California Drought – A Novel Statistical Analysis of Unrealistic Climate Models and of a Reanalysis That Should Not Be Equated with Reality

UPDATE:  See the update at the end of the post.

In the post The Obvious Failures of Climate Science That Mainstream Media Ignores, I promised to discuss the paper behind the National Science Foundation press release Cause of California drought linked to climate change. That paper was Swain et al. (2014) “The Extraordinary California Drought of 2013/2014: Character, Context and the Role of Climate Change”. It is included in the Special Supplement to the Bulletin of the American Meteorological Society (BAMS report) Vol. 95, No. 9, September 2014, Explaining Extreme Events of 2013 From A Climate Perspective.

Swain et al. (2014) begins:

California’s driest 12-month period on record occurred during 2013/14, and although global warming has very likely increased the probability of certain large-scale atmospheric conditions, implications for extremely low precipitation in California remain uncertain.

Basically, Swain et al. (2014) are saying that human-induced global warming increased the likelihood of the atmospheric conditions that caused the 2013/14 drought, but they have no idea what will happen in the future.

The NSF press release provides a brief overview of the Swain et al. (2014) (my boldface):

Climate scientist Noah Diffenbaugh of Stanford University and colleagues used a novel combination of computer simulations and statistical techniques to show that a persistent region of high atmospheric pressure over the Pacific Ocean–one that diverted storms away from California–was much more likely to form in the presence of modern greenhouse gas concentrations.

A “novel combination of computer simulations and statistical techniques”?

Let’s see, looking at the paper, Swain et al. used climate models stored in the CMIP5 archive, but those models have no bases in reality. The models stored in the CMIP5 archive cannot simulate (hindcast) global or regional surface temperatures or precipitation. We discussed and illustrated those failings in a series of posts in 2013 in advance of the IPCC’s 5^th Assessment Report. Topics discussed included:

• Global Precipitation
• Satellite-Era Sea Surface Temperatures
• Global Surface Temperatures (Land+Ocean) Since 1880
• Global Land Precipitation & Global Ocean Precipitation

Those posts were also cross posted at WattsUpWithThat. And I expanded on that series and presented the model faults in my book Climate Models Fail.

Swain et al. also used a reanalysis (computer model that has data as inputs). But a reanalysis is not reality. It is not the same as observations-based data. It’s a specialized climate model.

The National Center for Atmospheric Research (NCAR) presents the limitations of reanalyses at the end of the Summary of their webpage titled Atmospheric Reanalysis: Overview and Comparison Tables. There they write (my boldface):

Key Limitations:

• Reanalysis data sets should not be equated with “observations” or “reality”
• The changing mix of observations, and biases in observations and models, can introduce spurious variability and trends into reanalysis outpu
• Observational constraints, and therefore reanalysis reliability, can considerably vary depending on the location, time period, and variable considered

Yet Swain et al. treated the reanalysis they used as if it represented reality.

Out of the more than 3 dozen climate models stored in the CMIP5 archive, Swain et al. (2014) cherry-picked 3 (GISS-E2-H, HadGEM2-ES, NorESM1-M) for their study.

We discussed the recent high sea surface temperatures in the eastern extratropical North Pacific in the post On The Recent Record-High Global Sea Surface Temperatures – The Wheres and Whys. Figure 1 is Figure 15 from that post. The eastern extratropical North Pacific is the region with the recent persistent hot spot that contributed to the California drought. That region of the North Pacific cooled, not warmed, based on the linear trend from January 1989 to December 2012. That’s almost 2.5 decades where the sea surface temperatures cooled, before the unusual upswing.

Figure 1

And even with the addition of the most recent data, Figure 2, the sea surface temperatures of the eastern extratropical North Pacific show no warming since January 1989, based on the linear trend. (That was Figure 16 in the earlier post.)

Figure 2

We also showed in that post that the sea surfaces of the North Pacific as a whole had not warmed from January 1989 to December 2012, the year before the unusual warming episode. Figure 5 from that post is shown below in Figure 3.

Figure 3

Let’s take a look at how well the models simulated those absences of warming. Wanna guess?

According to the models selected by Swain et al (2014), the surfaces of the eastern extratropical North Pacific should have warmed more than 0.65 deg C from January 1989 to July 2014 if they were warmed by manmade greenhouse gases. That’s based on an unrealistic warming rate of more than 0.25 deg C/decade.  See Figure 4. But, as shown above in Figure 2, based on the linear trend of the data, they didn’t warm.

Figure 4

And as shown in Figure 5, based on the models selected by Swain et al., the surfaces of the North Pacific as a whole should have warmed more than 0.60 deg C from January 1989 to December 2012 if they were warmed by those nasty little manmade greenhouse gases. But once again, referring to Figure 3, based on the linear trend of the data, there was no warming.

Figure 5

It’s tough to employ climate models so you can claim that manmade greenhouse gases caused the California drought, when the models used by Swain et al. can’t simulate the lack of warming of one of the key metrics associated with it.

In summary, Swain et al. performed “novel” statistical techniques on dubious climate models and on a reanalysis that should “not be equated with ‘observations’ or ‘reality’”. Swain et al. also failed to note or consider that the sea surface temperatures of the eastern extratropical Pacific cooled from 1989 to 2012 and that the North Pacific as a whole showed no warming for that same time period, while the climate models they employed showed the oceans warming drastically during those periods.

And some people wonder why climate skeptics are skeptical of climate science.

SOURCE

The Reynolds OI.v2 sea surface temperatures and the outputs of the CMIP5-archived models used by Swain et al. are available through the KNMI Climate Explorer.

UPDATE

Based on a comment on the thread of the cross post of this article at my blog Climate Observations, some persons might be wondering why I compared models of sea surface temperature to data, when Swain et al (2014) focused on geopotential height (see note below).  My reply:

The large scale atmospheric circulation patterns are in part dependent on local sea surface temperatures. The oceans and atmosphere are coupled. If the models cannot simulate the sea surface temperatures properly, then they are not simulating atmospheric circulation properly.

Let’s confirm that.

Lead author Swain also wrote an article for The California Blog at WeatherWest titled Special update: The Extraordinary California Drought of 2013-2014: Character, Context, and the Role of Climate Change. In it, he acknowledged the relationship between ocean and atmosphere a number of times.  On page 2 of his article, Swain wrote (my boldface):

Several recent studies have examined precisely this possibility in assessing cause of the extraordinary persistence of the Ridiculously Resilient Ridge in 2013 and 2014. Wang et al. (2014) find that tropical West Pacific SST warm anomalies (associated with the West Pacific Warm Pool that acts as a precursor of El Niño) played a leading role in causing the strength and longevity of the Triple R by generating a recurring series of atmospheric “Rossby waves” that propagated from west to east across the Pacific Basin. Wang and Schubert (2014) find that the North Pacific SST warm anomalies during early 2013 created a “predilection” for dry conditions during the second half of the 2013-2013 “rainy season” in California, and Funk et al. (2014) also report that the observed Pacific SST anomalies during 2013-2014 contributed to the extremely low precipitation that was observed during 2013-2014.

However, Funk et al. also note that it’s possible the record-breaking warmth in the North Pacific (discussed further below) was actually a geographically remote response to the changes elsewhere in the Pacific–similar to the mechanism considered by Wang et al. It’s even possible that that the Triple R played a role in sustaining itself by reducing North Pacific storm activity and preventing vertical mixing of cooler sub-surface ocean water,  culminating in a self-reinforcing feedback loop by which atmospheric ridging led to warm SSTs , which in turn led to more ridging, and so on. Regardless of whether the record-breaking warmth in the North Pacific was the primary cause of the Triple R or merely a secondary one, it’s pretty clear that Pacific SST anomalies contributed to the persistent northeastern Pacific ridging and extremely low California precipitation observed in 2013-2014.

Swain confirmed that the sea surface temperatures of the eastern extratropical North Pacific and atmosphere above it are interrelated, and that the warm sea surface temperatures contributed to the California drought.

Looking back now at my post above, Figures 1 and 2 showed that the sea surface temperatures of the eastern extratropical North Pacific had not warmed for 2.5 decades, and had cooled prior to the unusual warming…while the climate models employed by Swain et al, Figure 4, showed the sea surface temperatures of that region should have warmed more than 0.65 deg C in those 2.5 decades if they were warmed by manmade greenhouse gases.

In other words, the climate models employed by Swain et al are not realistic representations of climate in the eastern extratropical North Pacific. This further indicates they have no value when attempting to determine the cause or causes of the California drought, and no values when trying to attribute that drought to manmade factors.

Note: If the term geopotential height is new to you, see the ECMWF webpage here.