Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Rural and Regional Affairs and Transport References Committee
Aviation accident investigations

HAINSWORTH, Mr Alasdair Horace William, Assistant Director Services, Bureau of Meteorology

HANSTRUM, Mr Barry Norman, Regional Director NSW/ACT, Bureau of Meteorology

JACKSON, Mr Gordon Edward, Head, Aviation and Defence Weather Services, Bureau of Meteorology


CHAIR: I welcome representatives of the Bureau of Meteorology. By the way, you do a great job at Wagga airport because I can ring up there and get the weather. It is fantastic.

Mr Hanstrum : Thank you, Chair.

CHAIR: I remind senators that the Senate has resolved that an officer of a department of the Commonwealth or a state shall not be asked to give opinions on matters of policy and shall be given reasonable opportunity to refer questions asked of the officer to superior officers or to a minister. This resolution prohibits only questions asking for opinions on matters of policy. It does not preclude questions asking for explanations or factual questions about how and when policies were adopted. Officers are also reminded that any claim that it would also be contrary to the public interest to answer a question must be made by the minister and shall be accompanied by a statement setting out the basis for the claim. The bureau has recently lodged a submission with the committee. Do you want to make any amendments or alterations to that?

Mr Hanstrum : No.

CHAIR: Would you care to make an opening statement?

Mr Hanstrum : I would just like to make some comments for the benefit of the committee that would set the scene regarding the particular meteorological hazard in this case, which was low cloud. Norfolk Island is prone to the incidence of low cloud. On average, if you fly into Norfolk Island, about 10 per cent of the time you will find that cloud is below the alternate minima for that airport. But on the particular night of this event the cloud base was around 200 feet. That is a very rare event indeed. The likelihood of encountering a cloud base of 200 feet is less than one per cent or one in 100, and even though the island is prone to these low cloud incidents this is an extreme event. It is very unusual indeed to see a cloud base of 200 feet at any time of the year at Norfolk Island and it is even less likely in November. The incidence of low cloud is at a maximum during the warmer months of the year. That is when the warm sea surface temperatures surrounding the island increase, so this adds to the available moisture in the air and raises the likelihood of low cloud.

The most prone period during the year is the summer months, from December through to February, so on this night there were particularly low odds of encountering cloud this low. The way cloud forms on the island is particularly fickle. It often forms locally due to the prevailing winds blowing across the island, and so it is very much driven by the island topography itself. Pilots report flying across the sea with clear conditions but then reaching the island to find that it is covered in cloud. So the island does have a significant impact on the weather in the area; it lifts the cloud, quite often resulting in a low cloud base over the island, which does not occur in the surrounding areas.

In relation to the submission that the bureau has provided, for the benefit of committee I thought it would be useful to go through the sequence of meteorological events that occurred during this incident, and essentially this chronology is provided as part of the—

CHAIR: Would you care to table that?

Mr Hanstrum : This is included, Chair, in the submission, and I have extra copies if you would like. So, just running through the sequence of observed and forecast weather on the day, in the period from around 0430 to 0630 UTC—and I will refer to UTC times here—the cloud base was around 2½ thousand feet with a few lower patches. At 0437 the routine aerodrome forecast was issued for Norfolk Island, valid for the period from 0600 to 2400. It referred to scattered cloud with a base of 2,000 feet and light showers from 1500 with the passage of a weak cold front, which was located to the south.

CHAIR: Are these New Zealand observations?

Mr Hanstrum : These are observations from the bureau's station at Norfolk Island. The bureau has an automatic weather station there as well as a weather watch radar.

CHAIR: Just to further clarify, are those observations made available to New Zealand because it is their air space?

Mr Hanstrum : Yes, that is correct. Those observations are transmitted to the Sydney office then to our head office in Melbourne through the computer message switching system and then distributed to air services and internationally to adjoining flight information regions through the Australian aeronautical fixed telecommunications network. The first real indication that conditions were deteriorating came at 0739. This was a

SPECI report. A SPECI report is an aerodrome weather report which indicates that significant change has occurred in the weather. In this case that SPECI report indicated that the cloud base had lowered below the alternate minima and was now broken at 1,100 feet.

Senator XENOPHON: When did Mr James, the pilot in command, in the context of these universal times, plan his flight and when did he actually depart Samoa? I just want to try to put it in context, in parallel.

Mr Hanstrum : From the ATSB report—I guess that is the source of the information regarding the pilot's activities for us—the pilot planned on the 0437 terminal area forecast.

Senator XENOPHON: At what time, though, did he plan that?

Mr Hanstrum : As I understand it reading from the ATSB report, he had access to that product, which was issued at 0437. So I guess some time after that, subsequent to 0437. At 0739, the first of these special aerodrome weather observations was sent and received by forecasters at the Sydney regional forecasting centre. On the basis of that deterioration in weather conditions, the terminal area forecast was amended some 24 minutes later, at 0803. The amended TAF, valid from 0800 to 2400, indicated below alternate minima conditions, with a cloud base at a thousand feet.

Senator XENOPHON: When was the pilot told that?

Mr Hanstrum : Both the special observation that was received at 0730 and the amended forecast issued at 0803 were around two to 2½ hours prior to the arrival of the aircraft at Norfolk Island.

CHAIR: So when was the deteriorated, risky weather put into the system?

Mr Hanstrum : That was put into the system at 0739.

CHAIR: And the pilot was still, two hours from destination, able to divert to Noumea?

Mr Hanstrum : Yes—

CHAIR: But it was put into the system but was not transposed into the pilot's ear by New Zealand?

Mr Hanstrum : What happened with air traffic in New Zealand is something we do not have knowledge of. We put in into our system and delivered it very shortly after 0739. My understanding is that the communication of these messages is very quick. One of my colleagues may wish to comment on that.

Mr Jackson : The information that we get from Norfolk Island is, as Barry said before, sent to Sydney office. It goes into a CMSS, or central message switching system, and then goes to the aeronautical information provider, which, in Australia's case, is Airservices Australia. It also goes by AFTN to air service providers in other FIRs. It takes only one or two seconds to get from Norfolk Island to New Zealand and to Fiji.

Senator EDWARDS: You said that, at 8.03, you got more information. But in your evidence you just said that, from 0800 hours, there was an amended forecast.

Mr Hanstrum : Yes, it was valid from 0800 hours to 2400 hours, but the actual issue time was 0803.

Mr Hainsworth : In essence it was backdated by three minutes to provide the pilots with the maximum amount of information and give them the impression that the change had occurred immediately and was valid from that moment.

Senator EDWARDS: How long before the plane ditched was that?

Mr Hanstrum : From reading the ATSB report, I understand that the aircraft made several attempts to land, commencing at just after 1000 hours. The period from 1000 through to 1030 was when the four missed—

CHAIR: We better let you finish your opening statement.

Mr Hanstrum : At 0803 the amended forecast had been issued. In the period from then until around 0930, the conditions fluctuated. But around 0930 there were further special weather reports which indicated a further lowering of the cloud base down to 200 to 300 feet and a reduction in visibility to 4,500 metres in showers. Following that observation, at 0958, the forecast was again amended to reflect up to one hour of cloud base at 500 feet and low visibility during that period. In the period between 1000 and 1030, which, as I have said earlier, was the period when the aircraft was making its attempts to land at the island, the SPECI reports indicated overcast cloud at 200 feet and that the visibility was 3,000 metres in showers. This low cloud event lasted around four hours. The poorest conditions occurred in, roughly, the half-hour period between 1000 and 1030—the period, as fate would have it, when the aircraft made several attempts to land. By 1100, a half-hour later, there was another SPECI report from the automatic weather station indicating that the cloud base had lifted slightly to 500 feet—up from 200 feet—and that the visibility had increased from 3,000 metres to 5,000 metres. Just less than half an hour later, conditions improved even more—the rain stopped, the cloud base lifted and the weather returned to above alternate minima conditions.

To summarise: a cloud base of 200 feet at Norfolk Island is a very rare event indeed, particularly in November. On the evening of the incident, there was around four hours—from 0739 to 1128—when the cloud base was below the alternate minima. There was a period of between half an hour to three-quarters of an hour when the weather was at its poorest—when the cloud base was around 200 feet and the visibility was 3,000 metres.

CHAIR: What was it when the pilot tried to land?

Mr Hanstrum : When the pilot tried to land was really when the conditions were worst. The cloud base was 200 feet.

CHAIR: Perhaps I could just clarify something: the airfield is at how many feet?

Mr Hanstrum : The airfield is at 171 metres. This is cloud above the airfield, not above sea level. So, that is really the summary of the events that occurred on the evening. Also in the bureau's submission is some information related to the weather radar at the time, which shows some very weak and light shower activity around. There were no thunderstorms associated with this storm. There was no particularly severe turbulence and there was no particular wind shear associated with this storm that would have affected the operation of the aircraft.

CHAIR: So, from your perspective—the bureau's, and being a farmer I am pretty interested in the bureau—you pass on your information to Airservices Australia, and then your responsibility ceases.

Mr Hanstrum : Yes. The bureau does not have responsibility to provide all—

CHAIR: I am sure it gets to—

Mr Hanstrum : Yes, that is right. Our role is to get the information to Airservices.

Senator FAWCETT: I would just like to again step back to the holistic system of what might have prevented the accident. One of the key points that came out was that it appears that, when the Fijian authorities transmitted just after 0800 the updated report that indicated the deterioration in the weather, it was garbled but there was no attempt made to actually clarify it, which goes perhaps to a mindset that the initial TAF was good—the first report given at 0800 was good. There does not appear to have been the awareness of how variable Norfolk Island is. Perhaps you could outline something for me. Do any of the training materials that the Bureau of Meteorology provides to the aviation community—and I know you used to have two fine blue manuals that looked at meteorology in general and aviation meteorology specifically—talk about locations in our region that are known to be extremely variable?

Mr Hanstrum : I think one of the really useful resources the bureau provides through its website is a list of station climatologies. In that information is the historical record of observations below certain cloud heights and also visibility below certain criteria. So whilst not explicitly referring to Norfolk Island as part of its educational material, the information is available through the bureau's site and is accessible. The statistics I provided in my introductory remarks are taken from that fairly accessible information available online.

Senator FAWCETT: What I am getting at here is the case of a person who has never experienced that, whatever it was—did you it was a two per cent chance?

Mr Hanstrum : Yes, less than that.

Senator FAWCETT: For a person who has never experienced that, there is nothing in the training materials to flag his or her attention to that variability. In the actual product that is delivered by you to Airservices distribution, is there any annotation to things like TAFs or SPECIs that indicate that whilst this was valid it could, at this particular station, vary considerably without notice? Is there any annotation to that effect?

Mr Hainsworth : Certainly not on the product itself that is issued, no.

Senator FAWCETT: So there is no correlation from that list of station climatologies you have talked about that would then automatically annotate particular forecasts to say, 'Be aware that there is a huge variability that may be associated with these reports'?

Mr Jackson : Our products are all ICAO compliant. Therefore, products that would go out for international airports, such as Norfolk Island, have to be in the ICAO standard format. So there is no additional annotation on the products.

Senator FAWCETT: I accept that. It does not stop us working with ICAO to change that, if indeed that is required. To come to the Unicom operator. I know that is not your direct responsibility but you have the weather station there on your radar. Does the Unicom operator receive your information directly or via ASA? Is he in a position to see the raw data coming from your weather station at Norfolk?

Mr Hanstrum : He would be able to access that information. There is an AWIS broadcast system there—and that would be the prime means by telephone.

Mr Hainsworth : He would have access to the radar either through the website or indeed they may have a direct feed there, but certainly through the website.

Senator FAWCETT: For the Unicom operator to become an approved observer recognised by CASA or the Bureau of Meteorology, what additional training would be required? What kind of a burden would there be on the Norfolk community for that person to be an approved observer as opposed to just the Unicom operator?

Mr Jackson : The procedure which is now in place is that they will be required to do a pre-reading in our one-week training course and then an additional week of observations. All up, it takes about two weeks' dedicated observer training, plus some pre-reading is required. That authorisation will be for about two to three years. We are finalising the new procedures now and the validity period for that.

Senator FAWCETT: Is it your understanding that the reason the Unicom operator did not contact the aircraft via HF was that he was not licensed to use the HF or that he was not an approved observer?

Mr Hainsworth : We are not aware of that.

Mr Hanstrum : No.

Senator XENOPHON: Further to Senator Fawcett's line of questioning, with such serious weather at Norfolk Island—it can get pretty ugly pretty quickly there—why depend for what is conveyed to Airservices Australia on an automatic weather station and a narrow radar signal rather than on human observations on the ground?

Mr Hanstrum : The automatic weather station has a ceilometer and vismeter, and that is really the standard by which observations are provided to the aviation industry around Australia. Most airports that aircraft operate into have an automatic weather station with ceilometer and vismeter as the standard means of reporting. That equipment performed very well and very reliably on the night of the incident and provided very rapid updated information to our forecasting staff in Sydney.

Senator XENOPHON: In terms of Senator Fawcett's final set of questions to you, don't you see any point in the person on the ground having the ability to communicate directly to the pilot in command? I think Mr Hainsworth was saying—

Mr Hainsworth : That is not what we said.

Senator XENOPHON: I am sorry, but I am just trying to clarify.

Mr Hanstrum : No, we do not see value in our weather observer broadcasting directly to aircraft. Our role is to—

Senator XENOPHON: I understand that, but alternatively would there be value in the person on the ground giving information directly to either the bureau to pass on to Airservices Australia or to air traffic control in New Zealand or Fiji or wherever the space is controlled?

Mr Jackson : The AWS has also got an AWIS system, which broadcasts the weather observations, so a pilot within a certain distance off the island can get the observations automatically using HF radio—which is the same way as the observer would be able to transmit that information.

CHAIR: So, is there an automatic repeater arrangement?

Mr Jackson : There is a broadcast system at Norfolk Island so that they can just dial up the frequency.

CHAIR: But it is an automatic repeat?

Mr Jackson : Yes, it automatically repeats.

Mr Hainsworth : And therefore the observation is available 24 hours a day, seven days a week, and it updates on a minute-by-minute basis. It provides more information than could possibly be provided by a manual observer.

Senator XENOPHON: But, in terms of the status quo of the forecasting at Norfolk Island, why does the AIP use them 'with caution'? Is that your understanding?

Mr Hanstrum : Sorry, AIP?

Senator XENOPHON: The aeronautical information publication.

Mr Hainsworth : I think probably the basis of that is that an automatic weather station has a single point of cloud base. It seeks the cloud base from a single point over the airfield. And there are occasions where you could have situations where the cloud is simply over that one point, or indeed where there is no cloud over that single point, which might be seen by a manual observer. So that could be the reason why they indicate—

CHAIR: So you cannot tune in, as you can here, to the radar and say, 'There's cloud out there at Mildura and down at Wangaratta'?

Mr Hainsworth : No, not necessarily like that. It means that it is a point; it is a laser that looks vertically—

CHAIR: My point is that the bureau's arrangements in Norfolk Island are very localised; they are not looking out. Does the weather come from east to west or just form locally?

Mr Hanstrum : In the winter months it often comes, as it does in southern Australia, from west to east—

CHAIR: Sorry, west to east.

Mr Hanstrum : but in the summer months it can also come from the north and east, as the influence of the monsoon—

CHAIR: How far out do you have the capacity to pick up the weather—only at the airfield?

Mr Hanstrum : We have a weather watch radar. The range of that radar, if there were heavy rain around, would be of the order of maybe 150 to 200 kilometres in a radius around the island.

Senator XENOPHON: Could you clear this up for me. You are familiar with the ATSB report on 22 February 2000 about the reliability of Norfolk Island forecasts?

Mr Hanstrum : Yes.

Senator XENOPHON: There was a recommendation issued to the Bureau of Meteorology. The subject is 'Reliability of Norfolk Island forecasts: safety deficiency', and it says:

The meteorological forecasts for Norfolk Island are not sufficiently reliable on some occasions to prevent pilots having to carry out unplanned diversions or holding.

Then it lists a number of incidents including some pretty large passenger aircraft—a BAe 146. You are familiar with that particular report?

Mr Hanstrum : Yes.

Senator XENOPHON: That was almost 13 years ago. Can you tell us what has changed since 22 February 2000 as a result of that ATSB recommendation issued to the bureau?

Mr Hanstrum : One of the recommendations in that report was that the bureau consider installing a weather radar at Norfolk Island.

Senator XENOPHON: And that was done?

Mr Hanstrum : That did happen.

Senator XENOPHON: When did that happen?

Mr Hanstrum : In 2003.

Senator XENOPHON: And other recommendations?

Mr Hanstrum : Other recommendations related to the way in which weather information was passed to weather forecasters. In 2002, a ceilometer and visibility meter facility was installed on the island, which transmits its information automatically and immediately to forecasters. That facility has largely superseded the need for there to be a call to our forecasting office from the observers, as was the case before that instrumentation was available to the forecasters.

Senator XENOPHON: Just so that I can move on, because I have series of questions to ask you, could you on notice provide details of the recommendations issued back on 22 February 2002; the extent to which they have been carried out; if so, on what date; and, if not, the reasons for those recommendations not being carried out? That just might put that in context.

Mr Hanstrum : Certainly, Senator.

Senator XENOPHON: That might be useful. In terms of aviation decision making, how long is it reasonable to rely on the conditions reported in a Met observation? Is that something you could comment on?

Mr Hanstrum : An important thing about forecasting and weather observations is that it is a continuous process. An observation is updated regularly, every half-hour, but if there is a significant change in the weather then that new information would be updated immediately, and that is done on a minute by minute basis. On the basis of those changing conditions, the forecasters are monitoring those observations and reviewing the forecasts accordingly.

Senator XENOPHON: My understanding in this case is that the pilot in command only got the 0803 details, not the 0739. Is that your understanding? But you did your job by providing both the 0739 and the 0803?

Mr Hanstrum : The 0739 special observation was issued. At 0803 that was followed by an amendment to the terminal area forecast for Norfolk Island.

Senator XENOPHON: In your submission at page 1 you say, 'Amendments to a TAF are issued in the event that one or more of the forecast elements described above varies by an amount that is significant to operations at the aerodrome.' What are the weather triggers for terminal area forecasts—TAF—amendment, and what are the standard amendment time frames once those significant variations are reported?

Mr Hanstrum : In the case of Norfolk Island the important cloud base is the alternate minimum, which is 1,169 feet. So as soon as the observed—

Senator XENOPHON: At Norfolk Island?

Mr Hanstrum : Yes, at Norfolk Island.

Senator XENOPHON: How many feet above sea level is the aerodrome there?

Mr Hanstrum : The aerodrome is around 500 feet above sea level, but all of these heights are in reference to the aerodrome, not mean sea level.

Senator XENOPHON: I am sorry, okay. Also in your submission you say, 'When the forecast or actual conditions at an airfield decrease below the alternate minima, aircraft flying to that airfield must either carry fuel for flight to an alternate airfield or fuel to allow the aircraft to remain airborne until the weather improves sufficiently for a safe landing to be conducted.' What does the bureau base that 'must' statement on, and how does it work for an unforecast deterioration in the weather?

Mr Hanstrum : That statement came out of the ATSB report in 2000. That is not a bureau statement; it is really a statement of fact about the way in which the bureau's observations and forecasts are used by aviation.

Senator XENOPHON: So has the word 'must' come into it since that 2000 report of the ATSB?

Mr Hainsworth : No, that is in Air Services regulation.

Mr Jackson : It is the wording that they used—ATSB.

Senator XENOPHON: And has that been since 2000?

Mr Hainsworth : Longer, much longer.

Senator XENOPHON: All right. How does the bureau monitor the accuracy of the AWIS cloud base and coverage detection—whether it is scattered, broken et cetera? What is the relevant performance of equipment on Norfolk now and at the time of the ditching? Is there any difference in the performance of the equipment now compared with 2009?

Mr Hanstrum : It is the same equipment that is in place there.

Senator XENOPHON: So there has been no change to the equipment?

Mr Hanstrum : No change to the equipment.

Senator XENOPHON: Could you have different equipment there that could make a difference in terms of accuracy of forecasting?

Mr Hanstrum : It is difficult to say. There is a limit. If we had observations all over the island, we may improve the forecast slightly, but the observation that we have there is representative of the conditions at the aerodrome and reflects the conditions adequately for forecasting.

Senator XENOPHON: Perhaps you may want to take this on notice: how does the bureau monitor the accuracy of forecasts on Norfolk Island and, say, Christmas Island? Is there a similar way of determining it? Is it identical, or are there any differences?

Mr Hainsworth : Yes, we maintain a verification system within the bureau that monitors the accuracy of our terminal area forecasts for all terminal area forecasts that we issue within Australia. That includes Lord Howe, Norfolk, Cocos and Christmas Island.

Senator XENOPHON: Just going back a second, is it your understanding—maybe it is outside your purview—that the pilot in command only got the 0800 details, not the 0803? What is your understanding of that? It is a separate issue—I think it is not appropriate to ask that.

CHAIR: You have a repeater station, where you can tune in on HF for the weather, right?

Mr Hanstrum : Yes.

CHAIR: What is the frequency?

Mr Jackson : I think we will have to get back to you on that one. Each different AWS has its own frequency.

CHAIR: You would be aware, I suppose, if you have been there long enough, that HF is very variable, depending on your frequency. I had four frequencies set up on my HF, where I could talk to Darwin easier than I could talk to Hay—and Broken Hill and the Royal Flying Doctor Service was sort of in between. As a bureau, is it your responsibility to check the repeater station?

Mr Jackson : No, it is not, actually.

CHAIR: Who is?

Mr Jackson : We provide the automatic weather station. Normally it is up to the aerodrome to provide the repeater station.

CHAIR: Would it be your responsibility, or would it not, that the equipment was the right equipment for the job?

Mr Jackson : It would be authorised; yes.

CHAIR: So in the authorisation of that, where did you set the optimum reception—1,000 kilometres off the island or 200 kilometres off the island—because HF varies considerably in its reliability depending on the frequency used? If you were responsible legally for that could you show us the workings?

Mr Jackson : I don't think we are responsible for allocating the frequency for that. We provide the automatic weather station and—

CHAIR: Given that you supply the information and you think the aerodrome—

Mr Jackson : In most cases the aerodrome is the one—

CHAIR: Who owns the aerodrome at Norfolk Island? I know the minister for public works—so-called, in an island of 1,400 residents at the time!—got the contract to build the extension, when he was the minister.

Senator Nash interjecting

CHAIR: That's right! As I said to him at the time, 'If you were on the mainland, mate, we'd lock you up.' Do you know what I mean? There are so many unknowns in this. The HF—as opposed to VHF, which Senator Nash would be familiar with, is very good in line of site—

Senator FAWCETT: Chair, can I clarify. You have been talking about HF. Norfolk Island may be an exception but my experience is that most stay with VHF and occasionally go to the NAV band frequency and you get it through your VOR. Is it definitely HF?

Mr Jackson : I may have made a mistake there. I will check up on that and check on the frequency.

Senator FAWCETT: Take it on notice, but I think you will find it is probably VHF seeing that the vast majority are.

CHAIR: That would be more or less line of sight.

Senator FAWCETT: That does not help you two hours out, necessarily.

CHAIR: That would absolutely change my thinking—

Mr Jackson : Sorry about that.

CHAIR: if it is VHF, because the poor bugger, if he is two hours away, ain't going to get it.

Senator XENOPHON: At page 2 of your submission you have said that the probability of encountering unforecast adverse weather conditions 'is therefore 2.7 per cent'. That is in relation to cloud forecasts at Norfolk Island. Having 27 flights out of 1,000 potentially affected seems quite high. How does that compare with similarly equipped mainland aerodromes?

Mr Jackson : For Sydney airport we have 99 per cent reliability, so it is slightly less, but at Sydney airport we have 24-hour forecasters who specifically do Sydney. So that is probably a gold standard. We are looking at a one per cent unreliability at Sydney airport.

Senator XENOPHON: But the difference between Sydney airport and Norfolk Island airport is that you have Canberra, Newcastle, Brisbane and Melbourne—a whole stack—of aerodromes to go to, fairly close by.

Mr Jackson : If you look at other aerodromes—we did a study of Lord Howe Island—you find that they have very similar statistics.

Senator XENOPHON: I will move on because of time constraints. The amended terminal area forecast issued at 0958 seems to indicate weather that was very optimistic for the first 1½ hours, in that the prevailing conditions were worst than forecast. Is that fair enough?

Mr Hanstrum : It was amended, valid 1000 to 2400, to indicate periods of one hour of cloud at 500 feet and visibility of 4,000 metres occurring.

Senator XENOPHON: But the prevailing conditions were worse than forecast.

Mr Hanstrum : The cloud base was worse than forecast. It was around 300 feet lower than indicated in that amended—

Senator XENOPHON: What I am trying to understand—and this is not a trick question; I am just trying to understand this—is: would that show up in your statistics as a bad forecast, or does the fact that it was a good forecast for the last 12½ hours outweigh the first 1½ hours? Mr Hainsworth is smiling, but I am just trying to understand how, statistically, it works out.

Mr Hainsworth : The verification of these things is extraordinarily complicated because we are always updating the forecast.

Senator XENOPHON: So when is it a good or a bad forecast? Do you take it on average, or how bad the weather is? I will be guided by Senator Fawcett on this, but I am just trying to work it out. If you have a good forecast for 12½ hours, does that outweigh the last 1½ hours?

Mr Hanstrum : No. It is just the first 12½ hours and then the last 1½ hours. So every hour we are making a comparison between what the forecast was—

Senator XENOPHON: How do you weight it, though? Do you weight it?

Mr Hanstrum : No, it is not weighted.

Senator XENOPHON: So then the first 12½ would outweigh the last 1½, wouldn't it?

Mr Hainsworth : The fact that it is below alternate minima would indicate—and if it was forecast below alternate minima, it would show up in the verification statistics that this has appeared where the forecast has failed. Regardless of whether it is an amended forecast or a routine forecast, it would indicate within the statistics that it was a bad forecast.

Senator NASH: For that period of time, rather than an overall analysis.

Mr Hainsworth : Yes. That is right. We look at it so that we can identify the percentage of time that it is below alternate minima or forecast to be below alternate minima. We can assess that from the statistics, but it is not weighted to say, 'The vast majority of the time it's fine.' No; we do look at the period where it is actually below alternate minima.

Senator XENOPHON: Going back quickly to what Senator Heffernan was asking: I have just had a message from a pilot—because a few pilots are listening in to this or watching it. What are the VHF and HF frequencies that you can tune into? This pilot tells me that he cannot find it. Is that something that is easy to tune into or to locate?

Mr Jackson : We would have to get back to you on the frequency.

Senator XENOPHON: Can a pilot readily find that?

Mr Jackson : The frequency should be in ERSA.

Senator XENOPHON: Should be where, sorry?

Senator FAWCETT: En Route Supplement Australia, a document published by Airservices that pilots can look up. It has got all the frequencies.

Senator XENOPHON: So that is where it should be, if you could just confirm that. You provide examples in your submission of infrared satellite images. Do they show high- or low-level cloud? I am sorry; I am not experienced in these issues.

Mr Hanstrum : They show the temperatures of the cloud tops. In that image where you see very bright colours or white shades, they are very cold clouds and therefore very deep and likely to be associated with quite heavy rain or thunderstorms. Where you see those grey shades, they are comparatively warmer cloud tops, which are lower, and it would be less clear from looking at that satellite image whether or not the cloud had the potential to produce rain.

Senator XENOPHON: Excuse my ignorance on these issues, but does that mean that high-level cloud obscures low-level cloud?

Mr Hanstrum : Yes, it does.

Senator XENOPHON: So what conclusion would you expect a pilot flying to Norfolk Island to draw from the 10.30 UTC satellite picture?

Mr Hanstrum : The conclusion that you would draw is that there is a band of cloud—it does not look particularly deep—but there is no information from that image itself related to exactly what the cloud base might be. It simply would say that there is a layer of cloud that is not particularly deep, but it does not provide the sort of information about how high the cloud base would be, for example.

Senator XENOPHON: So, in a nutshell, what conclusion do you think is reasonable for a pilot to take from that 10.30 image?

Mr Hanstrum : I think the message would be that there is a weak frontal cloud boundary that is passing the area. But there is not much more information about the surface weather conditions than that; it really is just a tool to get an overall impression about what the major weather systems are in the area.

CHAIR: Do you have a different responsibility or different thinking? Is there an ILS at Norfolk Island? I do not think there is.

Mr Hanstrum : No, I do not think so.

CHAIR: So the reporting requirements for somewhere, a speck in the ocean where there is no ILS, would be different, in terms of thinking, to one in Noumea where there is. You do not have to worry about that?

Mr Hanstrum : No. Our role is to just forecast—

CHAIR: So your description of 'dirty' weather—I would probably use a different word—for somewhere where there is no ILS would be no different to where there is an ILS?

Mr Hainsworth : No. We are there to forecast the weather conditions as we see them. We do not take into account what landing systems are there.

Senator XENOPHON: Further to what the Chair asked, you provided examples of the weather radar picture. Relatively speaking, the 10.30 radar picture seems quite benign, yet conditions for the airfield were pretty awful. Is that a fair summary?

Mr Hainsworth : Yes.

Senator XENOPHON: Excuse my ignorance on these things, but how does that come about? What conclusion would you expect to reach from that 10.30 UTC radar picture when flying to Norfolk Island?

Mr Hanstrum : It is the role of the professional meteorologists who are providing the forecast to make that interpretation and to translate that weather data, of which there are many sources—satellite imagery is one—into the forecast that the pilot would access and make direct reference to.

CHAIR: Wouldn't you have to have input into the condensation level to be able to predict that?

Mr Hanstrum : Yes.

CHAIR: And do you?

Mr Hanstrum : Yes.

CHAIR: At Norfolk Island?

Mr Hanstrum : At all of the aerodromes that we do forecasting for.

CHAIR: As you know, fog—

Senator XENOPHON: So, regarding the bureau's staffing at Norfolk and Christmas Islands, what is the current position in those two places?

Mr Hanstrum : At Norfolk Island there are three observing staff rostered on duty from around seven o'clock in the morning through until—

Senator XENOPHON: It is four o'clock sometimes? It could be as early as 4 am?

Mr Hanstrum : Normally it is 6 am to 7 am.

CHAIR: It is a lazy job!

Senator XENOPHON: And Christmas Island is the same?

Mr Hanstrum : I am afraid I am not in a position to comment on Christmas Island.

Mr Hainsworth : We have no bureau observers on Christmas Island. On Cocos Island we do, but not on Christmas Island.

Senator XENOPHON: What are Australia's obligations and protocols for the transfer of met information to the aviation met providers in neighbouring countries? Can you tell us?

Mr Jackson : We transmit all our observations and weather forecasts to neighbouring countries through the Aeronautical Fixed Telecommunication Network, the AFTN.

Senator XENOPHON: So it is not handled by Airservices on your behalf; you do it directly?

Mr Jackson : We put the addresses on it and it goes out. I am not clear on whether it actually goes through Airservices before it goes out or it goes directly from us, but, either way, we put the addresses on it to go to overseas countries.

Senator XENOPHON: What sort of research does that bureau conduct on aviation met issues? What sorts of research programs are there? What is the ongoing research on aviation met issues?

Mr Jackson : We have quite a bit of research. It is funded by the aviation industry, for fog and for thunderstorms. There is not one at the moment for low cloud, but we have had it in the past.

Senator XENOPHON: Would low cloud be an important issue to look at?

Mr Jackson : Low cloud is a very relevant issue, but we just have to change, depending on the funding that is available.

Senator XENOPHON: Would it be as important as thunderstorms, for instance?

Mr Jackson : I think thunderstorms are very important, especially with microbursts and downdraughts. We have a whole suite of criteria.

Senator XENOPHON: This is my final question. The pilot who has been following this hearing this morning said that he can only get a phone number—there is only a phone number on the site to get the frequency. You would have to ring up to get the frequency. I suppose this is an issue that is not directly related, in terms of getting the actual frequency. Could you confirm whether it is easily attainable?

Mr Jackson : Yes, we will.

Senator NASH: This may have been covered. I apologise that I had to leave the room for a little bit. There is the recommendation from the ATSB on reviewing how you get the forecasting to Norfolk Island. Has that been covered?

Mr Hanstrum : We touched on it briefly. The senator asked for a written response to the recommendations that were made in 2000 and what the bureau had done with those recommendations since then. We will prepare that.

Senator NASH: Thank you. In terms of the procedures and getting forecasting to Norfolk Island since the incident, have there been any changes to the practices of the bureau in that time, as result of the incident, being aware of it or just in general? Have the procedures remained the same?

Mr Hanstrum : The bureau is continually reviewing its forecasting services, so we have looked at this event in combination with the broad empirical knowledge of forecasters over many years of forecasting for Norfolk Island and added the additional knowledge from this event into our forecasting processes. It is not something we wait for a review to prompt us on; this is something we do as part of our routine business. In this case, we have looked at this particular event and we have added that to the knowledge base we have gathered over years of forecasting experience on the island.

Senator NASH: In terms of the knowledge base, you have three individuals on Norfolk Island; have they been there for a long period of time? I am happy for you to take that on notice. You have got all the forecasting and all of the fabulous stuff you do, but in terms of—

CHAIR: Do you put it up for bid? Do you have an auction system for a job there?

Mr Hanstrum : It is one of our more popular postings, yes!

Senator NASH: In terms of corporate knowledge and long-term on-the-ground knowledge and understanding of weather in a particular place, have they been there for a while?

Mr Hanstrum : The observers that were there during the time of this incident had been there for a couple of years at least. The standard posting term is two years but it can be extended to three, and mostly it is extended to three for observing staff at Norfolk Island. We do not rely on having that experience on the ground. That knowledge and experience are incorporated into our standard operating procedures, which are effective independent of who is on duty at the time.

Senator NASH: It just seems to be such a quirky geographical place that on-the-ground knowledge would contribute, I imagine.

Mr Hainsworth : The observers on the island lean heavily on the automatic weather station output, particularly with regard to cloud base and visibility. They utilise the information and the equipment that is there, so it is not necessarily as dependent on experience as it was in the past, prior to the installation of that equipment.

Senator NASH: When did that equipment go in?

Mr Hanstrum : The ceilometer and visibility meter—the ceilometer measures cloud base—went in in 2002.

Senator NASH: Thanks.

CHAIR: Thanks very much for your trouble and patience. That concludes today's hearing. Thank you to all those kind, gentle people up behind the glass there and the professional staff.

Committee adjourned at 12:03