which have complex structure and movement patterns, have a profound effect on tropical Australia. The role of sea spray in limiting the rate of intensification of tropical cyclones was investigated using theoretical studies, and a model of the boundary layer over the ocean was used to explain the maintenance of low-level jets that feed moisture and heat into developing cyclones (Figure 25).

Figure 25. Wind strength (m/s) in severe tropical cyclone `Vance' as modelled (top) and as observed (bottom), showing good agreement where observations were available. The model shows a jet maximum between 500 m and 2 km above sea level and about 30 km ahead of the cyclone eye.

Strategic Research

A substantial component of BMRC work continued to involve long-term sustained research and development on programs of significance to the Bureau, especially in delineating Australian weather and climate. An important part of this work focussed on the development and refinement of numerical models of global and regional weather and climate. This development work, across a broad range of spatial and temporal scales, was carried out within a unified framework so that a common infrastructure is utilised for modelling on both global and local scales.

The accuracy of predictions from numerical models depends critically on the accuracy of the initial state of the atmosphere, used as the starting point for the prediction. This dependency is due to the chaotic nature of the atmosphere, where small differences in the initial state can lead to substantial changes in the predicted state a few days ahead. In the southern hemisphere, where there are large expanses of ocean, satellite data play an important role in the specification of this initial state. The data are used to specify the vertical temperature and moisture distribution in the atmosphere. A variational technique was developed to assimilate spectrometer data from US polar-orbiting satellites into the BMRC global weather predictionmodel. It was found that the variational technique optimises the impact of the data within the model, leading to significant improvements in prediction accuracy in the Australian region (Figure 26).

Figure 26. Impact of the assimilation of satellite spectrometer data on numerical weather prediction in the Australian region, showing the reduction in the root mean square (RMS) error in predicted Mean Sea Level Pressure (hPa) particularly for forecast periods of 72-144 hours (3-6 days).

There was a continued program of research in BMRC aimed at reducing the uncertainties in the projections of climate change due to the enhanced greenhouse effect. Such projections are based on the global models used for weather and climate prediction in BMRC. Complex feedback processes within the climate system impact on the net response of the climate system to any externally forced changes, such as a change in the concentration of carbon dioxide in the atmosphere. In order to improve the performance of global climate models, diagnostic studies were conducted to identify and understand the important feedback processes. A recent study on the feedback between radiative transfer and cloud processes has evaluated for the first time the geographic distribution of contributions to global feedbacks, demonstrating that there are significant geographical variations (Figure 27). This result provides important new information on the processes which determine feedback strength, such as the relative roles of moist and dry areas for water vapor feedback.

Figure 27. Geographic distribution of changes in long wave radiation at the top of the atmosphere (W/m²/K) contributing to global water vapor feedback under a doubling of carbon dioxide.

Because climate change projections extend for decades or even centuries ahead, the calculations are necessarily made with models at relatively low spatial resolution. Work continued on relating the broad-scale results from the climate models to regional and local weather parameters through downscaling techniques. Earlier work in BMRC had verified the value of an analogue statistical downscaling technique for estimating extreme surfacetemperatures across Australia. The technique was recently applied to output from three global climate models simulating the climate change at a time when the concentration of carbon dioxide has doubled. It was shown that the statistical technique reduces systematic biases in model output, as well as producing local estimates of weather variables. The model studies showed that the downscaled estimates of extreme temperatures from the different models are found to be much more consistent than the estimates directly from the model output (Figure 28).

Figure 28. Improved consistency in regional scale projections of maximum temperature under the scenario of doubled atmospheric carbon dioxide. Projected increases in summertime maximum temperatures for the Murray Darling Basin from three global climate models (BMRC, CSIRO and LMD (France) ) show greater consistency using a statistical downscaling technique (right) than using the broad scale model output (left).

There is world-wide concern about the potential for enhanced greenhouse warming to lead to changes in the occurrence of extreme climate events. In order to determine whether such changes in extreme events have been occurring in recent decades, BMRC led a study to analyse trends in extremes in temperature and rainfall across the Asia Pacific region. Participants from 15 countries took part in a workshop in December hosted by BMRC in which daily rainfall and temperature data were quality-controlled and analysed in a consistent manner. One finding of the study is that there have been significant increases in the annual number of hot days and warm nights across the whole region (Figure 29).

Figure 29. Trends in the annual number of hot days and warm nights over the last 30 years across the Asia-Pacific region, showing some significant increases. Hot days (warm nights) are defined as events where the daily maximum (minimum) temperature exceeds the 1961-1990 mean 99^th percentile.

The oceans are widely recognised as both drivers of climate change and a reflection of changes which have occurred. Monitoring the oceans in order to understand and underpin weather and climate prediction, as well as to identify signs of climate change in the ocean, isincreasingly acknowledged as an important activity. In collaboration with CSIRO Marine Research, BMRC operates the Joint Australian Facility for Ocean Observing Systems (JAFOOS) which focuses on enhancing, analysing and using ocean measurements in our region. A major focus for the Facility in 1999 was contribution to the First International Conference on Ocean Observing Systems for Climate, including the preparation of a book based on the Conference proceedings.

As well as investigating the concept of predictability, research in BMRC included investigations of possible sources of predictability of Australian climate. The impact of the Pacific Ocean on Australian climate variability has long been recognised, but there has been increasing analysis in the last decade of the role of the Indian Ocean. Recent analysis showed, however, that the apparent influence of the sea-surface temperature (SST) patterns in the Indian Ocean are largely linked to the El Nino phenomenon of the Pacific Ocean. There is a residual independent impact of the Indian Ocean on Australian climate, but its behaviour is complex and not yet well understood.

Applied Research

Significant progress was achieved during the year on the collaborative project with CSIRO, the Victorian Environment Protection Authority (EPA) and New South Wales EPA to demonstrate the effectiveness of a model-based air quality prediction system. The system is based on the BMRC numerical weather prediction model, run at a resolution of 5 km, with emissions inventories developed by the EPAs and CSIRO driving the CSIRO chemical transport model. The system is being run routinely over Sydney and Melbourne for a four-month period covering the Sydney Olympic Games, to support the operational air quality forecasting services of the EPAs.

The Sydney Olympics in September 2000 is also a focus for an international nowcasting (i.e. very short range forecasting) demonstration project under the auspices of the WMO World Weather Research Programme (WWRP). Trials of the demonstration project, involvingsystems from the USA, Canada and the UK as well as from BMRC, were successfully conducted in September 1999 and February 2000 in order to ensure that products from the radar-based systems could be effectively fed into the Bureau's operational procedures for the generation of short-range forecasts and warnings.

The Bureau continued to support the application of the small autonomous aircraft, the Aerosonde, to meteorological functions. A trial was held in the Victorian Regional Office to test the utility of the Aerosonde in supplementing conventional observing systems for routine weather forecasting and analysis. The trial extended over the second quarter of 2000, and the results of the exercise are currently being analysed (Figure 30).

Figure 30. Aerosonde trials over Bass Strait, showing the aircraft tracks and the winds measured along track.

Cross-cutting activities

Scientific publications

Allied with ongoing research and development directed to the support and improvement of operational systems, the improved understanding of meteorology and oceanography, climate and climate change, BMRC staff produced some 126 research publications, including refereed journal papers, articles, book chapters, conference papers and miscellaneous reports, in 1999-2000. Fifty peer-reviewed papers were published in books and international journals. Internally-reviewed BMRC Research Reports numbered six. Details of papers contributed to refereed scientific journals are given in Appendix 10, together with a list of recent BMRC reports.

Research staff were involved in the peer review of scientific investigations performed throughout the organisation, served as reviewers for work conducted elsewhere, both within Australia and overseas, and continued to oversee the publication of the AustralianMeteorological Magazine, a highly regarded southern hemisphere journal of the atmospheric, oceanic and related sciences. This quarterly journal publishes original contributions in meteorology and closely allied fields submitted by scientists from around the world. In 1999-2000, 14 separate significant papers were published in the Australian Meteorological Magazine, with nine of them authored at least in part by Bureau scientists. Several BMRC scientists also served on the editorial boards of other international scientific journals.

The annual summary report of BMRC research was published, and the pre-print volume of the 1999 BMRC Modelling Workshop (Parallel Computing in Meteorology and Oceanography) consisted of 33 separate scientific papers, 16 of which were from overseas scientists. All reports were published in good time consistent with the need for thorough review and acceptance of only the highest scientific standards.

Collaboration

The BMRC continued to contribute strongly, both nationally and internationally to maintaining high standards in meteorological and related sciences. As well as their direct contributions through published scientific research, BMRC scientists served on a range of national and international working groups and external advisory committees, concerned with both scientific and science policy issues. In 1999-2000, 68 BMRC scientists served on such bodies, many of an ongoing nature, and 15 scientists were invited to give key presentations at international conferences and workshops. Of particular relevance are international working groups on numerical experimentation into large-scale weather and climate prediction, various international oceanographic bodies and the many aspects of the World Climate Research Programme and the WMO World Weather Research Programme.

External collaboration continued to contribute to both the vitality and efficient conduct of BMRC research. Collaborations ranged from one-to-one collaboration by individuals in their particular areas of interest to major scientific undertakings such as the Japan-Australia Mesoscale Experiment and the World Weather Research Programme's ForecastDemonstration Project, for which initial experiments were conducted ahead of the Sydney 2000 Olympics. All such collaborations enable a two-way transfer of expertise and intellectual capital and add significant value, at relatively low cost, to the Bureau's own research efforts.

During 1999-2000, more than 30 scientists, both from other Australian organisations and from overseas, visited BMRC for extended periods of one week or more to consult and collaborate with Bureau colleagues.

Systems implementation

The BMRC contributed to improvements in the quality of operational guidance material through the successful implementation of nine significant systems changes. The changes covered many different aspects of the Bureau's operations from data handling and assimilation through to improvements in resolution and performance of numerical models.

The operational LAPS was implemented on the NEC SX-4 supercomputer in July 1999. Enhancements to the model included a doubling in the horizontal resolution, bringing the computational grid down to 37.5 km, an extra 10 levels in the vertical, a new land surface scheme which includes fractional vegetation and stomatal resistance, and the routine generation of boundary layer fields such as screen level temperature, moisture and winds and the boundary layer height. This latter enhancement has been made possible through the implementation of an improved boundary layer formulation which includes non-local turbulent mixing, separate roughness lengths for momentum, heat and moisture with orographic and vegetation effects and substantially enhanced vertical resolution in the boundary layer.

The tropical version of the limited-area system (TLAPS) was implemented on the SX-4 in
September 1999, with increased resolution and utilising detailed analysis of GMS satellite data to better define cloud and moisture in the initialisation process.

A new version of the MESOLAPS model was implemented in November, with a horizontal resolution of 12.5 km over the entire Australian domain and a special 5 km resolution version specifically for the Sydney region to assist weather forecasters provide services during the Sydney 2000 Olympic Games.

A thunderstorm guidance system using output from MESOLAPS and based on a decision-tree approach was implemented operationally in NMOC. In addition to generating a pictorial representation of the severity of thunderstorms, it also generates a number of diagnostics relevant to severe weather forecasting.

A tropical cyclone version of the limited area prediction system (TC-LAPS) was implemented in December 1999. The performance of TC-LAPS was excellent over the summer season with the tracks of several tropical cyclones well-predicted.

A technique for assimilating satellite wave height data into the starting analysis of the wave model was incorporated into the operational system in August 1999 and is expected to lead to improvements in the forecasts of waves and swell.

An intermediate coupled atmosphere-ocean model, which depicts how the ocean and the atmosphere behave and interact and which is used to forecast sea surface temperatures, was implemented operationally this year. Such forecasts will be useful in warning of an El Niño event several months ahead.

A significant development was the TITAN (thunderstorm identification, tracking, analysis and nowcasting) system, which takes 3-D digitised reflectivity output from Bureau weather watch radars and performs automatic storm identification, tracking and forecasting thirty minutes to one hour ahead.

A system which combines the output from the LAPS model with surface observations to produce a mesoscale analysis of the atmosphere was implemented within the Australian Integrated Forecast System (AIFS). The system will be particularly useful in areas where theobservation network is reasonably dense, such as capital cities, and is one of the many forecast guidance tools available to the Bureau's support services for the Sydney 2000 Olympic Games.

Cooperative Research Centres

During 1999-2000, the Bureau continued its active involvement with the three Cooperative Research Centres (CRCs) in which it is a partner.

The Meteorology CRC (CRC for Southern Hemisphere Meteorology) program on climate dynamics provided a focus for strategic research on the mechanisms controlling large-scale atmospheric behaviour in the southern hemisphere. This research complemented the continuing work in the Bureau of Meteorology Research Centre (BMRC) on climate analysis and prediction in our region. There was particular progress in the development of techniques to analyse predictability and variability of climate. That research involved collaboration with CSIRO Atmospheric Research (CAR), and so it supported one of the many links between CAR and BMRC.

The Bureau has had a strong interest in the program on stratospheric ozone in the Meteorology CRC. The collaborative research of the CRC complemented the Bureau's operational ozone measurement program, which contributes to the WMO Global Atmosphere Watch. The CRC initiated a valuable project aimed at collecting ozone-sonde data from Macquarie Island, which is uniquely located for monitoring the mid-latitude variations in stratospheric ozone. The project involved collaboration with the Australian Antarctic Division as well as the Bureau so that the required operations could be maintained over a four-year period.

A significant output from the CRC has been the development of a system to diagnose and predict the level of ultraviolet (UV) radiation at the surface of the earth. The system was developed in collaboration with the Bureau, and it is now run routinely coupled to the Bureau's Global Assimilation and Prediction (GASP) system to produce a daily UV Index. The introduction of the system was carried out in conjunction with the Anti-Cancer Council of Victoria.

Through collaboration with the CRC, there have been enhancements to the capability of the BMRC global atmospheric model to simulate the variability of the stratosphere. The representation of processes such as gravity wave drag has been enhanced and the vertical resolution of the model in the stratosphere has been increased substantially, complementing continuing development of the model in BMRC.

Collaboration with the CRC also led to refinements in the parameterisation of radiative transfer and its interaction with cloud in the numerical weather prediction models in BMRC. The refinements included the introduction of a diagnostic cloud liquid water scheme as well as schemes representing cloud microphysical optics based on observational and theoretical results.

The Bureau also contributes two research meteorologists to the Antarctic CRC, who provide scientific support and professional meteorological services for major Antarctic research initiatives.

During July and August 1999 the Antarctic CRC co-ordinated a highly successful multi-disciplinary research expedition in Antarctic waters due south of Hobart. The 40-day study focussed on the oceanography, glaciology, meteorology and biology of the Mertz Glacier Polynya region, an area of persistently low concentration, thin sea ice adjacent to the Antarctic continent. Several Bureau staff were involved in the voyage. Analysis of results is continuing.

The Bureau collaborated closely with the Antarctic Division on plans for proposed inter-continental aircraft operations between Hobart and Australia's Antarctic Territory. Bureau staff are undertaking meteorological research, installing automatic weather stations at key sites and providing advice in preparation for flights. It is expected the Bureau will play a key role in this area in future years.

Erecting an automatic weather station (AWS) at Rumdoodle airstrip, behind Mawson station, Antarctica, in January. Meteorological observations from the AWS will provide vital input to research work in preparation for intercontinental aircraft operations between Hobart, Tasmania and Australia's Antarctic Territory.

Activities at the CRC for Catchment Hydrology are described under Hydrological Services.

2 Includes adjustments made to initial output allocations (as published in the Portfolio Budget Statement 1999-2000) in the light of further analysis of the overall resource situation following the Budget.

a Targets published in the 1999-2000 Environment and Heritage Portfolio Budget Statement (PBS).

b Cost re-balancing across the individual outputs post-budget has led to cost increases and decreases within the Major Output price. An overall price increase of approximately $3m is due to the combined effect of the transfer of trust accounts into the accrual framework and changes to the formula for calculating operating lease rentals (OLR) which led to a re-balancing of OLR across all Major Outputs.

Previous Chapter | Back to Index | Next Chapter