JoNova » Australia has lowest number of tropical cyclones in 1500 years?

A new paper suggests there is an “unprecedentedly” low number of tropical cyclones around Australia at the moment. (How much should we spend to avoid this dreadful outcome I wonder?)

I am a little skeptical of how we can be so sure of the cyclone activity in, say, the year 900 AD. But nonetheless, the study is worth a look. Haig et al took stalagmites from two places in Australia (Chillagoe, Qld, and Cape Range, WA) and got very nice long year-by-year records of ¹⁸O and ¹⁶O data. They calibrated these against observational instrumental records — though I note these are but a tiny 20 years of data (1990 – 2010), and that during a period described by mainstream climate science (cough) as “unprecedented”.

Assuming that it is possible to pick apart normal rain and cyclonic rain, and that cyclone activity did not just shift to be more than 400 km away (where these stalagmites won’t record the cyclones) then it does appear that there are usually more cyclones in Australia than now. Note the top graphs are the WA site which go back to 500AD, and the lower pair are the QLD graphs “only” going back to 1300AD. Both graphs show an increase in storm activity during the Little Ice Age around 1700AD. If things do get cold in the next 20 years, will we get more cyclones and will fans of man-made global warming still be blaming CO2 if we do?

Figure 3 | CAI over the last 1,500 and 700 years. a, c, Cape Range (a) and Chillagoe (c); black line indicates smoothing of the series using ref. 31 (smoothed data were not used in the statistical analysis). Grey shading indicates the r.m.s.e. of the model. Four values, which were more than 1 s.d. outside the d18OA range specified in Fig. 1, were removed from the series. b, d, Wavelet power spectra (Morlet wavelet) of Cape Range (b) and Chillagoe (d). Power increases from blue to red, black contours indicate regions above the 95% confidence level, and the white areas are regions subject to edge effects. The spectra have lag-1 autocorrelation coefficients of 0.75 (Cape Range) and 0.78 (Chillagoe). Software provided by C. Torrence and G. Compo (http://
atoc.colorado.edu/research/wavelets/)..

The paper has the usual obligatory vague claims pretending to connect this to anthropogenic causes, when there is nothing definitively unnatural shown in this long study, and given that the Earth has 1500 year and longer cycles, the long data is still not long enough anyway. We all know the debate is over and the science is settled, but note the 95% certainty is not quite evident in the conclusions (see my bolding).

The Australian region seems to be experiencing the most pronounced
phase of tropical cyclone inactivity for the past 550–1,500 years. The
dramatic reductions in activity since the industrial revolution suggest
that climate change cannot be ruled out as a causative factor. This
reduction is also in line with present projections for the late twenty first
century from global climate models, yet our results suggest that
this is occurring much sooner than expected. However, we cannot say
whether this downward trend in activity will be sustained.

These waffly phrasings of uncertainty will be studied one day as a comment on the politically distorted scientific culture of the early 21st century.

There are a couple of caveats, one is that normal rain and cyclonic rain are difficult to tell apart. The other is that these stalagmites can only record cyclones within 400km of the cave.

Tropical cyclones produce precipitation that is depleted in the heavier
oxygen isotope (¹⁸O) by >6%relative to average monsoonal precipitation,
owing to the recycling of water within the system, high
condensation efficiency and large size and longevity of such cyclones
as intense convective systems17. The resulting d¹⁸O content (expressed
as d¹⁸O5[(¹⁸O/¹⁶O )sample/(¹⁸O/¹⁶O )standard21)] x 1,000%) of tropical
cyclone precipitation at a site is influenced by a number of factors,
including the number of days since genesis (that is, rainout) and the
intensity of the storm, its source region18 and the distance of its centre
from the sampling path. Because tropical stalagmites are archives of
monsoonal d¹⁸O, signatures of past tropical cyclones are also recorded
within the d¹⁸Oof their carbonate layers, typically within 400km of the
storm centre19,20.

Figure 1 below shows a map of Australia with the two sites marked. The circles around the two centres are 400km radius, so all of the cyclones that track beyond those circles are not included. During the last 20 years, the two circles are very well located, but I wonder if weather patterns shift and cyclones simply move elsewhere. The black squiggly lines are the tracks of cyclones that wouldn’t register in these caves during the last 20 years.

Figure 1 | Site mapshowing the four-stage calculation of CAI. Chillagoe and Cape Range (black points) are shown with the 400-km radius around each study site. Tropical cyclones whose tracks did not lie within the study area
during the training period in Queensland and Western Australia are shown in black. Red shading indicates the coastlines most prone to tropical cyclones in both states. a, Tropical cyclones from the 1990–2010 training period and their corresponding Kt value (point size), showing the influence of Vmax, Rmax and distance; cumulative Kn,t values are shown in colour. b, Point size indicates Kn (individual storm averages) calculated from a and subsequent seasonal CAI values (gradated colour).

ABSTRACT

The assessment of changes in tropical cyclone activity within the context of anthropogenically influenced climate change has been limited by the short temporal resolution of the instrumental tropical cyclone record^{1, 2} (less than 50 years). Furthermore, controversy exists regarding the robustness of the observational record, especially before 1990^{3, 4, 5}. Here we show, on the basis of a new tropical cyclone activity index (CAI), that the present low levels of storm activity on the mid west and northeast coasts of Australia are unprecedented over the past 550 to 1,500 years. The CAI allows for a direct comparison between the modern instrumental record and long-term palaeotempest (prehistoric tropical cyclone) records derived from the ¹⁸O/¹⁶O ratio of seasonally accreting carbonate layers of actively growing stalagmites. Our results reveal a repeated multicentennial cycle of tropical cyclone activity, the most recent of which commenced around ad 1700. The present cycle includes a sharp decrease in activity after 1960 in Western Australia. This is in contrast to the increasing frequency and destructiveness of Northern Hemisphere tropical cyclones since 1970 in the Atlantic Ocean^{6, 7, 8} and the western North Pacific Ocean^{6, 7}. Other studies project a decrease in the frequency of tropical cyclones towards the end of the twenty-first century in the southwest Pacific^{7, 9}, southern Indian^{9, 10} and Australian¹¹ regions. Our results, although based on a limited record, suggest that this may be occurring much earlier than expected

REFERENCES

Haig, J., Nott, J. and Reichart, G. (2014) Australian tropical cyclone activity lower than at any time over the past 550–1,500 years, Nature 505, 667–671 doi:10.1038/nature12882 [Abstract]

ABSTRACT

REFERENCES

Related blog posts