REPORT BY THE DIRECTOR OF METEOROLOGY ON
THE BUREAU OF METEOROLOGY'S FORECASTING
AND WARNING PERFORMANCE FOR THE
SYDNEY HAILSTORM OF 14 APRIL 1999
FIGURES
Figure 1. Map of the storm area, showing place names, Local Government Areas (LGAs) and a schematic depiction of the radar-indicated track of the storm (See Figure 9 for a more detailed storm track.)
Figure 2. Average monthly distribution of severe thunderstorms in New South Wales.
Figure 3. Average hourly distribution of severe thunderstorms in New South Wales.
Figure 4. The structure of the forecasting and warning components of the Sydney Regional Forecasting Centre and the Sydney Airport Meteorological Unit.
Figure 5. Mean Sea Level Pressure Chart for 4 p.m. EST 14 April 1999.
Figure 6. Wind pattern over southeast Australia at the 500 hPa level (approx. 5.5 km) for 4 p.m. EST 14 April 1999.
Figure 7. Infra-red satellite images of the central N.S.W. coast region from 5.30 p.m. EST 14 April 1999 to 12.30 a.m. EST 15 April 1999.
Figure 8. Images from the Sydney radar (located at Letterbox, just north of Wollongong) from 4.15 p.m. EST to 9.55 p.m. EST 14 April 1999.
Figure 9. The radar-derived track and severity classification of the intense core of the thunderstorm.
Figure 10. Staffing arrangements in the Sydney Regional Forecasting Centre between 3 p.m. and midnight on 14 April 1999.
Figure 11. Variations in the total number of staff on duty at any given time in the Sydney Regional Forecasting Centre between 3 p.m. and midnight on 14 April 1999.
LIST OF ATTACHMENTS
REPORT BY THE DIRECTOR OF METEOROLOGY ON
THE BUREAU OF METEOROLOGY'S FORECASTING
AND WARNING PERFORMANCE FOR THE
SYDNEY HAILSTORM OF 14 APRIL 1999
The severe hailstorm which struck the eastern suburbs of Sydney on 14 April 1999 caused extensive damage which has been estimated of the order of $1Billion, making it possibly Australia's costliest ever natural disaster.
2 The storm was highly unusual in meteorological terms. Not only did it produce some of the largest hail ever recorded in Sydney, but it occurred at a time of year when severe thunderstorms are normally rare and at a time of day when the probability of storms developing, or existing storms maintaining their intensity, is low.
3 The meteorological indicators (observations, synoptic conditions, climatological record) did not exhibit the combination of background conditions which would normally be expected to lead to development of severe thunderstorms.
4 The few showers which developed on the ranges to the southwest of Sydney during the afternoon of 14 April were relatively benign, adding weight to the evidence that the environment ahead of a weak front moving up the south coast of NSW was not particularly conducive to the development of thunderstorms.
5 About 4.25 pm a thunderstorm began to form about 115 km to the south-southwest of Sydney, over land to the north of Nowra. Post analysis has shown that this storm subsequently developed into what is known as a "supercell", a rare but unusually severe type of thunderstorm whose structure, behaviour, intensity, and longevity is quite different to ordinary thunderstorms.
6 By around 6 pm the storm had moved up the NSW coastline with the western edge of the storm reportedly producing substantial amounts of hail in the Wollongong area (but no reports of giant hail). The storm then moved out to sea but continued to track north-northeast, parallel to the coast before crossing the coast again near Helensburgh at about 7pm and continuing along the coast to devastate Sydney's eastern suburbs from about 7.40pm. The storm eventually moved out to sea again east of Gosford about 9.30pm and had dissipated by 10pm.
7 The upper winds which, according to established meteorological theory and experience, would have been expected to steer any thunderstorms along the coastal ranges to the south of Sydney, were blowing towards the northeast (i.e. southwesterly winds). Radar observation of the few showers and storms that developed during the mid to late afternoon south of Wollongong confirmed that they were all moving northeastward out to sea, as would be expected with the southwesterly steering winds.
8 During the afternoon of 14 April 1999 the day Shift Supervisor in the Sydney Regional Forecasting Centre (RFC) made use of the Bureau's weather watch radar at Letterbox, which was the long standing radar for operational use, and a newly installed (February 1999) Doppler radar located at Kurnell (a terminal for which had been installed in the RFC for staff familiarisation, but was awaiting formal commissioning for routine operational use pending completion of training of all RFC forecasting staff). Both radars covered the storm.
9 On the basis of all the information available to him, and after consultation with others in the RFC, the day Shift Supervisor reached the judgement that, although there were a few thunderstorms developing over the coast and ranges to the south of Wollongong during the afternoon, they would continue to move northeastward out to sea, and there was only a low probability of thunderstorms directly affecting Sydney. He did not envisage that any thunderstorms that might occur would produce severe weather.
10 At around 6.35 pm to 6.40 pm, the RFC day Shift Supervisor started to hand over to his night Shift relief who normally occupies a professional forecaster position below that of Shift Supervisor. This night Shift Supervisor had almost 30 years' general forecasting experience and was competent to act in the position. He arrived at the RFC 20-25 minutes earlier than the scheduled commencement of his shift. He was briefed on the forecast policy set by the day Shift Supervisor including specific reference to the radar image of the storm near Wollongong which was assessed to be moving northeastward out to sea.
11 After approximately 15 minutes of handover briefing, a phone call from Wollongong reporting the occurrence of hail with the thunderstorm which was being tracked on the radar was received. Although the night Shift Supervisor accepted and concurred with his predecessor's forecast policy, including the judgement that there was no significant likelihood of severe weather in Sydney, he was concerned that some (non severe) thunderstorm activity associated with the western edge of the storm could affect the eastern edge of the Sydney Airport terminal area over the next hour or so. (The terminal area is contained within a 5 nm (8 km) radius of the airport and includes waters just offshore of eastern Sydney). He immediately phoned the Sydney Airport Meteorological Unit (SAMU) to ensure that the possibility of thunderstorms affecting the Sydney Airport terminal area was included in the Trend Type Forecast (TTF) for the Airport.
12 The SAMU forecaster had, in fact, already considered including thunderstorms on the forecasts for the Airport and had advised various Airport authorities and airline representatives of the possibility of some thunderstorm activity affecting the vicinity of the Airport by around 8 pm. At about 7.20 pm, after learning of reports of hail falling at Bundeena, some 16 km to the south of the Airport, he attempted to confer with the RFC Shift Supervisor but, when he was unable to make contact because of engaged phones (including a "silent line" to the Shift Supervisor's desk), he issued a Thunderstorm Warning for the Airport at 7.30 pm.
13 After also receiving phone reports of hail to the south of Sydney from various sources, including one of the Bureau's volunteer storm spotters, between about 7.30 pm and 7.45 pm, the RFC Shift Supervisor assessed the possibility that what he believed to be the still northeastward moving Wollongong storm cell might clip the eastern suburbs of Sydney and considered whether a Severe Thunderstorm Warning should be issued to the media. With only one other person in the office at this stage (a more junior aviation forecaster who had taken over the aviation shift at 7.20 pm after a 15 minute handover period), both forecasters were soon overwhelmed by phone calls from media and other callers, either reporting hail or seeking explanation of why there had been no warning. While he reached the view that it was now too late to provide useful warning for the city and eastern suburbs, the Shift Supervisor also considered whether a warning should be issued for the northern suburbs but concluded that the storm was probably already in a decaying phase and that it would be unlikely to be still producing large hail by the time it had crossed the Harbour.
14 Professional judgements taken on available information in the RFC during the afternoon and evening were reasonable, but with the benefit of hindsight, and on the basis of post analysis, the following conclusions have been reached:
15 As a result of these judgements, the Bureau's RFC did not provide any warning of the severe hailstorm. The SAMU provided some advance forecasts, warnings and advice to the aviation community.
16 Although the various decisions taken in the RFC during the afternoon and evening were individually consistent with the best professional scientific and operational judgements of the forecasting staff involved at the time, their combination resulted in the Bureau not providing a public warning for Sydney on the evening of 14 April. Post-analysis of events has resulted in the implementation of a number of operational measures in the Sydney RFC, which will be followed up in RFCs in other States, to minimise the risk of a recurrence of a similar set of circumstances to the Sydney storm situation.
17 The reason a public warning was not issued for the Sydney storm of 14 April lies in the still incomplete scientific understanding of severe thunderstorm development and movement, and an inability to adequately characterise the storm using the observing systems and technology available at the time. A significant step forward will occur when the Bureau's new Doppler Radar at Kurnell becomes fully operational after the completion of staff training over the next few months. This new radar opens the way for taking advantage of a number of new diagnosis and forecasting systems for severe thunderstorms which have been developed in the USA.
18 The major conclusions from this investigation, which are necessarily based on only a short period of detailed post analysis are:
19 The following action is being taken :
20 In addition to the follow-up action outlined above, the Director of Meteorology will commission two independent evaluations of the Sydney severe weather forecasting operations
21 These two experts will advise the Director of Meteorology on the most effective use of available resources to enable the further improvement of the Bureau's severe weather warning services
Figure 1. Map of the storm area, showing the locations of places mentioned in this report, including the Local Government Areas (LGAs) covered by the N.S.W. Severe Thunderstorm Warning Service. A schematic depiction of the radar-indicated track of the storm is also shown. (See Figure 9 for a more detailed storm track.)
The purpose of this report is to provide a summary analysis of available information on the meteorology of the severe hailstorm which struck the eastern suburbs of Sydney on the evening of 14 April 1999 and to explain why no advance warning was provided to the communities in the affected suburbs.
2 This report has been developed through the preparation of a series of preliminary working papers. Immediately after the storm had passed through Sydney the Bureau's Regional Director for New South Wales commenced the preparation of an initial summary of events. This summary became available on 18 April, but its preparation being so close to the event meant that inevitably it was both incomplete and, in places, inaccurate. It provided an initial assembly of essential data for subsequent analysis.
3 On 18 April in order to consolidate the Regional Director's summary paper, a Head Office group involving the Bureau's two Deputy Directors commenced the task of establishing a more authoritative assessment of all aspects of the Bureau's performance on 14 April in Sydney. By 26 April a preliminary version of the Director of Meteorology's report had been prepared with his full involvement. Until this time the Director had not had an opportunity to visit the Sydney RFC and meet with the officers who had been on duty at the RFC and SAMU on 14 April. The Director made this visit, and interviewed all key staff involved, on 29 April.
4 Following the Director's visit to Sydney, a revised draft of the preliminary report was completed on 30 April, but there was insufficient time to consolidate and incorporate all the detailed information provided by the operational staff into this preliminary report.
5 This final report, is the culmination of the efforts of many Bureau staff in Sydney and Melbourne to document the events on 14 April. The times of some of the events referred to in the report (e.g. the time of issue of forecasts) are considered to be quite accurate having been verified by reliable documentation. The times of some other events (e.g. discussions between forecasters) having been based only on personal recollection are approximate.
6 The Bureau will conduct further research into the meteorology of the event. Scientists of
the Bureau of Meteorology Research Centre will conduct high resolution, numerical model
simulations of the thunderstorm to better understand its dynamics and the processes governing
its movement, and there will be work done to test the emerging decision support systems, which
are driven by signals provided by Doppler radars such as the one at Kurnell, to better understand
their applicability in the context of the Sydney 14 April hailstorm.
7 Weather forecasting on all time scales is an extremely complex, and still largely unsolved, scientific problem. Although great progress has been made internationally and in Australia, over the past three decades as a result of international and national meteorological research programs, and operational forecasting skill has steadily increased to the point where there is now some real utility in prediction of general weather trends out to 5-7 days, the science and technology for forecasting local small-scale severe weather systems is still in a very early stage of development. In particular, while it is usually possible to identify with some confidence, up to 6-12 hours in advance, the general regions where severe storms can be expected to break out, it is not possible to predict the occurrence or behaviour of individual storms until their development is well underway. And, when they have developed, there is, so far, only limited skill in predicting their life cycle (growth and/or decay) on time scales beyond thirty minutes to an hour.
8 The 1998 Policy Statement of the American Meteorological Society on "Weather Analysis and Forecasting" describes the state of the art in forecasting on time scales out to 0-12 hours in the US, where the relevant science and technology are most advanced, in the following terms:
"These forecasts have shown considerable skill and utility, especially for predictions
of the evolution and movement of large- and medium-sized weather systems.
However, the accuracy of the forecasts decreases rapidly as the scale of the weather
features decreases and the time range of the forecasts increases. Forecasting the
evolution and movement of smaller-scale, short-lived, often intense weather
phenomena such as tornadoes, hail storms, and flash floods is less mature than
for predictions of larger-scale weather systems. The difficulty in forecasting
small-scale systems is due to insufficient computational ability, inadequate
observational capabilities, and limited understanding of the physical processes
that are taking place during these events. General areas where these systems are
likely to form can often be predicted up to 3 days in advance but the precise location
that such a small-scale storm will form cannot usually be forecast reliably with much
lead time. However, forecasts of small-scale features have improved markedly in
regions where weather-related phenomena are generated or modulated by fixed terrain
features, land-sea contrasts and land-use characteristics.
Despite the difficulties in predicting these small-scale phenomena, the lead time of watches and warnings has increased. For example, the lead time for tornado warnings has more than doubled in the last decade due to the improved observing systems provided by the NWS (National Weather Service) operational Doppler radar network and satellite imagery. These warnings and watches rely heavily on observing and detecting when conditions are favourable for the development of severe convection and then monitoring each storm's evolution. Forecaster interpretation of radar and satellite imagery and local spotters play a critical part in these very short-term forecasts." (Emphasis added.)
9 Until the end of the 1980s, the science and technology available in Australia was insufficiently developed to support a credible specialised severe storm warning service. Although Bureau of Meteorology forecasters were able to provide a general indication of expected thunderstorm activity on a broad areal basis with useful skill, the technology and techniques for forecasting the development and movement of individual severe storms were inadequate to enable the Bureau to provide operational severe storm warnings with sufficient reliability to prove useful, in aggregate, to the community.
10 On 1 February 1990, after many years' collaborative development work with overseas research and operational agencies, and on the basis of significant improvements in technological support through its ongoing re-equipment and modernisation program, the Bureau formally introduced a Severe Thunderstorm Warning Service for Capital Cities. This involved the implementation of improved data collection and processing systems, substantial research and educational support for severe weather staff and recruitment and training of volunteer storm spotter networks.
11 Over the nine years since its establishment, the Severe Thunderstorm Warning Service has performed well, with Warnings and Advices being issued for many of the severe thunderstorms affecting capital city and other major Metropolitan areas under weather watch radar. But, without the sophistication of the technological support that exists in the US (extensive Doppler Radar networks and high-frequency, high-resolution satellite imagery), the warning system inevitably falls somewhat short of the international state of the art.
12 While promoting the informed use of the Severe Thunderstorm Warning Service
(Attachment A), the Bureau has, nevertheless, attempted to ensure that the user community
understands the limitations of the Service and does not develop higher expectations of achievable skill levels than the available science and technology can be expected to deliver. For this reason, in the June 1998 issue of its Service Charter for the Community, the Bureau addressed the following exhortations to the users of its services:
We also ask that you understand that weather forecasting is a highly complex scientific problem and that the services you receive depend on the smooth operation of an integrated national and international meteorological services system. Although steady progress is being made, occasional significant forecast errors will still occur, as a result of inadequate data or the limitations that still exist in the international state of the art in meteorological science and technology." (Emphasis added.)
13 In line with worldwide understanding in the relevant intergovernmental fora that
professionally trained National Meteorological Service forecasters can, on average and once
having passed a certain threshold of skill, deliver positive net benefit to the community through
scientifically based weather forecasts and warnings, the Bureau has continued to provide, and
gradually improve, its severe storm warning services. They are provided in the expectation, and
on the basis, that, despite their limitations, they will be useful to the community more often than
not and that, over time, net benefits will accrue to the community. There is, however, no basis for
expecting that every severe storm warning issued will be correct or that an accurate warning will
be able to be issued for every severe storm that develops. Failure to accurately predict the
behaviour of a severe weather system stems inherently from the complexity of the process and
does not, of itself, provide any conclusive information on the professional capabilities of the
individual forecasters involved.
THE SEVERE STORM WARNING SYSTEM FOR SYDNEY
14 With the launch of the Severe Thunderstorm Warning Service for Capital Cities in 1990 as a component of the upgrading of the Bureau's severe weather warning services nationwide, the decision was taken to give high priority to the service for Sydney because of its large population and relative vulnerability to severe storms.
15 The Severe Thunderstorm Warning Service for Sydney is an integral part of the Severe
Thunderstorm Warning Service for New South Wales which is, in turn, embedded in the overall
round-the-clock forecast and warning operations of the New South Wales Regional Forecasting
Centre in Sydney. Because the occurrence of severe thunderstorms in NSW is highly
concentrated in the months October to March (Figure 2), maximum possible severe storm
forecasting expertise is maintained on duty during the October to March "Severe Thunderstorm
Season". This is achieved through the establishment, during the Season, of a roster of Severe
Weather forecasters involving the four severe weather experts from the Severe Weather Section of the Regional Office.
Figure 2. Average monthly distribution of severe thunderstorms in New South Wales.
16 Because the occurrence of severe thunderstorms is strongly peaked in the mid-afternoon
(Figure 3), the Severe Weather roster is arranged to provide maximum possible expert coverage
during the hours of the day (i.e. late morning to early evening) when severe storms are most likely.
The Severe Weather forecasters must however balance the requirements of the Severe
Thunderstorm Warning Service with the needs of the other high priority services (especially
bushfire forecasts and warnings) for which they are also responsible.
Figure 3. Average hourly distrubution of severe thunderstorms in New South Wales.
17 Organisationally, the Severe Weather Section is a component of the Regional Forecasting Centre. Its staff of four severe weather experts do not however participate in the routine rosters of Public Weather and Aviation forecasters who are responsible for all of the normal public, marine and aviation forecasting services for the state of NSW and the adjacent oceans. During the times covered by the Severe Weather roster, as described above, and at other times at the discretion of the Shift Supervisor, the Severe Weather Section staff are integrated into the operations of the Regional Forecasting Centre (RFC) to concentrate on the provision of severe weather services. At other times, they perform other functions related to the operation and improvement of severe weather services, including regular liaison with major user groups, user education programs, forecaster training and applied research.
18 Even when Severe Weather forecasters are on duty in the RFC, the RFC Shift Supervisor maintains overall responsibility for forecast policy for the entire State. This forecast policy is developed on an ongoing basis throughout the day as new information becomes available and provides general guidance to all forecasting staff including the specialised aviation forecasters serving the particular needs of Sydney Airport in the Sydney Airport Meteorological Unit (SAMU). Group consultation at various times during the day contributes to the development of the forecast policy. The formal organisational and operational staffing relationship between the RFC, the SAMU and the Severe Weather Section is shown schematically in Figure 4.
Figure 4. The structure of the forecasting and warning components of the Sydney Regional Forecasting Centre and the Sydney Airport Meteorological Unit.
19 One of the key roles of the RFC Shift Supervisor is to provide overall guidance and coordination of all forecasting and warning services around the clock for NSW. In the event of unexpected development of severe weather in the Region, he/she assumes responsibility for the provision of the Warning Service. The shift Supervisor is authorised to call in additional staff, including the specialised severe weather staff, if he considers extra assistance is needed.
20 The operation of the Severe Thunderstorm Warning Service is described briefly in Section 4 of Attachment A: The two components of the service in the Greater Sydney area are:
21 A severe thunderstorm is defined by the Bureau as a thunderstorm which produces either:
22 The concept of Severe Thunderstorm Advices was introduced in October 1993. Advices differ from warnings in that they (Advices) are intended to provide information on the likely development of severe storms over a broad area of the State, whereas warnings are shorter term and focus on the development and movement of individual storms.
23 The operational procedures governing provision of the Severe Thunderstorm Warning Service for Sydney are set out in the November 1998 edition of the "NSW and ACT Severe Thunderstorm Directive".
24 Severe Thunderstorm Warnings and Advices are disseminated to the community in a variety of ways. They are sent directly to television and radio stations and to emergency management authorities including the State Emergency Service (SES); they also appear on the Bureau's external Web site. The SES are alerted by telephone when a Warning has been issued.
THE METEOROLOGY OF THE STORM OF 14 APRIL
25 Many aspects of the development and movement of the Sydney hailstorm of 14 April are not fully understood at this stage and will require detailed study in a research mode over the next few years. The following paragraphs, however, provide a simple overview of the broadscale synoptic environment in which the hailstorm developed and of its subsequent movement, and intensity, post-analysed through the use of archived radar data.
26 The Mean Sea Level pressure chart for 4 pm EST 14 April (Figure 5) showed a high pressure system in the Great Australian Bight, a low pressure system just west of New Zealand and a cold front extending across the Tasman Sea to the southern NSW coast. At 6 pm the front was located just to the south of Wollongong. A thunderstorm formed in the vicinity of the front near Nowra around 4.25 pm and, by 6.00 pm, was located just off the coast near Wollongong. The upper wind flow in the region (Figure 6) was from a west to southwest direction. Such winds would generally be expected to "steer" any storms to the east and northeast. Broadscale indices of the instability of the atmosphere, while consistent with thunderstorm development, were not exceptional.
Figure 5. Mean Sea Level Pressure Chart for 4 p.m. EST 14 April 1999.
Figure 6. Wind pattern over southeast Australia at the 500 hPa level (approx. 5.5 km) for 4 p.m. EST 14 April 1999. (Wind arrows fly witht he wind and eash full feather represents 10 knots.)
Development and movement of the storm
27 The sequence of satellite images (Figure 7) and radar images (Figure 8) show the development and movement of the storm throughout most of its lifetime. In looking at these images, it is important to recognise that the radar is registering a reflected signal from raindrops and hailstones, whereas the infra-red satellite imagery is primarily registering the temperature of the tops of clouds. While the satellite images show the broad development of the thunderstorm complex in the vicinity of Sydney, the radar images are able to provide much more detail on the storm's intensity and direction of movement.
 |
 | |
| 5.30pm | 6.30pm | |
 |
 | |
| 7.30pm | 8.30pm | |
 |
 | |
| 9.30pm | 10.30pm | |
 |
 | |
| 11.30pm | 12.30pm |
Figure 7. Infra-red satellite images of the central N.S.W. cost region from 5.30 p.m. EST 14 April 1999 to 12.30 a.m. EST 15 April 1999 (The red dot denotes Sydney; the coastline extends from Bega in the south to Coffs Harbour in the North.)
Figure 8. Images from the Sydney radar (located at Letterbox, just north of Wollongong) from 4.15 p.m. EST to 9.55 p.m. EST 14 April 1999. The heaviest rain/hail areas are shown in red.
28 Figure 9 shows a detailed track of the storm derived from the sequence of individual radar images. The shaded area represents the horizontal extent of the most intense radar echoes and the small squares represent the storm centre (as measured by a weighted average of all the radar echoes within the storm). Beside each storm position is the time of the radar image, the speed at which the storm was moving at that time, and the direction in which it was travelling.
29 Three dimensional radar imagery can be used to assess storm severity, and a well-known
indicator that a storm may be severe is the presence of intense radar echoes at high altitudes. A
commonly-used criterion to classify a storm as severe (and one that is utilised operationally in the
Sydney RFC) is that radar echoes in excess of 49 dBz are observed above a height of 8 km. The
filled-in squares in Figure 9 indicate those times at which the April 14 storm may be regarded as
severe, based on this criterion.
Figure 9. The radar-derived track and severity classification of the intense core of the thunderstorm, including the location, time (EST) and estimated storm speed and direction of movement of the centre of the storm echo at 10-minute intervals.
30 At around 4.25 pm EST 14 April, the Letterbox radar first detected the storm beginning to develop to the north of Nowra. The storm initially tracked northeastwards towards the coast. At around 5.35 pm, the centre of the storm moved offshore to the north of Kiama and began accelerating towards the north-northeast. During the next half hour, the storm intensified, and by 6.05 pm was severe (based on post-analysis of the radar imagery as described above).
31 During the next fifty minutes, the bulk of the storm remained offshore and, apart from a brief time around 6.45 pm, was severe. At around 6.55 pm, the western edge of the storm intercepted the coastline again near Helensburgh, approximately 40 km south-southwest of the centre of Sydney.
32 Between 7.15 pm and 7.25 pm, the centre of the storm, which was still severe, passed just to the east of Bundeena, and during the next hour and a half tracked across coastal and eastern suburbs of Sydney. The storm weakened briefly following its crossing of Sydney Harbour, before picking up intensity again to the south of Mona Vale at around 8.35 pm.
33 At 9.00 pm, the storm reached Pittwater and subsequently tracked out to sea where it weakened.
34 The detailed chronology of the storm is given below.
| 4.25 pm | Storm begins developing approximately 115km to the SSW of Sydney over land to the north of Nowra, and begins tracking NE towards the coast. |
| 5.15 pm | Storm just west of Kiama moving NE at 26 km/hour. It has reached severe criterion with intense echoes extending above the 8 km threshold to 9 km. |
| 5.25 pm | Storm reaches the coast moving NE at 30 km/hour. Its intensity has dropped below severe criterion with intense echoes not reaching 8 km. |
| 5.35 pm | Storm moving NE at 37 km/hour. Intensity is still below severe criterion. |
| 5.45 pm | Storm now largely offshore and begins to move towards the NNE at 33 km/hour. Intensity is still below severe criterion. |
| 6.05 pm | Western edge of storm reaches Wollongong. Storm moving NNE at 33 km/hour. Intensity has reached severe criterion again with intense echoes to 8 km. |
| 6.15 pm | Storm moving parallel to coast (NNE) at 37 km/hour. Intensity meets severe criterion with intense echoes to above 8.3 km. |
| 6.25 pm | Storm moving parallel to coast (NNE) at 32 km/hour. Intensity meets severe criterion with intense echoes to above 8.3 km. |
| 6.35 pm | Storm moving parallel to coast (NNE) at 30 km/hour. Intensity meets severe criterion with intense echoes to above 9.2 km. |
| 6.45 pm | Storm weakens briefly to below severe criterion. Storm moving NNE at 33 km/hour. |
| 6.55 pm | Western flank of storm intercepts the coast again near Helensburgh. Storm moving NNE at 38 km/hour. Intensity meets severe criterion with intense echoes to above 9.1 km. |
| 7.05 pm | Western flank of storm continues to move in over the coast. Storm moving NNE at 37 km/hour. Intensity meets severe criterion with intense echoes to a height of 9.7 km. |
| 7.15 pm | Western flank of storm moves in further over the coast. Storm moving NNE at 37 km/hour. Intensity meets severe criterion with intense echoes to a height of 10.5 km. |
| 7.25 pm | Centre of storm has passed just to the east of Bundeena moving NNE at 39 km/hour. Intensity meets severe criterion with intense echoes to 10.3 km. |
| 7.35 pm | Centre of storm passes just to the west of the Kurnell Doppler radar moving N at 38 km/hour. Intensity meets severe criterion with intense echoes to a height of 8.6 km. |
| 7.45 pm | Centre of storm passing just to the east of Sydney Airport moving NNE at 44 km/hour. Intensity meets severe criterion with intense echoes to 10.9 km. |
| 7.55 pm | Storm passes over eastern suburbs of Sydney, moving NNE at 37 km/hour. Intensity meets severe criterion with intense echoes still extending to a height of 10.9 km. |
| 8.05 pm | Storm moving NE at 31 km/hour. Intensity has decreased, but still meets severe criterion with intense echoes to a height of 8.4 km. |
| 8.15 pm | Storm moving N at 39 km/hour. Intensity has dropped below severe criterion. |
| 8.25 pm | Storm moving N at 37 km/hour. Intensity is still below severe criterion. |
| 8.35 pm | Storm moving NNE at 28 km/hour. Intensity has increased markedly, with intense echoes extending to 10.1 km. |
| 8.45 pm | Storm passes just to the west of Mona Vale, moving NNE at 30 km/hour. Intensity meets severe criterion with intense echoes to a height of 8.9 km. |
| 8.55 pm | Storm is just south of Palm Beach, moving NNE at 34 km/hour. Intensity meets severe criterion with intense echoes to 9.8 km. |
| 9.05 pm | Storm begins moving partly offshore, moving NNE at 36 km/hour. Intensity meets severe criterion with intense echoes to 9.9 km. |
| 9.15 pm | Storm continues moving partly offshore and parallel to the coast (NE) at 41 km/hour. Intensity has decreased, but still meets severe criterion with intense echoes to 8.7 km. |
| 9.25 pm | Storm continues moving partly offshore and parallel to the coast (NE) at 33 km/hour southeast of Gosford. Intensity meets severe criterion with intense echoes to 9.1 km. |
| 9.35 pm | Centre of storm moving further offshore towards the NE at 32 km/hour. Intensity is still severe with intense echoes to 9.8 km. |
| 9.45 pm | Core of the storm now almost completely offshore due east of Gosford and moving NE at 32 km/hour. Intensity still severe with intense echoes to 9.6 km. |
| 9.55 pm | Storm begins collapsing rapidly and no longer meets severe criterion. |
35 The storm appears to have been a "supercell" thunderstorm, a particularly rare, severe type of system. Supercells are characterised by their longevity compared with normal thunderstorms - it appears that the 14 April storm lasted approximately 5½ hours, compared with the lifespan of a normal thunderstorm of only an hour or so. The structure of supercells is also more highly organised than other storms (refer to the brochure at Attachment A for a schematic diagram of the complex internal flow patterns of a supercell storm), and they are capable of the most extreme thunderstorm-related severe weather - large hail, high winds, flash flooding and occasionally tornadoes. Another characteristic of supercells is that they often track at an angle to the main environmental "steering" winds. The 14 April storm demonstrated this characteristic, moving towards the north-northeast for most of its life, at an angle of 40-50 degrees to the environmental flow.
36 The diameter of the hail which fell during the storm was measured to be up to 9 cm. This
is among the largest hail size ever recorded in the Sydney region.
KEY OPERATIONAL UNITS AND THEIR ACTIVITIES
37 The staffing arrangements in the Sydney Regional Forecasting Centre (RFC) and Sydney
Airport Meteorological Unit (SAMU) are summarised in Figure 10. The RFC Supervisor is a day
worker. The RFC itself is staffed around the clock by shift workers, with varying numbers of staff
on duty at different times in order to cope with varying routine workloads (Figure 11). The normal
RFC work routine is summarised in Attachment B.
Figure 10. Staffing arrangements in the Sydney Regional Forecasting Centre between 3 p.m. and midnight on 14 April 1999.
Figure 11. Variations in the total number of staff on duty at any given time in the Sydney Regional Forecasting Centre between 3 p.m. and midnight on 14 April 1999. (Top chart shows prescribed hours of duty. Bottom chart shows actual hours of duty on that day.)
38 The Sydney Airport Meteorological Unit (SAMU) is a highly specialised aviation weather forecasting unit for Sydney airport and its environs. The SAMU is co-located with the Sydney Airport Air Traffic Control Centre. The Unit's staff comprises an Officer in Charge (Senior Professional Officer Grade B) who works normal office hours, and 5 senior meteorologists (Senior Professional Officer Grade C), who work rotating shifts to provide 24 hour coverage. Apart from a 15 minute shift hand-over period, there is only one forecaster on duty at any time of the day or night. The office is well equipped with modern weather monitoring and forecasting facilities, including a radar display, and a continuous weather watch is maintained.
39 There are no support staff at the SAMU office itself, but support is provided from the Mascot Observing Office, located nearby within the airport. This office is staffed around the clock by trained observers who do half hourly surface weather observations (more frequently in defined bad weather), five upper air wind observations and two upper air temperature and moisture observations per day, and help maintain a radar weather watch. The normal SAMU work roster is summarised in Attachment C.
40 The chronology of key events at the RFC and SAMU on 14 April is given at Attachment D.
ANALYSIS OF THE BUREAU'S PERFORMANCE
41 The Bureau provides its Severe Thunderstorm Warning Service in order to provide the community with early warning of the occurrence of thunderstorms which threaten life and property. However, as indicated above, the preparation of severe thunderstorm warnings is not a clear-cut process. The most effective way to observe thunderstorms is through the use of radar, but the only way to be sure that one has occurred is by actual observation of the impacts, either by people in the affected community, including the special "storm spotter" volunteers, or by automatic observing equipment. Many such observations are not provided in real time. The current severe thunderstorm forecasting service aims to distinguish severe from non-severe thunderstorms according to criteria relating to storm impacts, but due to the limitations of current science and technology, do little to predict the degree of severity. To highlight one key difficulty, at the present time there is no way of determining the size of hail a thunderstorm will produce other than by direct observation. There are some promising techniques for estimating this from atmospheric variables, but these techniques remain under development and pseudo-operational evaluation.
42 Forecasters' decisions about whether particular thunderstorms are severe are made on the balance of evidence gained from a wide variety of sources, including conventional meteorological data, radar and reports from human observers where these are available, which they frequently are not. Since thunderstorms are relatively short-lived, these decisions must be made very quickly and there is little leeway to change them. Warning lead times are typically very short. Inevitably there are some false alarms when warnings are issued in real time for thunderstorms which are not, and do not become, severe, and some severe thunderstorms are not detected in time to give the affected community adequate warning. To judge the effectiveness of a severe thunderstorm warning service on the outcome of a single event, even one as significant as the 14 April hailstorm, is to ignore the current level of uncertainty and imprecision in the science and technology of severe thunderstorm forecasting.
43 In recognition of the difficulty of forecasting severe thunderstorms in Sydney and elsewhere, the Bureau has endeavoured to staff its specialist severe weather rosters during the times and in the regions that severe thunderstorms are most likely. During the October-March Severe Thunderstorm Season, additional staff, specially trained to recognise and forecast severe thunderstorms, assist the RFC Shift Supervisor by issuing warnings and liaising with key Bureau clients and the media during severe weather situations. In the off-season, these specialist staff are utilised to train the Bureau's forecasters in severe weather forecasting, and to develop improved techniques for operational implementation, as well as liaison with emergency management authorities and community education. A major training program was run in all Regional Offices between 1994 and 1995 specifically for Shift Supervisors and potential Shift Supervisors, to increase the overall skill levels of operational forecasting staff in severe weather situations.
44 Overall, the Bureau endeavours to maximise the proportion of severe thunderstorm events correctly forecast and to minimise the proportion of warnings which are "false alarms", by ensuring that forecasters with the highest skills and experience in severe weather are on duty during the severe thunderstorm season. For the Greater Sydney the severe thunderstorm warning service is monitored on the basis of Local Government Areas. Statistics have been kept of the number of warnings issued for the LGA's, and these warnings have been verified against observations. Table 1 summarises the results. The number of events for which warnings are issued has ranged between 3 and 12 in a season. The Probability of Detecting (POD) a severe thunderstorm has been between 30% and 50%, while the False alarm Ratio (FAR), or number of times a warning is issued and no severe storm observed, has typically been 70% to 85%.
Table 1. Summary statistics for thunderstorm warnings issued for the Sydney basin Local Government Areas, for 5 severe weather seasons. POD = Probability of Detection, FAR = False Alarm Ratio
| Year | Warned Events | POD | FAR |
| 94/95 | 6 | 0.40 | 0.68 |
| 95/96 | 12 | 0.52 | 0.75 |
| 96/97 | 3 | 0.50 | 0.88 |
| 97/98 | 10 | 0.43 | 0.84 |
| 98/99 | 5 | 0.31 | 0.71 |
45 In the context then of the historical performance of the Sydney RFC in warning for severe thunderstorm occurrence, the fact that a warning on April 14 was not issued was not in itself an unusual event, while nevertheless recognising that the storm itself was highly significant. This said, the assessment of the Bureau's performance must also consider the performance of the component systems which support the overall forecasting process.
46 The components which support the Severe Thunderstorm Warning Service comprise:
47 Radar is the primary observational tool used in thunderstorm monitoring. The imagery from two Sydney radars is available to forecasters in the NSW Regional Forecasting Centre (RFC). The Letterbox radar is located about 60 km southwest of Sydney and was commissioned in October 1982. It is a non-Doppler radar using 10cm radio waves to provide real-time information on the location and intensity of areas of rain within its range, including those associated with thunderstorms. The new Kurnell radar is located close to Sydney Airport and began operating in a running-in mode in March 1999. It is a specialised Doppler radar using 5 cm radio waves, which in addition to the conventional information, provides information on the wind field within rain areas, including the internal wind field of thunderstorms.
48 The Letterbox radar, is the primary radar for use by the Sydney forecasters. This is partly because they have considerably more training and experience in effective use of the data it provides. The Doppler wind fields provided by the Kurnell radar are extremely complex to interpret and cannot be easily used for thunderstorm forecasting without comprehensive forecaster training and operational experience. In addition, while the Kurnell radar is more sensitive and can pick up finer features in some situations, the 5 cm radio waves it uses are more likely to be partly absorbed when they pass through rain areas, and hence may produce less intense echoes from rain features beyond than would be observed on a 10cm radar.
49 Radar pictures are produced at 10 minute intervals by each radar, by a scanning procedure carried out automatically which takes 10 minutes. The date and time on each picture refers to the beginning of that time interval. An initial picture is available in the RFC about 3 minutes after the scan begins, and the full scan data gradually becomes available as the rest of the scan progresses. At the time of the 14 April storm, the configuration of the radar communications systems may have meant that there was up to a 5 minute delay before these same radar pictures were available in the SAMU; potentially, therefore, a picture labelled 10.00 pm would have been fully available in the RFC by 10.10 pm, and in the SAMU by 10.15 pm, by which time some of the information might have been up to15 minutes old. As part of the process of bringing the Kurnell radar up to full operational status the configuration of the radar communications systems has now been adjusted so that there should be essentially no difference in the time that radar data are available in the RFC and SAMU.
50 In the RFC, there are two radar display screens next to each other, both running a radar display system called 3D-RAPIC developed by the Bureau. A number of different types of display are possible on either screen. Data from weather watch radars around NSW can also be displayed. On the evening of 14 April, as is usual, the left-hand screen was used to display the latest Letterbox imagery, "merged" into a composite picture with imagery for the same time from other radars in NSW (but not Kurnell). The right-hand screen was displaying the Kurnell imagery.
51 Data from a conventional radar are viewed in two ways :
52 Storm spotters also provide valuable observational information in the forecasting process for severe thunderstorms. On 14 April, after some early storm spotter information was received from the Wollongong area, the storm approached Sydney through the Royal National Park, and so storm spotter reports which might have given more detailed advance warning of the nature of the hail were not available.
53 Apart from the paucity of storm spotter reports in the early stages, the observation systems important to severe thunderstorm forecasting functioned well on the evening of 14 April. The radars operated properly and there was no loss of conventional meteorological data or satellite data.
54 The key system for the assimilation of the data in the Bureau's RFCs and the SAMU is the Australian Integrated Forecasting System (AIFS). AIFS is capable of displaying a wide variety of data, including radar data in a simple fashion. AIFS is also used for composition and dissemination of forecasts and warnings. AIFS is supplemented by a number of special purpose computing systems including the Bureau's RAPIC system and a smaller personal computer used for displaying the geographic location of lightning strikes (the GPATS data).
55 The AIFS system and all ancillary data processing systems (e.g 3-D RAPIC, GPATS) functioned properly throughout the 14 April event. This said, they were not used in the same way on the evening of 14 April as they have been during the post-storm analysis. In particular, the use of 3-D RAPIC in the RHI mode was not a feature of the Shift Supervisor's operational routine on 14 April. The available synoptic data, and his assessment of the overall situation had indicated that such an analysis would not be justified as a priority use at this time.
56 As noted above, the Bureau has recently deployed a Doppler radar at Kurnell. This radar has the capacity to provide the fundamental input to severe weather forecasting decision systems now under development in the USA. Furthermore, the Bureau installed upgraded regional computing systems in the Sydney forecast offices in December 1997. These systems have the potential to serve as a platform for upgraded forecaster support systems.
57 With the experience of the 14 April hailstorm, and while noting that the existing systems performed well, it would seem opportune to reconsider how best to deploy the available resources to support the severe weather warning service. To do this the Director will commission an expert from the USA, who has long experience with severe thunderstorm forecasting, research and systems to evaluate the performance of the Sydney Region on 14 April and to advise on preferred future directions.
58 The Bureau relies on two types of communication system:
59 Throughout the event the data circuits performed as required. Observational data were collected and digital forecast products delivered by the Bureau's systems in a fully satisfactory fashion.
60 The Bureau uses telephones to communicate between operational facilities such as the RFC and SAMU, with external clients (SES, Media and members of the public) and between off-duty staff and operational staff.
61 The telephone numbers of the SAMU and the Shift Supervisor in the RFC are not listed in the telephone directories. All RFC numbers were changed in July 1996 when the office relocated. Nevertheless, because the Bureau provides information to a large range of clients, over time these numbers can become widely known. On 14 April it became apparent that the number of Bureau clients with access to these unlisted numbers is large and unsustainable during a severe weather event.
62 Since April 14 additional "silent" lines have been installed in both the RFC and SAMU. The numbers for these will be changed at least annually, and more frequently if required.
63 Bureau forecasting staff operate in a complex environment where there is often considerable scientific uncertainty to deal with, incomplete and possibly conflicting data, extremely tight deadlines and diverse client requirements to be met through the provision of well defined, traditional services. A high degree of professionalism is called for, along with decisiveness, good judgement and broad experience. Some events such as the 14 April hailstorm will only be experienced once, if at all, in a career of forecasting.
64 The Bureau's severe weather warning services are supported by clear directives which identify the criteria to be met in issuing advices and warnings. The Sydney RFC Severe Thunderstorm Warning Service Directive is updated annually. The Weather Services Handbook, the Aviation Services Handbook and other documents describe the operational environment in which the routine forecasting services are delivered. The routine, day-to-day forecasting operations are guided by the schedule of products to be prepared in a shift. The detailed procedures to be followed in routine operations such as shift handovers, meal break relief and preparation of many basic forecasts are not prescribed in documentation but determined by the professional judgement of the officers involved and the established operating procedures of the forecast office. This flexibility and reliance on professional judgement is necessary to permit the rapidly evolving science and technologies of meteorology from becoming burdened down by excessive or outdated process.
65 In this section the overall performance of the staff in the RFC, SAMU and those off-duty but involved in the events of 14 April is reviewed.
66 The day Shift Supervisor, through analysis of the available information, reached the conclusion that severe thunderstorms would not occur, and that any thunderstorms which might occur would be short-lived or move off the coast. By the end of the day shift in the RFC, the focus was very much on the next day, but the thunderstorm near Wollongong was noted as something to be watched (albeit it was not considered to threaten Sydney as it was confidently believed that it would move out to sea).
67 The handover between the day Shift Supervisor, a substantive Senior Professional Officer Grade C, and the night Shift Supervisor (an acting Senior Professional Officer Grade C) was scheduled for 7.00 pm to 7.30 pm. In the event, it occurred between 6.35 pm and 6.50 pm when the night Shift Supervisor arrived early and his relief subsequently departed early. The Technical Officer (Observer) advises that he left at 7.40 pm, approximately 20 minutes prior to his rostered shift completion time. He did so without consulting the Shift Supervisor. His work was finished and his potential contribution in the events which were about to unfold would have been small.
68 This said, it remains the Bureau's practice to operate the forecasting environment as a team process. Many staff arrive early by agreement with their team leaders and may also leave early and vice versa. The rule is that the work must be covered and the team leader (Shift Supervisor) must make the decisions relating to resource flexibility.
69 Two other oncoming RFC staff were slightly delayed, one by the hail storm (10 minutes) and the other by traffic (5 minutes).
70 The SAMU shift handover was scheduled to occur between 7.00 pm and 7.15 pm. It however extended over a longer period with the night shift forecaster (acting Senior Professional Officer Grade C) arriving slightly early at 6.50 pm and the day shift forecaster (Acting Senior Professional Officer Grade C) departing slightly late at 7.20 pm.
71 The SAMU staff were aware of the thunderstorm from the time of its development near Nowra, and at 6.45 pm the SAMU forecaster advised the Sydney airport owner (SACL), Qantas despatch and Ansett despatch that a lightning Alert and Airport Warning were being considered. The SAMU staff were not required to assess whether the storm was severe or not, but nevertheless acted prudently in issuing the Lightning Alert (7.16 pm) and the Airport Warning (7.30 pm) for hail and strong winds, liaising with key clients and issuing routine products such as the TTF (Trend Type Forecast).
72 The situation in which the night Shift Supervisor in the RFC found himself was extremely difficult. Until about 7.30 pm, he was not aware that he faced a severe weather situation. The Shift Supervisor had formed the view, (based on the data he had examined and the handover briefing), that a detailed analysis of the radar was not required because of the view he had formed of the storms movement and intensity. After 7.30 pm, the overwhelming number of calls convinced him of the severity of the event and he attempted to handle the telephones, which included calls from a large number of media organisations, in line with his decision not to issue a warning. By 8.00 pm he was of a view that the storm would dissipate as it crossed Sydney Harbour and so continued to deal with the telephones.
73 As the evening progressed, off duty severe weather staff, senior regional managers and the Regional Director attempted to call the RFC. The Region's Senior Professional Officer Grade C (Severe Weather) managed to contact the Shift Supervisor around 8.45 pm, offered his assistance, but the Shift Supervisor considered he could cope.
74 After discussion with the Regional Director, it was decided that the Shift Supervisor should call in additional support. He then acted to call-in two off-duty members of the Severe Weather Section to the RFC to assist. The commitment of off-duty Bureau staff to the task of providing assistance to the RFC, and services to the community, is a characteristic feature of Bureau operations in a severe weather situation.
75 In summary, it should be recognised that the Bureau's severe weather forecasting operations are in many ways unique. The atmosphere is a highly complex, 3-dimensional fluid whose evolution with time cannot be forecast precisely. This said, the severe weather forecasting operations also have many features in common with operational centres such as hospital operating theatres, emergency control centres, aircraft flight control decks, and other environments in which crucial decisions are made within immovable time deadlines based on less than complete information.
76 In order to ensure that the Bureau is able to learn from its performance on 14 April, and to build that experience into its national operations for severe weather forecasting, an expert experienced in the analysis of operational centres in times of crisis will be tasked by the Director to evaluate the overall effectiveness of systems and processes in place in the Sydney RFC, and at the SAMU on 14 April.
77 The major conclusions from this investigation, which are necessarily based on the short but intensive period of post analysis leading to the preparation of this report are:
78 The following action is being taken :
79 In addition to the follow-up action outlined above, the Director of Meteorology will commission two independent evaluations of the Sydney severe weather forecasting operations :
80 These two experts will advise the Director of Meteorology on the most effective use of available resources to enable the further improvement of the Bureau's severe weather warning services.
NSW REGIONAL OFFICE (RFC)
SHIFT SUPERVISOR SCHEDULE
Ongoing responsibilities
Specific Responsibilities - Dayshift
Local Time (EST)
| 0700 | Take over from night shift and familiarise with the current meteorological situation. |
| 0730 | Analyse 0600 EST surface chart |
| 0800 | Special fire forecasts if required. Respond to media inquiries. |
| 0815 | Formulate weather forecast policy for next four days. |
| 0930 | Lead chart discussion. Finalise forecast policy with input from staff. |
| 1000 | ABC Wagga radio broadcast. |
| 1010 | Analyse 0900 EST surface chart |
| 1030 | Warning summary, State and Metropolitan forecasts |
| 1130 | Issue four day NSW weather policy statement |
| 1200 | Analyse upper air charts |
| 1230 | ABC Country Hour Broadcast |
| 1330 | Analyse 1200 EST chart |
| 1500 | Review forecast policy prior to afternoon forecast issue. Liaise with other States to ensure cross border forecast and warning consistency. |
| 1530 | Warning summary, State and Metropolitan forecasts |
| 1630 | Sydney extended period forecasts |
| 1700 | Analyse upper air charts, Analyse 1500 EST chart |
| 1730 | Sydney rainfall probability forecast. |
| 1800 | Press issue of State and metropolitan forecasts. Update other forecasts if required. Issue high seas warnings if required. |
| 1900 | Hand-over |
| 1930 | Conclusion of shift. |
Specific Responsibilities - Nightshift
Local time (EST)
| 1900 | Take over from day shift and familiarise with the current meteorological situation. |
| 1930 | Analyse 1800 EST surface chart |
| 2000 | Analyse upper air charts |
| 2100 | Formulate weather forecast policy for next four days |
| 2200 | Analyse 2100 EST surface chart |
| 2200 | Review warnings |
| 2300 | Analyse 1200Z upper air charts |
| 0100 | Forecast policy discussion with aviation forecasters to identify critical aviation weather phenomena, especially in the early morning. |
| 0200 | Ensure all numerical model guidance has been received |
| 0300 | Four day forecast policy statement |
| 0330 | Analyse 0300 EST surface chart |
| 0430 | State and Metropolitan forecasts |
| 0445 | Coastal Waters forecasts |
| 0515 | District forecasts |
| 0515 | Town/city forecasts and town temperature/precis forecasts |
| 0530 | Town detailed and precis forecasts |
| 0545 | Newcastle town, district and coastal waters forecast |
| 0600 | Fire Danger ratings (In season) |
| 0600 | Special forecasts (eg, Special fire forecasts, Stadium Australia) |
| 0630 | High Seas warnings, if required. Variable number of radio broadcasts |
| 0700 | Hand over |
| 0730 | End of Shift |
SYDNEY AIRPORT METEOROLOGICAL UNIT
(SAMU) Services
SAMU Work Schedule (all times are in UTC and are a guide only)
| 0700 | Plot and analyse 05Z uppers |
| 0730 | Print Time Series from LAPS |
| 0800 | 0606 YSSY WEATHER BRIEFING (Copy to CENCO & Archive) |
| Complete Fog Sheet | |
| 0815 | Phone Qantas and Ansett for CODE GREY |
| 0820 | Complete Code Grey verification sheet |
| 0900 | Hand over (until 0915) |
| 0915 | Plot and analyse 08Z MSLP chart |
| 1000 | Phone SPOC to discuss TAF YSSY |
| Complete Fog Sheet | |
| 1030 | 1212 TAF YSSY, brief CENCO |
| 1045 | TAF policy in diary |
TTF's every 30 minutes (or as required with SPECI's)
Constant weather watch and retrieval of weather information from AIFS, radar, ceilometer,
GPATS, AWS, DIFACS & MDPDS.
SAMU Priorities:
Thunderstorm Procedures:
CHRONOLOGY OF EVENTS IN THE REGIONAL FORECASTING CENTRE AND SYDNEY AIRPORT METEOROLOGICAL UNIT ON 14 APRIL
An integrated chronology of activities and events at the Sydney Regional Forecasting Centre (RFC) and the Sydney Airport Meteorological Unit (SAMU) is given below. Many of the times are approximate, being based on staff recollections, although most are considered accurate to within 5 minutes.
| 4.30 am | RFC | The Sydney Metropolitan Forecast for Wednesday 14 April is issued, mentioning "Isolated showers afternoon and evening." |
| 11.25 am | RFC | Forecasts are revised and extended. The Sydney Metropolitan Forecast for the remainder of the day and Thursday is issued, mentioning "Chance of a shower this evening or overnight." |
| 2. 27 pm | SAMU | Forecaster issues Terminal Aerodrome Forecast (TAF) for 24 hours commencing 4 pm indicating a weak SE wind change is expected at about 9 pm with a few showers between 9 pm and midnight. |
| 3.42 pm | RFC | Sydney Metropolitan Forecast for Wednesday night and Thursday is issued, mentioning "Chance of a shower this evening or overnight." |
| 4.12 pm | SAMU | Sydney Airport Weather Briefing is issued essentially reconfirming the TAF issued at 2.27 pm. No reference to thunderstorms. |
| 4.25 pm | RFC | Forecaster notes thunderstorm development in vicinity of Nowra and assesses that it will move off the coast with the prevailing upper wind flow. |
| 4.25 pm | SAMU | Forecaster notes a thunderstorm has developed near Nowra moving NE at 10 knots (18 km/hr), heading out to sea. |
| 5.30 pm | RFC | The Senior Severe Weather Meteorologist is about to cease his out-of-season duties for the day (which do not include a shift work severe weather watch). He briefly talks to the day Shift Supervisor about the storm situation. At this stage the storm is not expected to be a problem for Sydney. The Senior Severe Weather Meteorologist then ceases duty. |
| 5.30 pm | SAMU | Forecaster notifies the Air Traffic Control Centre Coordinators (CENCOs) at Sydney and Melbourne Airports that thunderstorms are likely to affect air traffic around Sydney Airport but are not expected to pass directly over the airport. |
| 6.04 pm | SAMU | METAR (a routine half hourly aerodrome weather observation) reports lightning to south and cumulonimbus (thunderstorm cloud) 30 nm (54 km) to the south moving NE at 10 knots (18km/hr). Forecaster appends a TTF (Trend Type Forecast) indicating the expected trend in the weather for the next 3 hours based on the latest weather observation stating that no significant change is expected in these conditions at the airport and the current trend is expected to continue during the next 3 hours. |
| 6.15 pm | SAMU | The oncoming (evening shift) Sydney CENCO is concerned about the storms to the south that have been affecting air traffic approaching Sydney Airport. He discusses the potential storm threat with the forecaster, who is still of the opinion that the storms will not affect the airport itself. |
| 6.20 pm | RFC | Day Shift Supervisor receives phone call from State Emergency Service controller in Wollongong reporting that he has heard of hail blanketing some streets "like snow". |
| 6.34 pm | SAMU | METAR reports lightning to south, cumulonimbus cloud 27 nm to the south moving NE at 10 knots (18 km/hr). Forecaster appends TTF indicating no significant change to current trend expected for next 3 hours (ie storm still not expected to pass over the airport). |
| 6.37 pm | RFC | Day Shift Supervisor receives phone call from Storm Spotter in Wollongong reporting that he has been informed of severe hail and that he (the Storm Spotter) is confident from his own observations and his assessment of radar imagery he receives from the Bureau, that the storm is severe. |
| 6.40 pm | RFC | Night Shift Supervisor arrives early for duty just as the Storm Spotter call from Wollongong is taking place. At completion of call, day Shift Supervisor briefs night Shift Supervisor on situation, including discussion of storm and the Wollongong reports. Day Shift Supervisor advises that the storm is not expected to pose a threat to the Sydney area and that it will pass offshore. |
| 6.45 pm | SAMU | Forecaster decides that the storm is now clearly moving to the NNE instead of NE and a threat to the airport is now likely. Forecaster phones the Sydney Airport Corporation Ltd (the airport owner), Ansett Sydney Domestic Operations, and Qantas Operations Dispatcher to inform them of a possible impending thunderstorm and that a Lightning Alert and an Airport Warning are being considered. At this stage the SAMU forecaster is unaware of any storm spotter reports of the approaching storm. |
| 6.50 pm | SAMU | Qantas Operations Dispatcher phones forecaster for more information. Forecaster says it would be about one hour before the storms reach the airport. |
| 6.50 pm | SAMU | SAMU night shift forecaster arrives. Work continues. (The handover briefing is not completed until 7.20 pm.) |
| 6.50 pm | RFC | Briefing finishes and day Shift Supervisor ceases duty. There are four staff on duty in the RFC. |
| 6.50 pm | RFC | Night Shift Supervisor telephones forecaster at the SAMU to discuss whether thunderstorms are included on the Trend Type Forecast (TTF). At this stage, the night Shift Supervisor is satisfied that thunderstorms are being considered for the Sydney Airport terminal area (an area contained within 5 nm radius of the airport which includes water just offshore of eastern Sydney) but he still does not consider that they will affect more than the eastern edge of the Sydney terminal area. |
| 7.00 pm | RFC | Day shift Assistant Aviation Meteorologist ceases duty. Night Shift Supervisor assesses current forecast products and checks for latest TTF for Sydney Airport. |
| 7.00 pm | SAMU | Night shift Forecaster assesses information and determines that storm is now 15 nm (27 km) from airport and moving at 15 knots (27 km/hr). |
| 7.04 pm | SAMU | METAR reports thunderstorm and lightning approaching from the south. Night shift forecaster appends TTF predicting thunderstorms would affect the airport between 8 pm to 10 pm. |
| 7.05 pm | SAMU | Day shift forecaster speaks with Sydney CENCO, and advises Qantas and Ansett that thunderstorms are now forecast on the TTF. |
| 7.05 pm | RFC | Night shift Aviation Meteorologist arrives for duty and is briefed by the day shift Aviation Meteorologist on the situation. |
| 7.15 pm | SAMU | Night Shift Forecaster issues a Lightning Alert by phone to SACL Terminal Operations Centre, Ansett Sydney Domestic Operations, and Qantas Operations Dispatcher, advising of expected lightning to affect the airport. Confirmed in fax message at 7.16 pm. |
| 7.20 pm | RFC | Aviation shift handover briefing finishes and the day shift Aviation Meteorologist prepares to depart. Night shift Aviation Meteorologist assesses thunderstorm situation pertaining to aviation forecasts. There are four staff on duty in the RFC. |
| 7.20 pm | RFC | Sydney Metropolitan Forecast for Thursday is issued. This is a forecast for the next day and so does not mention the weather expected on Wednesday evening. |
| 7.20 pm | SAMU | Day shift forecaster completes his shift handover, finishes duty and departs. |
| 7.20 pm | SAMU | Night shift forecaster learns from Air Traffic Control (ATC) that hail has been reported at Bundeena. From about this time onwards (with some uncertainty as to exact time of commencement and cessation) a large number of phone calls are received, many of which are diverted to the answering machine. Calls are received from Qantas, Ansett, United, Sydney Tower, Air Ambulance, Helicopter pilots, Runway resurfacers, the Bureau's Mascot Observing Office, the Bureau's Regional Director, the SAMU OIC, and many others. |
| 7.30 pm | SAMU | Forecaster makes several unsuccessful attempts to contact Sydney RFC Shift Supervisor by phone to discuss the issuing of Airport Warning. |
| 7.30 pm | SAMU | Forecaster issues Airport Warning that thunderstorms with hail and wind gusts exceeding 41 knots are expected to affect the airport between 8 pm and 10 pm. |
| 7.30 pm | RFC | Night shift Aviation Meteorologist advises night Shift Supervisor of the forecast of storms on the latest TTF for Sydney Airport. Day shift Aviation meteorologist ceases duty. There are now three staff on duty in the RFC. |
| 7.30 pm | SAMU | METAR reports thunderstorm and lightning approaching from the south. |
| 7.32 pm | RFC | Night Shift Supervisor receives phone call reporting the occurrence of severe hail at Bundeena. The call takes approximately 5 minutes. The Shift Supervisor looks across at the radar image and notes thunderstorm over Bundeena but that the main part of the cell is over the water. His assessment is that the storm will move out to sea. |
| 7.36 pm | SAMU | Thunder heard at Sydney Airport. A series of SPECIs commence (special weather reports advising of hazardous weather conditions). |
| 7.37 pm | SAMU | TTF predicts thunderstorm conditions with hail between 7.45 pm and 10 pm and wind gusts to 42 knots. |
| 7.40 pm | RFC | Day shift Technical Officer (Observer) ceases duty. For the following 30 minutes (until 8.10 pm) there are now only two staff on duty in the RFC: the night Shift Supervisor and the night shift Aviation Forecaster). |
| 7.42 pm | SAMU | SPECI reports thunderstorm with rain at Sydney Airport. |
| 7.44 pm | SAMU | First SPECI reporting thunderstorm with heavy hail at Sydney Airport. These conditions and SPECIs continue until about 7.55 pm with wind gusts to 41 knots reported at times. |
| 7.45 pm | RFC | RFC becomes inundated with phone calls which are handled by night Shift Supervisor with assistance from night shift Aviation Meteorologist. |
| 7.55 pm | RFC | First report of severe hail near the city itself received when an off-duty Bureau officer rings in to report large hail in Newtown. |
| 7.55 pm | SAMU | Hail stops at Airport but thunderstorm continues with rain. Forecaster appends TTF predicting an easing of conditions, but intermediate periods of thunderstorms with hail and wind gusts to 42 knots (78 km/hr) until 8.30 pm and intermediate periods of thunderstorms and rain between 9 pm and midnight. |
| 8.00 pm | RFC | Night Shift Supervisor considers issuing Severe Thunderstorm Warning but regards it as too late, and concludes that the storm is likely to weaken. |
| 8.00 pm | SAMU | SPECIs and TTFs continue to be issued. Strong wind gusts stop (gusts now only 16 knots) but thunderstorm with rain continues. |
| 8.05 pm | RFC | Off-duty Severe Weather Meteorologist telephones RFC and offers assistance. Night Shift Supervisor indicates that he can handle the situation and continues taking phone calls. Off-duty Severe Weather Meteorologist dials-in remotely to Bureau systems at 8.18 pm to access radar and weather information and assesses the situation. |
| 8.10 pm | RFC | Night shift Technical Officer (Observer) arrives for duty. He had been delayed by 10 minutes by the storm on his way to work. This increased the staff on duty in the RFC to three which was the prescribed night shift staffing level. |
| 8.20 pm | RFC | Night Shift Aviation Meteorologist amends Richmond TAF, Williamtown TAF and ARFOR to reflect thunderstorm conditions. |
| 8.30 pm | RFC | Off-duty Severe Weather Meteorologist telephones off-duty Severe Weather Senior Meteorologist to discuss the situation. Off-duty Severe Weather Senior Meteorologist then calls State Emergency Services. |
| 8.45 pm | RFC | Severe Weather Senior Meteorologist telephones night Shift Supervisor and suggests that he contact the Regional Director. |
| 8.55 pm | RFC | Night Shift Supervisor receives a call from the SAMU night shift forecaster to discuss the proposed content of the Sydney Terminal Aerodrome Forecast (TAF) for the next 24 hours. |
| 8.57 pm | SAMU | Forecaster issues Sydney TAF for the 24 hour period commencing 10 pm predicting decreasing showers between 10 pm and midnight, then generally good weather conditions apart from a 30 per cent probability of reduced visibility due to mist between 3 am and 8 am. |
| 9.00 pm | RFC | Night Shift Supervisor contacts Regional Director, as does Severe Weather Senior Meteorologist shortly thereafter. Regional Director and Shift Supervisor agree that an off-duty Severe Weather Meteorologist be called in to work. |
| 9.10 pm | RFC | Night Shift Supervisor telephones Severe Weather Meteorologist and requests assistance. |
| 9.14 pm | SAMU | SPECIs and TTFs continue to be issued. Strong wind gusts to 38 knots (70 km/hr) with thunderstorm until 9.30 pm. Fluctuating shower and thunderstorm conditions continue until midnight. |
| 9.19 pm | RFC | Night Shift Supervisor amends the current Sydney Metropolitan Forecast for Wednesday night and Thursday, mentioning "Thunderstorms this evening with local hail, easing to a few showers overnight mainly in the coastal suburbs." |
| 9.30 pm | RFC | Severe Weather Meteorologist arrives at RFC and considers issuing a Severe Thunderstorm Warning but no warning is issued when it becomes clear from radar sequences that the storm is now moving out to sea and weakening. |
| 10.05 pm | RFC | Severe Weather Senior Meteorologist arrives at RFC and together with Severe Weather Meteorologist prepares Severe Thunderstorm Advice .indicating that the serious hail threat had eased, but that further storms were still likely with short periods of very heavy rain and localised flash flooding. |
| 10.18 pm | RFC | Severe Thunderstorm Advice is issued to the public. |
| 11.44 pm | RFC | Severe Thunderstorm Advice issued for coastal parts of Hunter District, including Newcastle, for large hailstones, destructive winds and very heavy rain. |
