Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
STANDING COMMITTEE ON INDUSTRY, SCIENCE AND INNOVATION
30/06/2009
Long-term meteorological forecasting

CHAIR (Ms Vamvakinou) —I declare open this public hearing for the inquiry into long-term meteorological forecasting in Australia being conducted by the House of Representatives Standing Committee on Industry, Science and Innovation. The inquiry arises from a request to this committee by Senator the Hon. Kim Carr, the federal Minister for Innovation, Industry, Science and Research. Written submissions were called for and 33 have been received to date. The committee is now conducting a program of public hearings and inspections. This hearing is the sixth for the inquiry.

I welcome our first witness today, from Qantas Airways Ltd. Although the committee does not require you to give evidence under oath, I should advise you that these hearings are formal proceedings of the parliament. Consequently, they warrant the same respect as proceedings of the House itself. It is customary to remind witnesses that giving false or misleading evidence is a serious matter and may be regarded as a contempt of parliament. We thank you for your submission and now welcome you to make a brief opening statement before we proceed to questions.

Mr Rennie —Thank you. I have managed the flight dispatch, flight planning, tactical and strategic areas for many years—15 in total. In the past week I moved into a new capacity as principal adviser, global operations development. I am here as an operational person, not as a meteorologist or a scientist. I am here to give the operational perspective for Qantas and how we see the importance of met in the future. To give a bit of background, Qantas is one of the world’s leading airlines in terms of operational activity with meteorology, including meteorology with volcanic ash. We are very active and enjoy a collaboration with the Bureau of Meteorology unlike any other country in the world. We have developed this collaboration with the bureau over many years and we enjoy this two-way activity. Globally, we also have close collaborative ties with IATA, the International Air Transport Association, and ICAO, the International Civil Aviation Organisation. We are innovative, with the World Area Forecast Centre, and Qantas was instrumental in improving the resolution for the world area forecast that is produced by Washington and the UK daily—it is now four times per day. We have been instrumental in improving both spatial and temporal resolutions.

We understand that meteorology is not a perfect science, and since we have understood that we are much better at it. Operationally, Qantas has a met department, which is not unusual in aviation among airlines, but these days our model is quite different, whereby the meteorology department within Qantas adds value. It is a very integrated department within the Qantas integrated operations centre and it emphasises the importance that we place on meteorology and the environment.

Regarding the future, we see meteorology being absolutely instrumental in the way that we progress. There is improvement in efficiency—and safety of course—and, from an air traffic management perspective, it is going to be vital. At a recent meeting with the Federal Aviation Administration in the US, where we were talking about the next air traffic management system, they produced documents stating that they have measured the current delays in airspace and that the cost to airlines is upwards of US$28 billion, of which they consider that two-thirds is avoidable. What this means is that they have put in a huge effort within the national met authority and other authorities in the US to improve this aspect. We are right there with them.

In terms of the next air traffic management system, which will be global and starts in 2015, we will be having what are known as business trajectories, where aircraft will receive a trajectory in 4D—that is, 3D plus time—and included in that is the weather. So, when we produce a trajectory within our operations office and that is accepted by air traffic control, that will take into account the weather. So the products and the accuracy we require will need to improve. From an air traffic perspective, with the congested airspace today and what is projected, they see it as the only way to reduce separation standards so they can manage the airspace. It is that important. Qantas as an airline have measured the environment very closely and we see it as being fundamental to improving safety; efficiency—fuel saving to the industry; and, of course, CO2 emissions.

CHAIR —On the issue of the products, in your submission you said:

Qantas develops probability forecasting products to ensure that operational efficiency is maintained in a range of circumstances.

Can you tell us what type of products you are developing and what type of circumstances they are being applied to. Are these products applicable to other industries?

Mr Rennie —The probability forecasting is really central to the future. A meteorologist, when he develops a strategy for a forecast, works in probabilities. Nothing starts and ends just like that. It is worked up in terms of probabilities. We learned in the 1990s just how important probabilities were. We developed a product in Australia called Code Grey and if you look at the aviation weather site today you will see Code Grey. That is a probability forecast for long-haul flights into Australia. It was developed by Qantas with the bureau and everyone can use it. It is simply a probability of the alternative scenario occurring rather than the forecast. The forecast may say one thing; this is the alternative scenario. When we are managing the risk of getting an aeroplane in it is very important that we understand that risk.

CHAIR —Are you being pre-emptive in the way you approach this?

Mr Rennie —Exactly and it is being pre-emptive in terms of fuel policy. This is not a safety aspect; it is simply an efficiency aspect. There is a probability of an alternative scenario occurring in X amount of hours, therefore should we or should we not take that into account? We put that through a risk filter and come up with an answer to that. It is not precise but what we have done since then is to improve the way that we manage risk at both destination and en route.

We have developed a product internally called OpRisk whereby we look at all of our forecasts, all of our destinations, in the light of probabilities. We may have a low pressure system or we may have a perfect day forecast, but what is the alternative in that time frame? In other words, we are looking at forecasts over maybe 24 hours but we are really looking at the half an hour when that aeroplane gets in so that is what we refine that OpRisk to. It is simply a refinement of that forecast. When you consider our global operations we look at one-hour fly-in and also at 16-hour fly-in. They are quite different in the way they have to be managed in terms of the environment. If you look at the back end of the forecasts, the accuracy drops off to a level whereby we must take account of that by measuring that risk. That is exactly what we do.

We forged products with Weathernews in Japan for Japanese arrivals. We found that in Tokyo there was consistently low-visibility weather which had not been forecast. For years we carried additional fuel. We forged a probability forecast with Weathernews. They have an independent look at the forecast at a very precise time, and at the forecast at the alternative airports, versus the official product from the Japanese meteorological authority and they give us another scenario. We measure that scenario in terms of risk. Since we forged that we have saved a lot of fuel and we have not had one failure—touch wood. We have had that now for five years.

Mr RAMSEY —A failure is where you are getting low on fuel and you have to divert.

Mr Rennie —Exactly. In this case it is very costly, obviously. We look at it in terms of misses, or failures, and false alarm rates. Unfortunately, at the moment the false alarm rate is quite high with probability forecasting. So that is where our efforts are going at the moment. We also have a probability forecast with the UK Met Office for low-visibility arrivals into London and Frankfurt. In terms of verification of the product, they have a very high success rate. When you consider operating an aircraft, a flight crew have to be recently trained in low-weather minima and autolands. If we have a pilot that is not recent then we can only approach in much higher minima. Therefore, in low-vis ops we cannot make an approach, so we have to divert until that weather becomes good. We can now look at probability forecasts up to three days out and measure the risk of each flight.

We also have probability forecasts with HKO of Hong Kong. Hong Kong airport is significant in the level of turbulence and windshear that it experiences. It is also subject to cyclones and high winds. So we have a probability of crosswind coming across Lantau Island and creating windshear across those approaches. We have spent a lot of time with the HKO in developing products with them, particularly with the probability forecasting. Our probability forecasting is now with the WMO, which is the World Meteorological Organisation. They are the chief scientists of the world, and they work out all the equations and so forth. They are now seriously looking at probability forecasting. It is the future. The FAA and Eurocontrol have recognised that probability forecasting is the future—and how each airline then measures that risk.

CHAIR —This is a product that Qantas has developed.

Mr Rennie —Yes.

CHAIR —Is this a world-leading innovative approach?

Mr Rennie —Yes. We use probability forecasting more than any other airline.

CHAIR —Why don’t other airlines emulate that type of product?

Mr Rennie —Good question. It is a question I ask myself.

CHAIR —I might change my attitude towards flying forever after today.

Mr Rennie —Regulatory rules require fuel policy. Qantas is very adept at flying long-haul operations and our fuel policy reflects that. In terms of our rules, we have a lot of flexibility with our fuel policy. To keep that developing, we do not needlessly carry too much fuel—in other words, you have perfect weather and you have very long alternates. Australia is not endowed with lots of airports, particularly for large aeroplanes. It is very costly to fly across the Pacific and carry Melbourne, Brisbane or Adelaide as an alternative airport. From a 747 perspective, to carry such an alternate across the Pacific would cost in the order of five to six tonnes of gas. That is a lot of money every day, four times a day. So we have a fuel policy that we tailor to that flight. It is well within the CASA rules, obviously. It is not a safety issue; it is simply that we can forge such policies.

In Australia we are not endowed with airports. It is a unique operating environment in that—at night, for example—we do not have a lot of ATC at airports; we do not enjoy that. And there are very strict rules for flying, whether it be twin jets or ETOPS rules. So we work very closely with CASA in forging fuel policy and the operational rules.

So we need to be smart. Other airlines that carry lots and lots of fuel around the world do not necessarily have to be that smart. But, as the FAA found out when they provided relief to their airlines to do this, in terms of the US carriers, they now operate across the Pacific in much the same way as we do. So we have a meeting of the minds here, certainly with American airlines.

Mr RAMSEY —Can I ask a question from personal ignorance: what is your alternate on Jandakot Airport?

Mr Rennie —I do not know. In terms of Jandakot, because we do not operate there, in terms of heavy metal, I cannot say. I can tell you about Perth or anywhere Qantas operates, but unfortunately—

Mr RAMSEY —What happens in Perth? Where are your alternates?

Mr Rennie —In Perth, we have alternates for the 737 and that size aeroplane. We can go to Kalgoorlie, Geraldton or further up north, or down south to Albany. For the 767 and 330s, we are more restricted, but we have Kalgoorlie, and then we go up to Learmonth. For the 747, it is Learmonth, before Adelaide.

Mr RAMSEY —Not close, is it!

Mr Rennie —Not close!

Mr RAMSEY —Boy oh boy, that is a lot of fuel.

Mr Rennie —That is exactly right. And when you look at the cost of fuel and the emissions, unless we had got very smart with how we operate, and we measured that risk, we would be in a much worse situation today.

Mr SYMON —I have lots of questions. Unlike other witnesses to this inquiry, your organisation appears to have a reach, obviously, into a lot of other countries’ met systems. Everyone we have heard from until now has come up and spoken about the Australian Bureau of Meteorology and CSIRO and various other things that we do here. My question is: in regard to the information you receive from Australia compared to other countries in the world, is the information you get here of a higher standard or a lesser standard? Are there places that do it better than we do?

Mr Rennie —Yes. When we look at weather, we have to look at the conditions. Not being a meteorologist, I may not be able to answer all of that question. We certainly measure forecasts around the world, and we have some verification internally. We certainly push very hard ICAO and the bureau of met and other met authorities to introduce verification. Verification is a key to improving the accuracy of their forecasts. We find, for example, that forecasting in Australia is generally quite good—as a general statement. What we find, though, is that forecasters can be overconservative, for example. We need to improve verification to identify that.

On fog forecasting: we find that, in Australia, fog is obviously very disruptive. It is incredibly disruptive. We need to improve on that, but we are no worse and no better than any other country. And that is because Perth may be harder than Sydney for forecasting fog, for example. So you cannot take it from one country to another; it is really one location. In terms of Perth fog, we have had several workshops in the last 12 months on how to improve. Obviously, when you mention Jandakot or Perth, there is an element of it being over on the other side of the country, and you do not always have those alternates. And we do not want to be diverting en route, which is what we would do if there were unforecast fog.

We need to understand the probability of that forecast fog occurring in a 12-hour time frame. For our flights from Japan that arrive in the early morning, knowing about the fog at four o’clock in the morning is no good, because we are already up there. We need to have an understanding of the possibility of something occurring. If it does take us by surprise, which it has and it can do, then we simply divert en route—we end up in Darwin or wherever. However, that is not quite the aim of what we are trying to achieve.

If we get there and it is below minimums then we will divert early anyway, because it is better to divert early, tank up and then fly on. But, unfortunately, there is a high false-alarm rate with fog forecasting at any time. You may say it is forecast over six hours but it may only appear for half an hour. That is the issue. So we need to continually evolve products with the bureau that say, ‘We have a long-term forecast, or a probability forecast, and this is now what we are doing, in the now-casting sense,’ so that, if it is going to appear at three hours, then we have a high predictability of that actually occurring so we can do something.

Mr SYMON —I would like to switch questions completely now and ask you about timetabling and scheduling and how that is affected with long-term forecasting. Obviously, it is a question of resources as to how many hours you can keep a plane in the air per day, and you need to set your schedules quite some time in advance. How much of a factor is the long-range forecast in that side of the operation?

Mr Rennie —Let us look at long-haul operations—say, from the US to Australia. If we see that there is a high probability that a 747 is going to end up in Brisbane, for example, because there is a very, very good chance that fog will appear, then we know where flight crew are. The flight crew that actually end up in Brisbane, if they do end up in Brisbane, can elect to fly on, if they are within hours. They may be out of hours, which means they cannot, of course. If they are within hours then they may elect to fly on. If they are not, our crewing department will know and will have, on standby, a flight crew ready to operate that aeroplane.

Mr SYMON —I probably have not got my question across properly there. Let us say, for instance, you were thinking of setting up a new route from Sydney to—who knows where; let us say somewhere in Mozambique. At that stage, when you were thinking about that, the long-term forecast would, I would think, have to come in there to say, ‘How long is the flight going to take?’ ‘What are the conditions we expect over the next year?’ and ‘How will it affect our baseline?’ Surely that is looked at, at some level?

Mr Rennie —That is a good question. Within our department now, that is exactly what we do. Whenever we look at sectors—and we are continually looking at sectors, whether they be ‘what ifs’, or ‘proposed’, or whether we are actually going to do it—we have a level of sophistication. We have within our computer statistics on winds—12 months of statistical winds. We know where we can access climatology, which is what is going to happen in 12 months at that airport. There is a big drive now to look at the actual time of day.

Mr SYMON —So that is the fourth dimension again.

Mr Rennie —Yes. There have been instances, in the recent past, whereby we have said to the commercial area that it is not the best to operate into that airport at this time of day, because of high fog rates or whatever so that, from a commercial aspect, it is simply not viable. We look at that and then we have that picture out. We are going to Mozambique, so we understand the possibilities—what we are likely to receive 12 months a year. Then we ask: ‘What are the alternates?’ ‘What is their climatology?’ ‘What is the en route climatology?’ So then we turn all that into what we call load planning and we say to our commercial area, ‘This is the load for this aeroplane.’ They may say, ‘Let’s try it with a different type of aeroplane,’ and we will do exactly the same with that. We will come up with a load plan, so it is a route and a load plan for that service. They will then do a business case on whether or not Mozambique is viable.

Mr SYMON —I have one last question and it is only a short one. Do you share information with the Bureau of Meteorology? You obviously get information from them. Do you collect information for them as well?

Mr Rennie —Absolutely. Operationally, sharing of information is widespread. There are very few barriers to that.

Mr SYMON —Do you share it with other airlines?

Mr Rennie —Yes, and that is within Australia and globally. It is not a commercial aspect. There are some commercial aspects to what we do and we do not share those. That is obvious. But in terms of how we operate there is no barrier there.

Mr RAMSEY ——Graham, you spoke earlier about US$28 billion for the US industry in in-air wastage time, or something or other—I do not know how you described it but you know what I mean. I presume that is an annualised figure, is it?

Mr Rennie —Yes, it is. There were other numbers that were produced by the FAA which came up with higher numbers because they actually broke them up into the effect to the airlines, the passengers and other factors. It came up to about US$42 million.

CHAIR —Million or billion?

Mr Rennie —Sorry, billion. That is the second time I did that.

CHAIR —It is just that the difference is—

Mr Rennie —I cannot comprehend those numbers! They actually broke it up. It was a greater number. So $28 billion was a reference to the loss for airlines. There are many factors: your inconvenience, for example. If you could not make it today with your flight in, what was that inconvenience to you? How do you put a dollar on that? That is the hard bit and that is the number that is not really in there. It is really about fuel, crew costs and CO2 as well.

CHAIR —Does it have a relationship to the price of air fares?

Mr Rennie —I do not know.

CHAIR —I was wondering. If you develop efficient forecasting systems and less fuel, you might—

Mr Rennie —It is a cost-based business, absolutely. Our call is all about meteorological accuracy. If we were 100 per cent accurate every day in forecasting around the world, I can assure you our fuel bill would be hugely different.

Mr RAMSEY —One of the primary focuses of this inquiry is seasonal forecasting. I am wondering whether there is any advantage to Qantas in having better seasonal forecasts. We understand it affects the tourist industry—a wet English summer might stuff up Wimbledon and the Ashes, for instance—but is there any advantage in Qantas actually having better seasonal forecasts or does Qantas just keep filling up the jets and then, when one does not fill, do a cancellation?

Mr Rennie —I guess, really, if you have a schedule then you have to go to that schedule, regardless. What we are about operationally is to ensure that in safety and efficiency we do the best we can. I think all organisations these days are getting smarter, in terms of what you are saying there. There is so much intelligence that floats around. For example, in our industry, in terms of seasonal forecasting, we have cyclone forecasts and we know each season for the Pacific what the forecast is for the cyclones for that season, for the next 12 months. I look at that and we pass that information around. We discuss it. We say, ‘Okay, we can expect more in Australia, fewer in Australia, more in the western Pacific, more in Japan or more in Hong Kong’—whatever. That allows us then to be better with our planning. These days everything we do is about planning. In terms of the strategic planning, it allows us to have that in place.

Mr RAMSEY —Effectively, what would that mean? If you know that you are going to have more cyclones down the east coast of Australia, it is all well and good to say you know there are going to be more, but you do not know what dates they are going to be on. How would that change what you do in an operational sense?

Mr Rennie —If we were going to have, for example, an increase in cyclones in the western Pacific, which is basically an ordinary scenario for us, we would either put in extra routes or we would do exactly what we have just recently done, and that is to introduce user preferred routes into that area. In terms of air services, Qantas and Australia were at the forefront of developing the current air traffic management system. That allows us in the Pacific and Indian Oceans—and soon in Australia and the Tasman—to use user preferred routes. Right now, when you go from Sydney to Perth, you go on an airway. There is a chart that has all these lines, but with user preferred routes the lines disappear, fundamentally. We can operate where we want to operate in terms of the winds and the restricted areas within the normal operational boundaries. So we can now optimise our flying, reduce our fuel and so forth. If we are going to have more cyclones we need to get smarter on how we plan for cyclones. So we are making improvements to our planning system and saying: ‘We’re going to have more cyclones within a certain time frame. How do we operate around that?’ We are now doing mathematical calculations in terms of free flight to actually motor around that type of environment much better. It is the same for volcanic ash.

Mr RAMSEY —Does that have any implications for safety?

Mr Rennie —It improves it. That is what we are all about.

Mr RAMSEY —I mean there must be a compromise somewhere in abandoning the traditional flight paths, where we know the traffic is.

Mr Rennie —Because user preferred routes are at random, it is in fact safer in that sense. We have operated user preferred routes across the Pacific, from the US to Australia, and from Sydney to Buenos Aires and South Africa, for quite a few years now. If you are talking about randomness, it has allowed the separation standard to be reduced to levels that are really world beaters and world leaders. That is the benefit of the current air traffic management system, which was tested in the Pacific throughout the nineties.

Mr RAMSEY —So why have we persisted so long with the traditional flight paths?

Mr Rennie —Because of traffic. Between Sydney and Melbourne we will always have a line. There is no benefit in using user preferred routes between Sydney and Melbourne, because if you were to random-route between Sydney and Melbourne you are almost certainly going to come up with the same line. When we operate from the east coast to the west coast, Broome or Alice Springs, we will have a user preferred route, and Airservices or ATC will agree to that. We will file it and they agree or they disagree. From an air traffic management perspective, the other aspect of that is that we can now do dynamic airborne re-routing in the Pacific area, whereby as soon as we get a wind update we can improve the flight. We simply re-run that flight plan from a certain position and uplink that straight into the flight management system of the aircraft, and then we can operate that route on an improved basis. If that tropical cyclone slows down or speeds up, we can motor around it again.

CHAIR —That was very interesting. Thank you very much.

[9.35 am]