Home » Climate Change
Category Archives: Climate Change
This article was originally published on my LinkedIn profile.
I’ve seen plenty of coverage of the recent Climate Council report on the likely increase in bushfire losses. But in reality these growing costs could be far higher and even more unequally distributed.
The figures for economic losses are based primarily around insured losses (which in economic terms aren’t really a loss, assuming that the bushfire risk has been priced correctly), and this assumes that in the future most properties will still be insured against bushfire. But the larger conflagrations brought on by climate change as well as the closer coupling of disaster events (the longer fire seasons will make it more likely that multiple large bushfire disasters happen in any one year) will increase the costs for insurers and make it more likely that any one bushfire season could bankrupt them. In theory, the larger the loss in the worst case scenario, the higher premiums need to be in order for insurance companies to remain solvent. In practice, insurance companies push up premiums after large disaster events, even for different hazards. Together these mean that as climate change increases bushfire risk, in particular the risk of catastrophic outcomes, insurers will charge ever higher premiums and some may withdraw from the market entirely. This will lead to far fewer people being covered when a bushfire disaster strikes, with the government and charities being left to pick up the slack.
Perhaps it’s time then that the Federal Government revisited the idea of a national disaster insurance scheme – to ensure that future generations can meet the costs of these disasters.
The reborn Climate Council has recently put out a new report on the projected impact of climate change on bushfires. Most of the content is backed up by sound research and pretty decent science, but some of the media coverage is focussing on a supposed doubling in the need for fire-fighters by 2030.
I’ve long been wary of claims that climate change requires massive investment in the emergency services – the problem of increasing bushfire or flood risks are better dealt with using disaster mitigation measures like land-use planning, building controls
Let’s unpack this claim a bit. The executive summary of the Climate Council report says it as thus:
By 2030, it has been estimated that the number of professional firefighters will need to approximately double (compared to 2010) to keep pace with increased population, asset value, and fire danger weather.
This was sourced from a 2013 report titled Firefighters and climate change: The human resources dimension of adapting to climate change prepared by the National Institute for Economic and Industry Research for the United Firefighters Union (I’m going to leave aside the obvious question of bias here and stick to the content).
It focusses on full-time firefighters of which there are about 11,500 in Australia in addition to the FTE of 1700 part-time firefighters and 220,000 volunteer firefighters. It’s these 11,500* who are counted as professional firefighters in the report.
I’m going to focus on the projections for NSW for time constraints.
In its future projection of fire-fighting resource requirement the NIEIR report examines two drivers:
- Population and asset growth
- Increased response due to climate change
First, population and asset growth. Here’s the projections of firefighter numbers that the report says would be required to maintain existing levels of firefighter cover of population and assets in NSW.
|NSW Firefighter projections(NIEIR)||3604||3999||4424||4706||4991|
This is an annual growth rate of 2.1-2.2% to 2020 and 1.2-1.3% between 2020 and 2030.
But the population of NSW is not growing this fast. Over the next 20 years the ABS estimates that population growth will average somewhere between 0.9% and 1.3%. So that leaves the rest of the projected growth due to increases in asset value. But does asset value make sense as a basis for estimating fire cover?
Not really – assets are property and a property will increase in value over time (due to inflation, rising replacement costs, increasing land value etc.) even though the property itself (and thus the number of firefighters needed to put it out if it’s on fire) remains unchanged. Population growth should be able to account for the increasing physical units of stuff (houses, businesses, vehicles etc.) that firefighters provide protection to. Maybe, though there’s other drivers?
So let’s look at responses by Fire and Rescue NSW over the last 15 years. The following chart shows their activities per 100,000 of population (culled from their annual reports)
The first thing we notice is that fires make up a relatively small proportion of the overall activity of Fire and Rescue NSW (and this is similar to other urban fire organisations across the country). The second thing we note is that the total number of incidents as a proportion of population is unchanged over the last 15 years. That is – population alone can explain any increase in the activity of Fire and Rescue NSW.
One thing that might be harder to see on the graph is that the number of fire incidents is actually going down. This is particularly the case with structure fires which have decreased by 32% over the last 15 years, per 100,000 population. Other fires have also dropped, by 42% – though the reduction in fire response has been offset by increases in non-fire rescue (57%) and Hazmat incidents** (20%). The increase in non-fire rescues is likely due to expansion in the land rescue areas that FRNSW is responsible for.
The decline in structure fires makes sense – we’re building safer buildings (code compliant, better electricals, fire safety measures etc.) and we’re also behaving safer (mostly because fewer people are smoking in bed). Landscape fires (bush and grass fires) make up about a third of the ‘Other Fires’ category, but I don’t have enough data to break this category down across a decent time series to identify any trends.
The Total Incidents tells the story though – there’s no basis for assessing growth in overall demand of full-time firefighters on anything but population. So on the basis of population growth the resource requirements for FRNSW full-time firefighters looks something like this:
|NSW Firefighter projections(population based)||3516***||3572||3807||4042||4230|
This produces a much more sensible growth rate.
Now let’s turn to climate change. Again here are the NIEIR projections for NSW, this time with added rows for a Low (H2) and High(H3) climate change scenario.
The NIEIR methodology hinges on a correlation they found between the number of Landscape fires in Victoria as set out in the 2012 Report on Government Services and the number of total fire ban days (both partial and statewide) declared in Victoria in any one fire season as enumerated in the CFA Annual Reports. By my count that’s a measly 5 data points.
Here’s the plot of the number of total fire ban days versus the number of landscape fires in Victoria.
There’s a reasonably strong positive relationship there (correlation coefficient ~0.9) – but it’s only 5 data points. NIEIR could really have done better than that. The response could be substantially more noisy than appears or even non-linear. They’ve also assumed that this relationship would hold everywhere.
NIEIR then uses an excellent report prepared in 2007 by the Bushfire CRC, CSIRO and the BoM for the Climate Institute which examined, amongst other things the expected change in days with FFDI>50 (on which Total Fire Bans are generally declared) under a variety of climate change scenarios.
Now here the outline of the methodology begins to get a bit shaky. They seem to have applied some sort of modifier to take into account fewer bushfires in urban and arid areas (which is sensible) and matched similar sites in western and northern Australia with those in the Lucas paper (which is not sensible – the climate regimes are unlikely to change in similar ways) to produce area estimates of the increase in firefighters due to both climate change and population/asset growth.
Now here comes the kicker – best as I can figure out (by reverse engineering the math) NIEIR has assumed that the increase would apply to all incident types responded to by professional fire fighters, not just the landscape fires that make up about 5-10% of all incidents responded to by full-time firefighters. Even a doubling in the number of Extreme fire danger days, which is likely in some but not all areas, would only increase the overall taskload of NSW full-time firefighters by 4-6%. This is actually within the annual variation in incident numbers in NSW, suggesting that on this crude measure it would be possible for additional response to climate change to be handled within population adjusted resourcing.
On the other hand the resource requirements of the volunteer fire services (where a much larger propotion of the taskload is made up of bushfires) could be more severely impacted. Unfortunately I don’t think that these type of analyses really provide much information for emergency managers to plan future resource requirements in response to climate change. The impact of climate change on emergency services will be most keenly felt in extreme events – where you’re more interested in surge capacity and where part-time firefighters and volunteers play a much larger role. A scenario approach to modelling resource requirements during these extremes would provide much more insight into what we really do need for the greater amount of extreme weather in the future.
In the meantime it sadly seems that the Climate Council lacks the resources to adequately check its sources, especially when they come from the gray literature. Here’s hoping that they can encourage some work to get the real answers on human resource requirements for extreme fires under climate change.
* The report also discusses some of the problems in counting how many full-time firefighters there are in Australia, it uses a figure of 12,041 which is calculated from Census responses. I’ve taken the figure from the Productivity Commission’s Report on Government Services – so this is an approximate figure.
** I’m not very confident about this figure – FRNSW have changed their way of categorising hazmat incidents over the years, so a small change here may have been offset by a small change in one of the other incident categories – probably other. I’ve applied all the usual statistical tests and these changes are significant – just as there is no significant change to the Other incidents category or the Total incidents overall.
*** Here I’ve used the number of full-time firefighters contained in the FRNSW Annual Report.
Climate change is a big topic in disaster management. I have earlier outlined that claims of a big impact on severe weather by climate change (at least presently) are largely overblown. As a broader risk management issue climate change is a big one.
If left unchecked, climate change could have some huge, civilisation altering, consequences over the next couple of hundred years. This is a big risk for humankind and possibly the largest over a timescale of 100-300 years.
Which is why it really gets my goat to hear claims that the warming has stopped.