Hansen gets slammed on his 'extreme weather' alarmism

More on Extreme Weather and the Greenhouse Effect   NYT.com

Martin P. Hoerling, a federal research meteorologist specializing in climate dynamics, has written the following expansion and defense of his criticism of some assertions made in an Op-Ed article on climate change by James E. Hansen of NASA. His initial criticism was posted on the Dot Earth blog.   

Several scientific conclusions and claims were expressed in the NYT Op-Ed piece by J. Hansenwith which I raised concerns and objections on a scientific basis.   While allowing for imprecision in how nuanced climate change science is sometimes communicated in such venues, there were nonetheless statements in the Hansen NYT piece that drew my attention because they stood contrary to peer-reviewed literature.  Some of these claims could also be tested and/or falsified by simple tests using data available in the public domain, with examples given below.

The Hansen NYT piece asserts:  

"Over the next several decades, the Western United States and the semi-arid region from North Dakota to Texas will develop semi-permanent drought, with rain, when it does come, occurring in extreme events with heavy flooding. Economic losses would be incalculable. More and more of the Midwest would be a dust bowl. California’s Central Valley could no longer be irrigated. Food prices would rise to unprecedented levels."

To which I replied:  

“I am unaware of any projection for "semi-permanent" drought in this time frame over the expansive region of the Central Great Plains.  He implies the drought is to be a phenomena due to lack of rain (except for the brief, and ineffective downpours).  I am unaware of indications, from model projections, for a material decline in mean rainfall. “

Supporting Material for My Statement  

The peer-reviewed study by Milly et al. titled “Global Pattern of Trends in Streamflow and Water Availability in a Changing Climate”(Nature, 2005) used the IPCC CMIP3simulations to diagnose the relative change in surface runoff for the period 2040-2060 compared to 1900-1970.  A version of their published Fig. 4 is shown below.  Runoff is a holistic indicator of surface water balance that integrates the effects of changes in mean precipitation and its characteristics, and also changes in temperature via evapotranspiration.  These published results indicate no appreciable change in surface water availability over the semi-arid region from North Dakota to Texas, or for the Midwest as a whole, within this ensemble of  CMIP3 simulations.    The authors point out the various uncertainties in such regional scale projections, not the least of which one must also include the uncertainty in the consequences of changes in land cover and land use.

The recently published report by the U.S. Global Change Research Program titled “Global Climate Change Impacts in the United States”provided a synthesis of the current understanding of probable regional climate change impacts. This synthesis document states, in the section on Water Resources on pg. 45, “Precipitation and runoff are likely to increase in the Northeast and Midwest in winter and spring.” Drawn from such works as the Milly et al. study, this synthesis is clearly quite contrary to the assertion made in the Hansen NYT piece.Regarding flooding that is mentioned in the Hansen piece, it is important not to confuse heavy downpours with hydrologic flooding on a river basin scale, for instance.  The IPCC Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation (SREX, 2012http://www.ipcc-wg2.gov/SREX/) states a low confidence for both observed and projected changes in the magnitude and frequency of floods.

Summary  

The claim in the Hansen NYT piece that the Midwest would be a dustbowl in coming decades thus runs contrary to peer reviewed literature and recent assessments by the U.S. Global Research Program that emerged from the synthesis of current understanding by an expert team of scientists.

Regarding observed changes in climate of the Great Plains, I stated: 

“Indeed, that region (Great Plains) has seen a general increase in rainfall over the long term, during most seasons (certainly no material decline).  Also, for the warm season when evaporative loss is especially effective, the climate of the central Great Plains has not become materially warmer (perhaps even cooled) since 1900.  In other words, climate conditions in the growing season of the Central Great Plains are today not materially different from those existing 100 years ago.  This observational fact belies the expectations, from climate simulations, and in truth, our science lacks a good explanation for this discrepancy. “

Supporting Material for My Statement

The lack of a warming trend over the central United States during the past century, sometimes called the U.S. “warming hole”, has been especially noted in the peer-reviewed literature (e.g., Kunkel et al. Journal of Climate, 2006, Knutson et al. Journal of Climate2006).  Particularly striking has been a cooling trend in summertime temperatures at many meteorological observing stations located between the Rocky and Appalachian Mountains over the period 1901-2010 (Fig. 2, top).   This region of summertime cooling has generally coincided with a region of summertime mean precipitation increase (Fig. 2, bottom).  The data is the monthly Global Historical Climate Network data available at http://www.ncdc.noaa.gov/ghcnm/v3.php

To date, these regional patterns are not well understood on physical grounds. This, and other regional examples of climate trends, illustrate the need for a more comprehensive assessment on the causes of regional and seasonal differences in climate trends that considers multiple possible contributing factors, including atmospheric dynamics and coupled ocean-atmosphere processes, land surface and biological processes, atmospheric chemistry and aerosol and human-induced climate change.

Summary  

The certainty language expressed in the Hansen NYT piece about the coming dustbowl fate for the Great Plains region and Midwest is contrary to the low confidence of regional climate change projections for coming decades as documented in USGCRP and IPCC reports.  Not only are various regional patterns of trends that have been observed over the last century poorly understood, but the projections of regional changes in coming decades are highly uncertain.

The Hansen NYT piece asserts:

"The global warming signal is now louder than the noise of random weather..."  

To which I replied:  

“This is patently false.  Take temperature over the U.S. as an example. The variability of daily temperature over the U.S. is much larger than the anthropogenic warming signal at the local, weather time scales. Depending on season and location, the disparity is at least a factor of 5 to 10. I think that a more scientifically justifiable statement, at least for the U.S. and extratropical land areas is that --- Daily weather noise continues to drum out the siren call of climate change on local, weather scales.”

Supporting Material for My Statement  

Weather---as experienced on a daily basis and at any particular location--- is highly variable, but the challenge offered in the Hansen NYT piece is that the noise of such variations are now being drowned out by the global warming signal.   Debating this latter point should not, of course, be confused with opening a debate about the rise in global mean temperatures.  The IPCC (2007) has stated that warming of the climate system is unequivocal.  The unequivocal rise in global average temperature and the attribution that most of this rise is due to the rise in anthropogenic greenhouse gas concentrations, owes to the small natural variability of global mean temperatures compared to the large magnitude of the human-induced warming signal in global averaged temperatures.    In other words, for global mean conditions, the signal is much louder than natural variability of globally averaged temperatures.   The science is clear, we know that the planet has warming over the past century, and we are very confident as to why such warming has (and continues) to occur.

But that appears not to be the point of the NYT piece, as impliedby the subsequent context of Hansen’s statement regarding extreme local weather events. The “random weather” called out in the NYT piece is unlikely meant to refer to the variability in global mean temperature, but rather to the local conditions we regularly encounter in our own backyards and that swing back and forth across a range of conditions.  Thisrange (the random noise, as per Hansen) is in fact not smaller than the global warming signal, as shown from several lines of evidence below.

The recent published peer-reviewed study by Hawkins and Sutton (2012, Geophysical Research Letters) diagnoses the so-called “time of emergence” of climate signals from the noise of random variability at a local level.   Their analysis compares an estimate of a human-induced change in surface air temperature against an estimate of its natural variability for seasonally averaged data.   For a signal-to-noise ratio of two (signal being double the magnitude of the noise), they find that the time of emergence is after 2050 for most mid-latitude regions during cold and warm seasons.  The tropics, where noise of temperature variability is less than in mid-latitudes, the time of emergence is appreciably earlier.

But even that analysis is not quite germane to the Hansen assertion regarding the noise of random weather.   Figure 3 presents an analysis of the intensity of the variability of daily averaged surface temperature across the United States.

The data is daily temperature at Cooperative observing stations, which is available at http://www.ncdc.noaa.gov/oa/climate/research/gdcn/gdcn.html.  The top panel is the intensity of daily temperature variability during 1901-2010, averaged across the 12 calendar months January-December.   On average, daily temperature variability is about 5°C over North Dakota, to as small as 2°C over Florida.  The recent update, by NCDC, of the annual mean global mean warming signal is +0.51°C (for 2011 relative to a 20th Century reference).  It is thus evident that daily surface temperature variability is on order 5 to 10 times greater than the global warming signal (see Fig. 3, bottom).   Consistent with the published work of Hawkins and Sutton, it is obvious that the time of emergence of the global warming signal from this weather noise is far in the future under the assumption of continuing global warming.

Summary

The global warming signal is much smaller than the typical daily variability of surface air temperature over the United States.  Most of the magnitude of daily weather extremes owes its causes to natural internal fluctuations and not to global warming.  A possible exception could be imagined if global warming were also to increase the variability of daily temperatures (and not just increase the mean temperatures), but no compelling evidence to such effects has been shown.   While globally averaged temperatures have risen during the past century, the cause for which is very likely human-induce climate change, the signal of this change is still barely audible among the loud noise of daily, backyard weather fluctuations.

Weather, of course, is more than temperature variability.  While this discussion has involved temperature, weather involves rain, storms, winds, severe convection, clouds among others.   In this regard, it is important to reiterate the statement in IPCC SREX (2012) in their Executive Summary which states that “many weather and climate extremes are the result of natural climate variability”, and that “even if there were no anthropogenic changes in climate, a wide variety of natural and weather extremes would still occur”.

The Hansen NYT piece asserts:

" We can say with high confidence that the recent heat waves in Texas and Russia, and the one in Europe in 2003, which killed tens of thousands, were not natural events — they were caused by human-induced climate change."

To which I replied:

“Published scientific studies on the Russian heat wave indicate this claim to be false.  Our own study on the Texas heat wave and drought, submitted today to Journal of Climate, likewise shows that that event was not caused by human-induced climate change. These are not de novo events, but upon scientific scrutiny,one finds both the Russian and Texas extreme events to be part of the physics of what has driven variability in those regions over the past century.  Not to say that climate change didn’t contribute to the those cases, but their intensity owes to natural, not human, causes.”

Supporting Material for My Statement

The principal conclusion by Dole et al. (2011, Geophysical Research Letters) is that the extreme magnitude of the 2010 Russian heat wave was mainly due to internal dynamical processes (associated with atmospheric blocking), and that it was very unlikely that warming attributable to GHG forcing contributed substantially to the heat wave's magnitude. Rahmstorf and Coumou (2011, PNAS) concluded that a strong warming over western Russia (which they attribute primarily to GHG forcing) multiplied the likelihood of a record heat wave. They estimated an 80% probability that the 2010 July heat records in Moscow would not have occurred without climate warming. Barriopedro et al. (2011, Science), on the other hand, conclude that the magnitude of the 2010 event was so extreme that despite GHG warming, the likelihood of an analog over the same region remains fairly low at this time. This is consistent with estimates in Dole et al., which showed a very low probability of an event of this magnitude in 2010, but a rapidly increasing likelihood of crossing given thresholds in future climate, based on results from CMIP3 model runs.

Toward attempting to reconcile these conclusions, Otto et al (2012, Geophysical Research Letters) conclude that there is no substantial contradiction between studies by Dole et al. and Rahmstorf and Coumou, in that the heat wave “can be both mostly internally generated due to magnitude, and mostly externally-driven in terms of occurrence-probability”. Further discussion on this matter including an extensive list of recent published work on the Russian heat wave is available at http://www.esrl.noaa.gov/psd/csi/events/2010/russianheatwave/index.html

Summary

Analysis of various forced model simulations indicates that human influences did not contribute substantially to the magnitude of the Russian heat wave. Even accounting for a possible stronger warming signal, as suggested by Rahmsdorf and Coumou, these were still appreciably smaller than the peak magnitude of the event (which reached 10°C over Moscow during July).   Barriapedro et al. (2011) conclude that the magnitude of the 2010 event was so extreme that despite an increase in temperatures due to human climate change, the likelihood of an analog over the same region remains fairly low until the second half of the 21st century.   These results are thus consistent also with the Hawkins and Sutton (2012) results regarding the time of emergence of a climate change signal at local scales.

go to Revkins blog for graphics in above piece