Disaster Impact Analysis - Whitsunday and Rockhampton
In early 2017, Cyclone Debbie devastated parts of the Whitsundays in Queensland and disrupted the local economy. The actual scale of that economic disruption has been reported on through anecdote, and in economic models that are, at best, an educated guess. Moreover, we have few options for verifying the numbers our models generate. Now, though, we can use real, not modelled, data to more accurately quantify this disruption.
Climate change is exacerbating natural disasters
The science is complex, but the concept it quite simple: climate change is expected to make natural disasters more common and more extreme. That means more, or more extreme, flooding, heat waves, droughts, storms and bush fires. In Australia, we have experienced the eight of the hottest years on record in the last 15 years; extreme bushfires occur more frequently; heatwaves are expected to increase in number and severity; and we will see less rainfall. The result is an expected tripling in the cost of Australia’s natural disaster bill to $33 billion by 2050.
All of this has implications for infrastructure, water management, agriculture, biodiversity, housing, health and even political stability. In this setting, then, the ability to quantify the impacts of extreme events is invaluable for forward planning and mitigation. That’s where ‘big data’ can help.
Cyclone Debbie hits the local economy in surprising ways
By looking at bank transaction data for the Whitsunday council area before, during and after Cyclone Debbie, we can see where spending changed in the region.
But first, you need to know two things about the economy in the Whitsunday region: firstly, tourism is central to the economy; and secondly, tourism is very seasonal.
Figure 1 shows how visitor spending on food, dining and entertainment dropped when Debbie hit and then stayed flat over the next few weeks. Interestingly, it didn’t look like it exactly collapsed. If we look in the previous year, we see the bottom of the trough in visitor spend is roughly the same, with or without the Cyclone. But what Debbie did appear to do is first delay the new tourist season in the late April to early May period. Then, and perhaps more interestingly, we see an enormous spike in spend well above anything we see in the previous year.
We asked the local council and while they didn’t have the numbers we had access to, they knew what caused the spike. A quick look online and we could see as well: tradies working on the clean-up and reconstruction (Figure 2).
In summary, the decline in visitor spend was less than expected, but the rapid spike represented a significantly higher spend than at the same time the previous year. In food, dining and entertainment alone, it was several million dollars per month at its peak. It will be interesting to see whether this activity is maintained.
For example, major accommodation establishments such as Hayman and Daydream Islands, have taken this opportunity to bring forward their major redevelopment and refurbishment plans. If this disaster is taken as an opportunity to redevelop the local area, greater sustainable activity may be possible beyond the temporary construction boom — even if construction is scheduled to continue for years.
A few weeks after Debbie, the rain generated by the Cyclone caused flooding in Rockhampton further south. Again, we can use bank transaction data through Spendmapp to track the effects of this. This time, rather than monthly visitor spend data, we look at total daily spend, but just on grocery stores and supermarkets. Here we see several features.
1. April 1. An initial spike of $1.2m in grocery stores and supermarkets expenditure. This can probably be explained by people stocking up on food and supplies.
2. April 4. The floods hit and spending drops to $600,000. It stays this low for the best part of a week.
3. April 13. Spending spikes (to $1.7m). As with Whitsunday, this is also likely to be clean-up and reconstruction efforts.
These trends all make intuitive sense. But until now, we had to rely on simulated and modelled data to estimate their true magnitude. Now, with the aid of real bank transaction data, we can see the exact economic impact of such events, and better predict and plan for future impacts.
I think for the next blog, we’ll look into measuring the total impact of multiple, major natural disasters (cyclone, flood and fire) across Australia over a 12-month period. Stay tuned.