BOM’s Climate Forecast Failure

Senator ROBERTS: Thank you to the bureau for appearing today. I’ve handed out some documents for circulation. They’re copies of BOM forecasts versus actual. I think you’ll be familiar with them, for sure. That’s a contradiction in terms, ‘think’ and ‘for sure’. Anyway, I’m sure you’ll be familiar with them.

On 19 September 2023 the Bureau of Meteorology’s weather forecast read, ‘Warmer and dryer conditions would be more likely over spring and summer,’ linking the Indian Ocean Dipole with El Nino using the words, ‘The last time this occurred was 2015,’ which was a very dry year, especially in Queensland. The bush listened to that, and a lot of other people did too—investors as well.

On 30 November the Bureau of Meteorology predicted ‘a high chance of warmer than usual days and nights across Australia, below average rainfall likely for much of the tropics’. The actual weather: northern Queensland was flooded in December—big floods—by Tropical Cyclone Jasper; inland Queensland was flooded in January by Tropical Cyclone Kirrily; South-East Queensland was flooded in December and January.

I’ve circulated your entire forecast for 2023 split into five periods. Each period forecast, except one, was for drier weather than occurred. One was about right. None predicted more rain than occurred, much less than I would have hoped. My question is simple: is your weather model fundamentally flawed?

Dr Johnson: No, Senator, it’s not.

Senator ROBERTS: It’s costing nearly $1 billion to upgrade your computer system, the ROBUST Program as it’s called. Is that still the cost, and can you provide an itemisation? It appears a ridiculous amount of money.

Dr Johnson: Firstly, as I’ve answered to senators in this chamber over many years, the costs associated with the ROBUST Program are cabinet in confidence; they’re not for publication. As I also answered—I think it was at the previous hearing or perhaps the one before—in response to a question from Senator Pocock, the upgrade of the Australis computer system is not part of the ROBUST Program; it is a separate program of work.

Senator ROBERTS:Could you explain the Australis versus the ROBUST, and which one is—

Dr Johnson: ROBUST is a complex program to upgrade the bureau’s ICT and observing systems, fundamental ICT—

Senator ROBERTS:What’s ICT?

Dr Johnson: Information and communications technology.

Senator ROBERTS:Thank you.

Dr Johnson: It upgrades our underlying information and communications technology infrastructure, our observing networks—all sorts of things.

Senator ROBERTS:And recording devices?

Dr Johnson: I’ll get to that in a second. That includes radars, automatic weather stations, automatic balloon launchers—all sorts of things that observe the environment—as well as our underpinning technology infrastructure.

The ROBUST Program, again, has three dimensions. It has a security dimension—in other words, investment to improve the security of the bureau’s systems from threats from our country’s adversaries. There’s stability. Prior to the investment in ROBUST, many of the bureau’s systems were very old, many decades old, and we were experiencing challenges in keeping them stable and operational. And then there’s resilience so that, in the event of an outage, the capacity of the bureau to respond and have our systems back online is improved. So there are three dimensions to ROBUST: security, stability and resilience.

There is a supercomputer dimension to ROBUST, which is a second supercomputer, a disaster recovery machine. Prior to ROBUST, our disaster recovery functions were executed within a single machine in a single place. The arrangements going forward will be different. I’d rather not disclose those in detail, for security reasons, but the ROBUST program funded a second supercomputer for disaster recovery purposes. That is a different machine to the Australis machine, which has often been asked about in Senator Pocock’s questions. That was a separate program to Robust, Senator. You’re conflating two bits of technology uplift in two separate programs.

Senator ROBERTS:The total cost is a billion dollars for both?

Dr Johnson: No. As I said, I’m not going to speak about the cost of ROBUST. The cost of the Australis upgrade is roughly, I think, $44 million—something of that order.

Senator ROBERTS:Dr Johnson, you’re required to produce any information or documents that are requested to this committee. There’s no privacy, security, freedom of information or other legislation that overrides this Senate committee’s constitutional powers to gather evidence. You’re protected from any potential prosecution as a result of your evidence or producing documents to this committee. If anyone seeks to pressure you against producing documents, that’s also a contempt. If you wish to raise an immunity claim, there are proper processes around that, and it is up to the Senate whether to accept that, not you or the minister. Can you please take on notice to produce that document to the committee and the cost—

Dr Johnson: Which document are you referring to, Senator?

Senator ROBERTS: The cost.

Dr Johnson: Of ROBUST?

Senator ROBERTS: Yes.

Dr Johnson: The decision around the cost of ROBUST—

Senator McAllister: Just take it on notice.

Dr Johnson: We’ll take it on notice.

Senator ROBERTS:Thank you, Minister. David Burton of Inigo Jones, long-term weather forecaster, uses sunspots and planetary cycles to correctly forecast weather decades in advance. He’s got a track record because he’s got investors who invest as a result of his successful forecast. He posted 12 months ago that the good rains would start after 20 November 2023. There was no El Nino, and cyclones were likely. David Burton has no computer models and uses a $20 calculator because he understands the cycles. He got the weather right; you got it wrong. Hayden Walker, another long-term weather forecaster, correctly forecasted severe storms in the areas where they did occur. Will you talk to these private forecasters to work out why their systems are right and yours was wrong? Theirs are actually history. Yours are models—aren’t they?

Dr Johnson: I reject the whole premise of your questions. Our forecasts, as I’ve indicated at the previous hearing on this subject, were remarkably accurate. I’m happy to go through them again. What we said is on the Hansard in terms of the seasonal outlook. We were very clear, as the year progressed, that we were moving out of a dry warming trend into a moistening trend. We were also very clear in our messaging that, irrespective of the ENSO status and the seasonal forecast, we know that in northern Australia, in particular, there is always the risk of severe weather—cyclones and floods—under any climatic situation. I don’t agree with the premise of your question.

Senator ROBERTS: This is not just northern Queensland, where we know that it’s prone to storms, but western Queensland and southern Queensland. We know that your bureau declares El Nino and positive Indian Ocean Dipole events. David Burton said there was no El Nino and cyclones were likely. David Burton quite often gets it right. He’s paid a considerable amount of money because David Burton’s, Hayden Walker’s and, prior to them, Inigo Jones’ and various other people’s methods have been in use for decades. Farmers, investors and businesses pay for their forecast. They have to go out into the market and sell.

Dr Johnson: I understand that, and millions of Australians rely on our forecasts every day, including farmers and folks in the business community. I just reaffirm to you, as I did at the previous estimates, just how remarkably accurate our forecasts were over the period. I’ve certainly said in previous hearings and in other forums that we acknowledge that some of the messaging that we gave during the previous spring and summer didn’t get through in a manner that we would like.

That’s not to blame the recipients of that messaging. It’s just a fact. People heard a message around an ENSO status and thought, ‘That’s it; it’s going to be hot and dry.’ We update our forecasts every week, and we regularly updated our outlooks, and those outlooks proved to be very accurate.

We also affirmed, in all those messages that, particularly during the summer and irrespective of the ENSO status, the risk in this country of thunderstorms, floods and tropical cyclones remains. In fact, at the national severe weather forum here in Canberra, I made that very clear in my own presentation: one thing that this country has taught us is that severe weather can occur at any time. We’re very clear in our messaging around that. I’m only going to comment on our forecasts and warnings. Others are welcome to comment on those made other parties, but I stand by the quality of our forecast. I did so at the previous hearing, and I’ll continue to do so.

Senator ROBERTS:Well I do agree with you—

CHAIR: Senator Roberts, we’re going to have to rotate the call.

Senator ROBERTS:I note that your own forecast—

Dr Johnson: I’d like to have a look at this—

Senator ROBERTS: That’s produced by you.

Dr Johnson: I know, but a very quick and cursory—and maybe it’s not helpful, Chair—is that you’re comparing two different measures of data there. One’s talking about chance of exceedance and another one talks about actuals. The two are fundamentally different concepts. Just because you have a map of Australia with colours doesn’t mean to say that the two datasets are comparable. Let me have a look at it. If you have a specific question, I’d be happy to take it on notice.

Senator ROBERTS:Your faulty forecast caused farmers to offload cattle. The benchmark Eastern States Young Cattle Indicator sat as high as 1,192 cents per kilogram of carcass weight in 2022, but by late 2023—after your forecast hit the streets—it had tumbled all the way to 349 cents per kilogram. That’s less than a third. Do you accept responsibility for that loss to the Australian capital market as graziers sold stock because they feared overstocking in the looming dry?

Dr Johnson: Let’s just get the data right here. The Eastern States Young Cattle Indicator was about 1,200 cents per kilogram in January 22, and less than 400 cents by October 23. We released our El Nino declaration on 18 September, so by 18 September that particular measure of cattle prices had dropped somewhere by around 80 per cent. To somehow draw a conclusion that because we issued that declaration on 18 September that resulted in a run on the cattle markets, is just not supported by the data.

As I said at my previous hearings on this matter—in response to, I think, a question from Senator Davey—there are a whole range of factors influencing the decisions of primary producers. I’ve talked to a lot of primary producers and absolutely our advice and our outlooks contribute to their decision-making. But to a person to a farm business that I’ve spoken to there are a whole range of other things that they take into account in making a decision to sell their cattle. So this notion that the bureau declaring an El Nino at 18 September can explain an 80 per cent drop of the cattle prices from January 22 to October 23 is just a nonsense, frankly.

Senator ROBERTS:Thank you, Chair, can we come back?

— *** —

Senator ROBERTS:Dr Johnson, could you please repeat your dates and cattle prices? Eastern Young Cattle Indicator—

Dr Johnson: I’ll try if I can. Let me just find the brief that I have and the advice I have received. I’ve been advised of Rural Bank data that shows the Easten Young Cattle Indicator declined from approximately 1,200c a kilogram in January 2022 to less than 400c a kilogram by October 2023. As I said, we declared an El Nino on 18 September. So, just to reaffirm by that calculation, I’m advised that the cattle price had completed more than 80 per cent of its downward run by the time we declared that El Nino in September.

Senator ROBERTS:Great. Thank you very much.

Dr Johnson: That’s the advice I have.

Senator ROBERTS: The Financial Review blames the Bureau of Meteorology: ‘How the BOM’s big dry weather forecast cost millions’ and ‘Bureau of Meteorology’s botched weather call crushes Elders’ earnings’. There’s ‘BOM mistakes hit farmers but slash inflation’, and then we’ve got others there. Your botched prediction cost more than just farmers; it cost mum-and-dad investors in Elders millions, with the share price dropping 25 per cent. Do you accept that this was the fault of your forecast?

Dr Johnson: Again, I’ve already answered this question. We absolutely stand by our forecast. Our forecasts are remarkably accurate. As I’ve said at previous estimates hearings, commentary in the media, frankly, has been largely ill-informed and inaccurate, and we’ve sought to correct the record where we can. Take, for example, the Australian Financial Review article which asserted that our El Nino declaration had been linked to Elders’ earning advice. I’ve had a look at the Elders’ advice, and it was not stated or even implied in their earnings outlook. I don’t care what the Australian Financial Review reported. My reading of what Elders actually said was that it didn’t state or even imply that the bureau’s El Nino declaration affected earnings for the period 1 October to 30 September. There are lots of things written in the media. Again, we talked about this last time. If you actually have a look at the facts of what we said and when we said it, our forecasts were remarkably accurate given how complex it is and the sheer area that we’re seeking to provide forecasts for. The forecasts are not perfect; they will always contain uncertainty.

Also, the long-range forecasts can’t explicitly predict the emergence of cyclones—individual, specific events. There will be times when you’ll have an anomalous specific event, and hence why, in our public commentary, we seek to affirm to the public and to industry that there’s always the risk of severe weather in this country and there is always the risk of cyclones in the tropics and subtropics in this country. It is a forecast. It is an estimate of a point in time in the future based on the data at a particular point in time. We update it every week. I strongly encourage those who follow our services—and many millions of people do—to continue to check those updates. The situation changes all the time. We continued to update a point-in-time statement back in September as more information came to hand.

CHAIR: Senator Roberts, we are going to wrap up, so can you please make this your last question.

Senator ROBERTS: There are lots of people who are saying that the Bureau of Meteorology colours its diagrams to make it look hot and dry but we actually see—and this is rainfall over 124 years—no pattern or trend, no declining trend or increasing trend, just natural variation. And that’s from the BOM. Why the doom and gloom? Why depress expectations for rural output, which also depresses investment, training and employment in the bush, reduces the standard of living and increases the cost of living?

Dr Johnson: I’m not sure I understand your question, Senator.

Senator ROBERTS:Why are you so negative and preaching fear and doom when there’s nothing to suggest that, and why do you use colours to exaggerate it?

Dr Johnson: I don’t agree with your statement that we are preaching doom and gloom. We’re simply reporting the observations we’re making of the environment around us, and we’re reporting, to the best of our ability, what our guidance is for the time ahead. We do that objectively, using world-class, internationally peer reviewed, highly regarded scientific methods, and we’ll continue to do so. I think we’re entirely objective in our pronunciations and our public statements.