Are droughts on a decreasing trend?

This post has been [updated].

Yesterday I noticed a tweet by Matt "Northern Rock" Ridley, brother-in-law of Owen Paterson.
He said "Global decline in droughts since 1982" and linked this Nature article by Zengchao Hao et al. (pub online March 20140.

Ridley is dedicated to the proposition that all is well in the world and global warming is not a problem.

The paper does indeed carry this figure which shows a decline in the incidence of drought from 1982-2012. [I have been told by one of the authors that the trend shown here is not significant]*

(click on the figures to make them bigger)

Now the Pacific Decadal Oscillation, which is the climatic correlate of the El Nino/La Nina oceanic cycle, has also been in decline since the 1980s. Here it is, from the excellent Wood for Trees website, declining away.

So I use the MSPaint facility on my computer to combine the Hao's drought index with the PDO index. This involves pasting in, and a bit of pulling around to get the time frames to fit, so it is perfectly possible that distortions have crept in. 

This is what we get:






The red line is the PDO index. The brown dots represent volcanoes - El Chichon in 82, and Pinatubo in 1991. Big volcanoes have a cooling effect on global surface temperatures.

There is a modest agreement and correlation between the two curves in 1986-90, 1993-2000, and weak agreement from 2005-2012. Pinatubo seems to have broken the relationship in 91-92, and they are out of phase in 2002-4.

Just for comparison, here are global Sea surface temperatures overlaid on Hao's drought Index. The correlation is far less clear, and I have had to remove the trend, since global sea surface temperatures are on a rising trend.

Now I make no claims about the significance of the apparent correlations between droughts and PDO. Ocular comparisons are fallible. We need a proper climatologist to compare the two datasets and apply mathematical tools to judge whether the correlation is significant. [The authors have kindly given me their dataset, and I will try to find an Excel-competent person to compare the dataset with the PDO dataset].

If the correlation is significant, it confirms at very least that when the PDO goes into positive phase, global droughts will increase. Indeed, the correlation between positive PDO and droughts in the Northern USA is already known.

Now here is the long term pattern for the PDO since 1900, smoothed out:

This figure suggests that we are in a PDO downswing (which explains the predominant La Nina episodes over the last decade or so, which have contributed to the pause in surface temperature warming). It also suggests that an upswing is due in 10-15 years' time, with the next peak due around 2030.
We have to expect more droughts worldwide about that in 2030. 

So do not be misled by the ex-chairman of Northern Rock into thinking that increasing CO2 levels in the atmosphere will mean that droughts will become ever more infrequent. When the PDO does turn positive, it will collaborate with the duvet that our fossil CO2 is throwing around the planet to create an unwanted set of circumstances that will make Ridley's crash of  Northern Rock look like a minor bump.

*But one of  the authors, Prof Amir AghaKouchak, PhD, PE, gives this paper, which shows a significant drying trend in the Southern Hemisphere, and this paper which shows some significant regional trends.