Dennis Jensen (Tangney, Liberal Party) Share this | Hansard source

At the outset I commend the member for Hughes on his speech. It was not planned that we would be playing a bit of a tag team. I will get onto the issue of data integrity a little later on, which the member for Hughes spoke a bit about. This amendment to allow commercial advertising on the Bureau of Meteorology website is, in the main, something that should be welcomed, as the member for Hughes has pointed out. The previous government saddled us with a massive debt that needs to be repaid. The interest on that debt is $2,000 for a family of four every single year—just interest repayments. So finding new sources of sustainable revenue for government and creative ways of monetising long-run sunk investment is something that I specially welcome.

However, I do have concerns with respect to the specific text of the bill. I am principally concerned about the level of safeguards and oversight in the bill. I feel that, while it is undoubtedly a good thing to get more money into the bureau, this bill gives too much power to the director of the bureau of met. This is not to say that the current director of the bureau of met would use this legislation as something like an enabling piece of legislation or as a tool for nefarious purposes. But, in simple terms, this bill gives the director the final call when it comes to what advertising gets up on the Bureau of Meteorology website. On the face of it, that seems to be fair enough.

However, there is some quite insidious wording in the bill. Indeed, when looking over the parliamentary secretary's second reading speech on the issue, and the honourable member for Riverina confirmed my fears, there is the phrase 'in the bureau's interest'. Regulatory capture, rent-seeking behaviours and soft-touch self-regulation are things that do not serve the community interest. I put it to members of this place that, instead of reading 'the bureau's interest', the bill should make it explicitly clear that it needs to be in the community interest.

Community interest may, and could, include groups of companies sceptical of climate change, which could be fairly described as a position that is not necessarily in the bureau's interest. For example, there are specific issues relating to data manipulation conducted by the Bureau of Meteorology that may be questioned by a body that chooses to advertise on the website. To explain: the data gathered from what are known as the ACORN-SAT sites, which are the designated sites from around Australia from which temperature and other data is collected—and these feed into Australian and global statistics—are processed using a term known as 'homogenisation'. It is these homogenised data, and not the raw data, that are then used for compiling the temperature data statistics. But is homogenised data a more correct representation of what is real than the raw data?

I went to the Bureau of Met in Melbourne to discuss this and some other issues. Based on this discussion, some interesting things came to light. These issues are very important when the Bureau of Meteorology starts talking about differences in records in mere hundredths of a degree as being significant. Take the case of thermometers. When an ACORN-SAT site is checked, the thermometer that is at that site is checked against a calibrated thermometer. If it is out by 0.3 degrees Centigrade it is still designated as fit for purpose—an error which is much greater than the Bureau of Meteorology is claiming as significant.

The homogenisation itself has a trend. I asked the Bureau of Meteorology for the homogenised and raw data trends for each ACORN-SAT site. I also asked for the physical area that they covered to be able to apply a weighting factor for each site. I got the former, but not the latter. So assuming equal weighting, which was the only thing that I could do, the homogenisation trend would be about 0.3 degrees Centigrade for the last 100 years. In other words, if the raw data showed no warming whatsoever over the last century, the homogenised data would indicate that Australia is 0.3 degrees hotter today than it was 100 years ago. This homogenised data is generated using an algorithm comparing data from the ACORN-SAT site with sites that may be 1,000 kilometres away.

I looked at the specific case of Bourke and questioned them on the homogenisation used there. In particular, in the early and the mid-1950s there was a homogenisation in the temperature data increasing the temperature by around half a degree. In a very short space of time it is almost a straight line vertical of approximately half a degree. The problem is the Bureau of Met was unable to explain physically why this adjustment was made; it was just that the algorithm had done its thing. As a scientist, I would say that the first thing that should be asked is: 'What physical factor caused this?' If none can be found, the raw data should be used in preference to the processed data. The lack of curiosity—for want of other terminology—is very concerning. This indicates a predisposition to simply trust the computer models rather than accept the data. This is a problem because, after all, there are assumptions built into the computer models.

On a slight tangent, this is why I have called for an audit of the Bureau of Meteorology by the Bureau of Statistics. I was very concerned to find that there is only one formally trained statistician in the Bureau of Meteorology. The data and the statistical handling of that data is critical. An audit of the Bureau of Met's handling of the data and the methodologies that they use is very important. How would the Director of the Bureau of Met respond to advertising by a group that sought to point out and contest or contradict the Bureau of Meteorology's data and data handling?

A division having been called in the House of Representatives—

Sitting suspended from 16:20 to 16:39

I have no doubt that they would determine that the advertising would be considered not in the bureau's interest. They could also determine that said advertising was misleading or deceptive. The problem with this is that the definitions, as spelled out in the explanatory memorandum, are too broad. What may not be in the Bureau of Meteorology's interest may very much be in the Commonwealth or national interest. The problem with the definitions as they stand is there appears to be an implicit assumption that what is in the Bureau of Meteorology's interest is also in the Commonwealth's interest and vice versa. This is not necessarily the case and this aspect needs some rethinking to ensure that it does not, by default of the potential bias of the organisation, become the antithesis of another point, which is advertising of a political nature. In other words, the decision not to allow some advertising may, in effect, become political in itself.

This bill originates from the 2011 Munro review whose modus operandi was to find new ways for the bureau to respond to extreme weather events in the future. New revenue sources could go a long way to address that goal. New money could be put to greater investments in IT and data collection systems. Moreover, new money could go a long way if invested in communications infrastructure. As with most things in government, finding a home for new money is never much of a problem. So, yes, I agree that with the move to give clarity to the Bureau of Meteorology regarding advertising on its website. This move is overdue and should be replicated by other government agencies and departments. Information, viewer attention, brand recognition and click loyalty are all very valuable resources that many government departments and agencies own but do not currently utilise.

In the current economic situation, it is welcome and necessary to see greater innovation and utilisation of capital. I am gravely concerned regarding the provision that is given to the director of the Bureau of Meteorology to reject advertising that causes offence.

This bill is in many ways too prescriptive as to what cannot be included on the website or what groups can advertise on the Bureau of Meteorology's website. I am of the belief that we should open this opportunity up to as wide a potential pool as possible. The one caveat that I would have is around protecting the integrity and good name of the bureau and, consequently, the Australian government. The same belief goes for protecting scientific integrity. Scientific integrity, to my mind, means recognising other and all credible scientific voices in the community. This means recognising the legitimate concerns many in the community have regarding the Bureau of Meteorology's upward homogenisation of its datasets.

Perhaps one improvement to this bill would be a community interest charter that could inform how a director makes his or her call on what advertising gets up. In turn, if done correctly—in other words, through wide-ranging consultation with community stakeholders—this charter could then be rolled out to other government websites. This would save a great deal of time, money and energy in the long run as well as give certainty to future direction.

See this speech in context