Simple Statistical Model Using Recent Droughts

Changes in the exceptionally dry years (droughts) have been estimated in the Drought Exceptional Circumstances Report (DECR) in two ways: (a) a statistical modification of the observed rainfall data (Box 3); and (b) analysis of simulations from 13 climate models. Up until now I have been looking at the modeling in approach (b). Today I started to look at approach (a). As mean rainfall declines the probability of exceptionally low rainfall increases. This is graphed in Box 3 (see also Table 6).

The parameters used in this simple extrapolation exercise have curious inconsistencies with their source. The DECR report says:

3. Likely Change in Climate Averages. Over Australia, a warmer and drier climate is expected. Median estimates for 2030 indicate a warming of about 1°C, relative to 1990, with less warming near the coast and more warming inland, a 3 to 5% decrease in rainfall with slightly larger decreases in central and south-western areas and little change in the far north, and a 2 to 4% increase in potential evaporation (Figure 7)(2).

2. CSIRO and Australian Bureau of Meteorology. Climate change in Australia, Technical report. 148pp (2007).

Only, the Climate 2007 report doesn’t say that. What it says is:

5.2.1 Median precipitation change by 2030. Best estimates of annual precipitation change represent little change in the far north and decreases of 2% to 5% elsewhere.

Climate 2007 estimates 2-5% reduction in some regions, not 3-5% in all regions. The DECR estimates at least double literature reference with no apparent explanation. But not satisfied with the arbitrarily doubled figure of 3-5% in all regions, the DECR goes on.

A 10 per cent decrease in mean annual rainfall across most of Australia is a possible scenario by 2030 (see Section 3). This decrease roughly doubles the risk of exceptionally low rainfall in five of the study regions, and almost triples the risk for the Vic&Tas and SW WA regions.

Where does the 10% decrease in rainfall as a possible scenario come from? The Climate 2007 report mentions 10% only in relation to the range.

5.2.2 The uncertainties in the precipitation change by 2030. The range of precipitation change in 2030 allowing for model to model differences is large. Annually averaged, the 10th to 90th percentile range is around -10% to +5% in northern areas and -10% to little change in southern areas.

The possible scenario used in DECR, the scenario that is the basis for the high range estimates, is the 10 percentile range of the uncertainty of rainfall reduction in the Climate 2007 report — the 1 in 10 chance.

From the 10% reduction in mean rainfall it is concluded that:

This decrease roughly doubles the risk of exceptionally low rainfall in five of the study regions, and almost triples the risk for the Vic&Tas and SW WA regions.

This finding is referred to the summary as the high scenario.

Under the high scenario, EC declarations would likely be triggered about twice as often and over twice the area in all regions. In SWWA the frequency and areas covered would likely be even greater.

Well yes, a 10% reduction in rainfall is a high scenario, its actually a 0.1 unlikely possibility, and it is within the range of Climate 2007 CI (just). But in arbitrarily choosing this value they are sacrificing the most likely results of this method (which are not reported) for the most extreme scenario.

The DECR then misrepresents the Climate 2007 report by claiming it predicts a 3 to 5% decrease in rainfall when it actually predicts little change in the far north and decreases of 2% to 5% elsewhere. And if that is not enough, they add the word “likely” to the DECR summary to make the high scenario sound even more “likely”. Here they are sacrificing scholarship in order to exaggerate global warming effects.

Under the high scenario, EC declarations would likely be triggered about twice as often and over twice the area in all regions.

Truth in advertising would require they also report that the median decrease in rainfall over Australia, say 1-2%, would result in 5 to 10 times less reduction that claimed, i.e. the analysis shows frequency and severity of droughts would only be expected to increase by 5-10% in 2030. Also while quoting the limit of the CI (so called high scenario) of doubling of drought frequency, they would also quote the (equally probable) upper limit of the CI of no change or increase in rainfall, with no increase or mild decrease in drought frequency.

This shows they will sacrifice the most obvious and scholarly interpretation for the sake of extreme global warming effects. Hey, why spoil a good story for the sake of the truth?

Advertisements

0 thoughts on “Simple Statistical Model Using Recent Droughts

  1. Pingback: coupon sites

  2. Pingback: grumpy cat memes

  3. Pingback: bateria do laptopa hp compaq

  4. Pingback: pokazy gotowania

  5. Pingback: livecooking

  6. Pingback: witryna firmowa

  7. Pingback: zobacz tutaj

  8. Pingback: kliknij link

  9. Pingback: strona firmy

  10. Pingback: strona www

  11. Pingback: Kraków

  12. Pingback: depilacja

  13. Pingback: zobacz tutaj

  14. Pingback: komputer79.bcz.com

  15. Pingback: kliknij

  16. Pingback: kliknij tutaj

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s