Comparison Between Observational Data and Model Projections for Cairns (AU) Hot Days

Guest essay by Dr. B Basil Beamish

On Friday 29 October 2010, The Cairns Post published an article based on a Queensland Government report entitled “Climate change in Queensland: What the science is telling us” (QDERM, 2010). The article referred to predicted impacts of climate change affecting the Far North Queensland region.

One of the statements contained in the article was: “The number of days over 35C in Cairns is expected to triple and the Gulf and Cape can expect longer drier spells interrupted by more intense rainfall.” The reference to tripling was based on an assumed high emissions scenario by the year 2050.

Since the release of this report five more years of data have become available and it is worth comparing the actual trend in hot days for Cairns with respect to the model projections as given in the QDERM report.

QDERM (2010) presented a table on the projected increase in the average number of hot days per year for several regional centres across Queensland (Figure 19, page 29). The model projections for Cairns are shown in Table 1 and are based on three different climate model scenarios: (1) 2030 with a medium emissions scenario; (2) 2050 with a low emissions scenario; and (3) 2050 with a high emissions scenario. The values in the brackets are the range for each scenario projection (10^th and 90^th percentiles).

Figure 1 shows the trend in hot days for Cairns since 1910 to 2015 using the publicly available ACORN-SAT data from the Australian Bureau of Meteorology. The emission scenario projections contained in Table 1 are also shown in Figure 1 for reference. A 30-year moving average has been added to the graph to illustrate how the average number of hot days is trending from a climate perspective.

Figure 1. Trend in the number of hot days for Cairns from 1910-2010 and model projections from the QDERM report for the average number of hot days by 2030 and 2050.

As 2030 will be the first of the projection dates to be reached it provides an opportunity to look at the number of hot days required to reach the model projections for that year. For example, the average number of hot days required to be maintained from 2011-2030 to reach the median value can be calculated as follows:

Projected average number of hot days by 2030 (medium emissions) = 6

Therefore, the number of hot days projected for 2001-2030 is:

6 x 30 = 180

The number of hot days recorded from 2001-2010 is:

(2 + 4 + 1 + 1 + 2 + 1 + 3 + 4 + 0 + 3) = 21

Therefore, the number of hot days from 2011-2030 required to reach the model projection is:

180 – 21 = 159

Average number of hot days required each year from 2011-2030 is:

159/20 = 7.95

Similar calculations are used for the 10^th and 90^th percentile values and the corresponding results are shown in Table 1. It should be noted that there has never been a sustained period of hot days greater than or equal to 6 in the entire Cairns temperature record. 30-year moving averages have been added to Figure 2 for the 2030 medium emissions scenario for each of the median value and percentile levels based on the average number of hot days required every year from 2011 onwards to achieve the model projection value for each of these parameters.

Figure 2. Trend in the number of hot days for Cairns from 2011-2030 required to achieve the model projections contained in the QDERM report.

In the years from 2011-2015 Cairns has experienced the following number of hot days, 2011 – 1; 2012 – 0; 2013 – 8; 2014 – 0; and 2015 – 5. This is an average of 2.8 hot days per year, which is well below that required to reach the model projection for 2030. This is visibly apparent in Figure 2 as the observed 30-year moving average has already deviated significantly below the 10^th percentile line.

A fourth option (“No change”) has been added to Figure 2 for comparison, which assumes that the number of hot days for Cairns from 2016-2030 will not be any different to the number of hot days from 2001-2015. If this eventuates then the projected 30-year moving average returns to a level similar to the late 1960’s and deviates even further from the model projections.

On Wednesday 28 January 2015 another article was published in the Cairns Post that stated: “CAIRNS may no longer be a comfortable place to live by the end of the century with the amount of days where the temperature rises above 35C set to increase threefold.”

This was based on the release of a joint CSIRO and Bureau of Meteorology report that contains projections of the Far North’s climate by the end of the 21^st Century (McInnes et al, 2015). It would seem that the threefold increase in the number of hot days has now been extended by 40 more years to 2090 as a result of these latest model projections (Table 2).

Results of applying the same analysis to the CSIRO/BOM model projections as for the QDERM model projections for 2030 are shown in Figure 3. In this case the time starts from 2016 not 2011 and the average number of hot days required from 2016-2030 to reach the projected value is 8.67 with a range from 6.47 to 13.47. The increase in the number of hot days to reach the model projections is higher compared to the QDERM model projections due to the shorter timeframe to 2030.

Figure 3. Trend in the number of hot days for Cairns from 2016-2030 required to achieve the model projections contained in the CSIRO/BOM report.

Over the past 105 years there has been no consecutive 15-year period where the average number of hot days has reached 6.47. In fact the highest consecutive 15-year average is 4.73 for 1923-1937. Consequently, the probability of reaching the lower limit of the 2030 RCP4.5 model projection is practically 0. For each year that the number of hot days falls below the average number required to reach the model projection value, the number required in subsequent years will increase. If this happens it will reinforce that the model projection is unrealistic. Alternatively, if the “No change” option eventuates it would also prove the model projection to be invalid.

In conclusion, the following key points are made as a result of this comparison between observational data and model projections:

• Any model projection needs to be validated against actual observations before any confidence can be placed in the model results.

• Accountability of model projections should be mandatory and comparisons similar to this one should be performed on a five yearly basis, particularly if the models are continually adjusted.

• As each five years of data becomes available it should become readily apparent whether the model projections are either validated or not.

References

Queensland Department of Environment and Resource Management (QDERM), 2010. Climate change in Queensland: What the science is telling us, Queensland Climate Change Centre of Excellence, 93p.

McInnes, K et al, 2015. Wet Tropics Cluster Report, Climate Change in Australia Projections for Australia’s Natural Resource Management Regions: Cluster Reports, eds. Ekström, M et al, CSIRO and Bureau of Meteorology, Australia, 48p.