...

[AvantHiatus]

We will only know in hindsight if the negative regime is over already.  The historic record is too short to provide much guidance. 

It would be over because the Ocean's are filling with heat and not because of natural variability. Assuming it does not return in 1-3 years. A disconnection between the IPO states and ocean temperatures is very unusual, in my opinion.

[ORH_wxman]

It would be over because the Ocean's are filling with heat and not because of natural variability. Assuming it does not return in 1-3 years. A disconnection between the IPO states and ocean temperatures is very unusual, in my opinion.

 

 

The PDO is a detrended dataset...your post makes no sense.

[AvantHiatus]

The PDO is a detrended dataset...your post makes no sense.

Perhaps but who is to say that a -PDO of the 21st century is like the +PDO of the 20th century? Not necessarily claiming this but it would be difficult to assume that traditional cycles would remain constant and uniform in light of different atmospheric chemistry.

[ORH_wxman]

Perhaps but who is to say that a -PDO of the 21st century is like the +PDO of the 20th century? Not necessarily claiming this but it would be difficult to assume that traditional cycles would remain constant and uniform in light of different atmospheric chemistry.

 

I don't think that was claimed anywhere...on the contrary if you read a few posts up, we were just discussing how the past behavior of the PDO has been anything but constant and uniform.

 

Adding tons of heat to the ocean certainly didn't stop it from going negative after the big +PDO regime of the 1980s/1990s.

[AvantHiatus]

A secondary heatsink in the Atlantic Ocean would support a higher ECS as the average global temperature has remained relatively constant since 1999 (within 0.15C of peak warming). Oceanic heat transport combined with lower solar activity should of lowered surface temperatures substantially.

[chubbs]

Agree with authors that there is a roughly 60-year natural cycle in global temperatures (see my post in climate sensitivity thread). Hard for a non-expert to evaluate the relative importance of the Atlantic vs Pacific but suspect that the Atlantic and Pacific are coupled since cycles in both oceans appear to be similar in length.  Need to account for this natural cycle in evaluating global temperature trends over periods shorter than 60 years.  Over longer periods of time the natural cycle is much less important since warm and cool periods cancel out. 

[LocoAko]

http://www.nature.com/nclimate/journal/vaop/ncurrent/full/nclimate2355.html
 
 

Contribution of natural decadal variability to global warming acceleration and hiatus
 
Masahiro Watanabe, Hideo Shiogama, Hiroaki Tatebe, Michiya Hayashi, Masayoshi Ishii &Masahide Kimoto
 
Nature Climate Change  (2014) | doi:10.1038/nclimate2355

Received 23 May 2014  |  Accepted  24 July 2014  |  Published online 31 August 2014

Reasons for the apparent pause in the rise of global-mean surface air temperature (SAT) after the turn of the century has been a mystery, undermining confidence in climate projections^1, 2, 3. Recent climate model simulations indicate this warming hiatus originated from eastern equatorial Pacific cooling⁴ associated with strengthening of trade winds⁵. Using a climate model that overrides tropical wind stress anomalies with observations for 1958–2012, we show that decadal-mean anomalies of global SAT referenced to the period 1961–1990 are changed by 0.11, 0.13 and −0.11 °C in the 1980s, 1990s and 2000s, respectively, without variation in human-induced radiative forcing. They account for about 47%, 38% and 27% of the respective temperature change. The dominant wind stress variability consistent with this warming/cooling represents the deceleration/acceleration of the Pacific trade winds, which can be robustly reproduced by atmospheric model simulations forced by observed sea surface temperature excluding anthropogenic warming components. Results indicate that inherent decadal climate variability contributes considerably to the observed global-mean SAT time series, but that its influence on decadal-mean SAT has gradually decreased relative to the rising anthropogenic warming signal.

 

http://onlinelibrary.wiley.com/doi/10.1002/2014GL060527/abstract
 
 

Drivers of decadal hiatus periods in the 20th and 21st centuries

 

Nicola Maher^1,2, 

Alexander Sen Gupta^1,2and

Matthew H. England^1,2

 

Article first published online: 20 AUG 2014
DOI: 10.1002/2014GL060527
©2014. American Geophysical Union.
 
The latest generation of climate model simulations are used to investigate the occurrence of hiatus periods in global surface air temperature in the past and under two future warming scenarios. Hiatus periods are identified in three categories: (i) those due to volcanic eruptions, (ii) those associated with negative phases of the Interdecadal Pacific Oscillation (IPO), and (iii) those affected by anthropogenically released aerosols in the mid-twentieth century. The likelihood of future hiatus periods is found to be sensitive to the rate of change of anthropogenic forcing. Under high rates of greenhouse gas emissions there is little chance of a hiatus decade occurring beyond 2030, even in the event of a large volcanic eruption. We further demonstrate that most nonvolcanic hiatuses across Coupled Model Intercomparison Project 5 (CMIP5) models are associated with enhanced cooling in the equatorial Pacific linked to the transition to a negative IPO phase.

[AvantHiatus]

Nice literature, the science has really advanced in the last few years.

[nflwxman]

http://www.nature.com/nclimate/journal/vaop/ncurrent/full/nclimate2355.html

 

 

 

 

http://onlinelibrary.wiley.com/doi/10.1002/2014GL060527/abstract

 

 

Great stuff. The second paper is very interesting and hold a lot of truth, IMO.  We've had a lot of natural and human made elements that should have conspired to cause a cooling the last 12 years or so, but temperatures have barely flatlined and even increased.  It would take a lot of volcanos in the future with increased forcing to stop an upward decadal trend.

[stadiumwave]

'57-'58 shows you can get a pretty long stretch of solid positive PDO values within a longer term decadal -PDO regime.

 

If we go longer than about one year of solid positives, then it would be more unique. You have to consider though that not all + or - PDO regimes are equal....the middle 20th century was a very potent negative regime...much more so than the negative regime in the late 19th and early 20th century. The positive regime from roughly 1920-1945 was not as strong as the one from 1977-2007 (or 1998, if you prefer that as the cutoff).

 

We like to compare everything to the 1945-1976 and 1977-1998 periods...but there is some evidence that those were quite strong compared to the previous 100 years...so expecting the PDO to mimic those periods exactly is probably a bit unreasonable. Comparisons should be taken with that in mind.

 

 

From 1957-1960 there were 25 months of +PDO values:

 

http://jisao.washington.edu/pdo/PDO.latest

 

We've just had 7 consec. months at this point...nothing even remotely unusual.

[chubbs]

An interesting read. Since that paper was published the PDO has at least temporarily flipped to positive and global temperatures have responded. To keep the hiatus going will probably need the negative PDO to reassert itself combined with falling AMO/NAO like the 60s and early 70s. Of course the longer the hiatus lasts the larger the radiative dis-equilibrium which only serves to warm the ocean more rapidly.  The more likely scenario is that the hiatus ends in the near future as the memory of the 1998 nino fades.

[bluewave]

An interesting read. Since that paper was published the PDO has at least temporarily flipped to positive and global temperatures have responded. To keep the hiatus going will probably need the negative PDO to reassert itself combined with falling AMO/NAO like the 60s and early 70s. Of course the longer the hiatus lasts the larger the radiative dis-equilibrium which only serves to warm the ocean more rapidly.  The more likely scenario is that the hiatus ends in the near future as the memory of the 1998 nino fades.

 

The interesting thing is that the PDO at least temporarily flipping positive hasn't resulted in a decline in the anomalous

trade wind pattern that is holding this El Nino event closer to the weaker side of the spectrum.

This was the first time during an El Nino event that there was no significant WWB pattern

west of the dateline during a fall(SON) El Nino event. This goes to the record breaking

trade wind pattern that has been a staple of the hiatus as noted in several recent papers.

[chubbs]

The interesting thing is that the PDO at least temporarily flipping positive hasn't resulted in a decline in the anomalous

trade wind pattern that is holding this El Nino event closer to the weaker side of the spectrum.

This was the first time during an El Nino event that there was no significant WWB pattern

west of the dateline during a fall(SON) El Nino event. This goes to the record breaking

trade wind pattern that has been a staple of the hiatus as noted in several recent papers.

 Don't think there is any PDO predictability other than persistance, so an interesting period coming up with short-term and longer-term trends conflicting.

[nflwxman]

The interesting thing is that the PDO at least temporarily flipping positive hasn't resulted in a decline in the anomalous

trade wind pattern that is holding this El Nino event closer to the weaker side of the spectrum.

This was the first time during an El Nino event that there was no significant WWB pattern

west of the dateline during a fall(SON) El Nino event. This goes to the record breaking

trade wind pattern that has been a staple of the hiatus as noted in several recent papers.

 

Correct me if I'm wrong, wasn't there some papers that speculated that the anomalous trade winds were more related to a disparity between CATL and WPAC temperatures rather than the PDO itself?

[bluewave]

 Don't think there is any PDO predictability other than persistance, so an interesting period coming up with short-term and longer-term trends conflicting.

 

The other interesting feature of the hiatus has been the NH winter temperatures contribution to the slower

rise in temperatures. We would also need to be able to predict this feature going forward. If we saw such 

winter cooling during the 77-97 +PDO era, then the warming rate per decade would have been lower.

 

http://iopscience.iop.org/1748-9326/7/1/014007

[bluewave]

Correct me if I'm wrong, wasn't there some papers that speculated that the anomalous trade winds were more related to a disparity between CATL and WPAC temperatures rather than the PDO itself?

 

Yes, I posted those links further back in this thread. The WPAC flipped warm right after the 97-98 Super EL Nino

and has not returned to the previous cooler state. 

[chubbs]

The other interesting feature of the hiatus has been the NH winter temperatures contribution to the slower

rise in temperatures. We would also need to be able to predict this feature going forward. If we saw such 

winter cooling during the 77-97 +PDO era, then the warming rate per decade would have been lower.

 

http://iopscience.iop.org/1748-9326/7/1/014007

 Certainly hope that trend continues and produces a string of +PDO/-AO winters. None of this is predictable in detail though particularly regional effects.

[stadiumwave]

Excellent Bluewave....thanks for posting!

[chubbs]

Crunch time for the hiatus. If it ends in 2015 people are going to look back 20 years from now and wonder what all the excitement was about. Looking at ocean temperatures it is not like the 1940-70 hiatus at all.

 

[ORH_wxman]

Crunch time for the hiatus. If it ends in 2015 people are going to look back 20 years from now and wonder what all the excitement was about. Looking at ocean temperatures it is not like the 1940-70 hiatus at all.

 

Hiatus.png

 

 

I view it more as the slowdown vs expectations rather than the exact demarcation of where there is a technically 0 slope on the sfc temps. Even though it is a fun excercise to pick the endpoints...but from a scientific standpoint in climate science, those endpoints are relatively meaningless.

 

The hiatus can end from a statistical standpoint, but the underlying "problem" of the slower warming isn't really solved until we warm at a fairly rapid rate again. Climate models warm us very quickly, so it is difficult to keep up with that warming rate....we have to basically see no regression during the next La Nina period, and then another big step jump during the next El Nino.

 

I know a lot of the blogosphere is obsessed with the exact dates of the hiatus and the exact hundreth of a degree of the temperature metrics and whether it is a new record or 2nd place or 3rd place, etc, etc...but the actual science isn't really worried about that. That's not where the true debate is. The hiatus in a truly scientific sense (rather than trivial statistics) has called into question what we thought we knew about both equilibrium sensitivity (ECS) and more importantly, transient climate response (TCR).  

 

 

 

Also, those aren't ocean sea surface temps on your graph, it is heat content. The SSTAs actually had a negative slope from the late 90s until the big spike in the past year.

[StudentOfClimatology]

Agreed that 2015 is a crucial year for those who believe the hiatus to be an indicator of a slowdown. Here's NCDC, and HADCRUT4:

[chubbs]

I view it more as the slowdown vs expectations rather than the exact demarcation of where there is a technically 0 slope on the sfc temps. Even though it is a fun excercise to pick the endpoints...but from a scientific standpoint in climate science, those endpoints are relatively meaningless.

 

The hiatus can end from a statistical standpoint, but the underlying "problem" of the slower warming isn't really solved until we warm at a fairly rapid rate again. Climate models warm us very quickly, so it is difficult to keep up with that warming rate....we have to basically see no regression during the next La Nina period, and then another big step jump during the next El Nino.

 

I know a lot of the blogosphere is obsessed with the exact dates of the hiatus and the exact hundreth of a degree of the temperature metrics and whether it is a new record or 2nd place or 3rd place, etc, etc...but the actual science isn't really worried about that. That's not where the true debate is. The hiatus in a truly scientific sense (rather than trivial statistics) has called into question what we thought we knew about both equilibrium sensitivity (ECS) and more importantly, transient climate response (TCR).  

 

 

 

Also, those aren't ocean sea surface temps on your graph, it is heat content. The SSTAs actually had a negative slope from the late 90s until the big spike in the past year.

 Agree we shouldn't get hung up on small plus or minus slopes in short-term trends. AGW and climate models should be evaluated on 30+ years of data. If the current hiatus ends next year then it will not be long enough or sharp enough to have much impact on long-term temperature trends. The increase in ocean heat content indicates radiative imbalance so there is plenty of forcing for surface warming when the natural forces discussed in this thread reverse.

[ORH_wxman]

 Agree we shouldn't get hung up on small plus or minus slopes in short-term trends. AGW and climate models should be evaluated on 30+ years of data. If the current hiatus ends next year then it will not be long enough or sharp enough to have much impact on long-term temperature trends. The increase in ocean heat content indicates radiative imbalance so there is plenty of forcing for surface warming when the natural forces discussed in this thread reverse.

 

 

We should be accelerating the warming though and not reducing it.

 

There's no doubt we will have a positive 30 year warming trend from 2000-2030 in my mind...but it may be under-shooting model projections by quite a lot. The models don't account for a slower period of warming...they just warm us faster and faster which is why there's a debate on what is wrong with them. That's where the literature is right now...TCRs, sensitivity, and which variables (including natural oscillations) the models are missing.

[chubbs]

We should be accelerating the warming though and not reducing it.

 

There's no doubt we will have a positive 30 year warming trend from 2000-2030 in my mind...but it may be under-shooting model projections by quite a lot. The models don't account for a slower period of warming...they just warm us faster and faster which is why there's a debate on what is wrong with them. That's where the literature is right now...TCRs, sensitivity, and which variables (including natural oscillations) the models are missing.

Models are never going to be able to accurately predict the details of ENSO/PDO/AMO etc. so of course they will look bad during a hiatus.  We are  stuck with the reality that temperatures are going to increase somewhere within a fairly wide uncertainty band. The probability is highest on the low side but the upper end cannot be ruled out.  The models are probably never going to be good enough to give a detailed roadmap.

 

[ORH_wxman]

Models are never going to be able to accurately predict the details of ENSO/PDO/AMO etc. so of course they will look bad during a hiatus.  We are  stuck with the reality that temperatures are going to increase somewhere within a fairly wide uncertainty band. The probability is highest on the low side but the upper end cannot be ruled out.  The models are probably never going to be good enough to give a detailed roadmap.

 

FIG_11-25_UPDATE.png

 

 

Of course the model mean won't get every small detail correct, but they aren't getting the larger picture correct either. That's the whole point. The models have no idea on natural variability.Theoretically, if they did, then we would see many models within a large ensemble forecast a hiatus, but they don't. Something like 2% of them forecasted anything close to what has happened. That's inadequate. It means that even the models that try and put us in a natural cooling cycle for 15-20 years are overwhelmed with their GHG forcing to produce higher trends than what actually happened. We can chalk this up to either a >2 sigma natural variation event or that the models are probably fundamentally flawed. We can't completely rule out idea #1, but the smart money would be on the latter scenario.

 

There's recent papers on this exact topic and that is where the science is headed right now.

 

When I discuss it though, i always get this underlying feeling that many people take it as the same thing as denying global warming or something similar which is patently ridiculous. A slower warming trend vs a fast warming trend isn't that strange of a concept to debate. Though I know it is more popular on the blogospheres to have the strawmen fights between deniers and alarmists...it's never anything in between where the real science is. Always about zero warming vs the worst case scenario.

 

It may not be as exciting as that, but I do find it very interesting where the actual debate in the literature is and not just the rhetoric on the blogs.

[chubbs]

Of course the model mean won't get every small detail correct, but they aren't getting the larger picture correct either. That's the whole point. The models have no idea on natural variability.Theoretically, if they did, then we would see many models within a large ensemble forecast a hiatus, but they don't. Something like 2% of them forecasted anything close to what has happened. That's inadequate. It means that even the models that try and put us in a natural cooling cycle for 15-20 years are overwhelmed with their GHG forcing to produce higher trends than what actually happened. We can chalk this up to either a >2 sigma natural variation event or that the models are probably fundamentally flawed. We can't completely rule out idea #1, but the smart money would be on the latter scenario.

 

There's recent papers on this exact topic and that is where the science is headed right now.

 

When I discuss it though, i always get this underlying feeling that many people take it as the same thing as denying global warming or something similar which is patently ridiculous. A slower warming trend vs a fast warming trend isn't that strange of a concept to debate. Though I know it is more popular on the blogospheres to have the strawmen fights between deniers and alarmists...it's never anything in between where the real science is. Always about zero warming vs the worst case scenario.

 

It may not be as exciting as that, but I do find it very interesting where the actual debate in the literature is and not just the rhetoric on the blogs.

 

There have been several recent papers however like the one below which show that the models do not overpredict if the modeling is properly aligned with actual conditions. Climate models do not have the capability to make 10-15 year forecasts.

 

Markus Huber & Reto Knutti, (2014), "Natural variability, radiative forcing and climate response in the recent hiatus reconciled", Nature Geoscience, Volume: 7, Pages: 651–656, doi:10.1038/ngeo2228

http://www.nature.com/ngeo/journal/v7/n9/full/ngeo2228.html

Abstract: "Global mean surface warming over the past 15 years or so has been less than in earlier decades and than simulated by most climate models. Natural variability, a reduced radiative forcing, a smaller warming response to atmospheric carbon dioxide concentrations and coverage bias in the observations have been identified as potential causes. However, the explanations of the so-called ‘warming hiatus’ remain fragmented and the implications for long-term temperature projections are unclear. Here we estimate the contribution of internal variability associated with the El Niño/Southern Oscillation (ENSO) using segments of unforced climate model control simulations that match the observed climate variability. We find that ENSO variability analogous to that between 1997 or 1998 and 2012 leads to a cooling trend of about −0.06 °C. In addition, updated solar and stratospheric aerosol forcings from observations explain a cooling trend of similar magnitude (−0.07 °C). Accounting for these adjusted trends we show that a climate model of reduced complexity with a transient climate response of about 1.8 °C is consistent with the temperature record of the past 15 years, as is the ensemble mean of the models in the Coupled Model Intercomparison Project Phase 5 (CMIP5). We conclude that there is little evidence for a systematic overestimation of the temperature response to increasing atmospheric CO2 concentrations in the CMIP5 ensemble."

[ORH_wxman]

There have been several recent papers however like the one below which show that the models do not overpredict if the modeling is properly aligned with actual conditions. Climate models do not have the capability to make 10-15 year forecasts.

 

Markus Huber & Reto Knutti, (2014), "Natural variability, radiative forcing and climate response in the recent hiatus reconciled", Nature Geoscience, Volume: 7, Pages: 651–656, doi:10.1038/ngeo2228

http://www.nature.com/ngeo/journal/v7/n9/full/ngeo2228.html

Abstract: "Global mean surface warming over the past 15 years or so has been less than in earlier decades and than simulated by most climate models. Natural variability, a reduced radiative forcing, a smaller warming response to atmospheric carbon dioxide concentrations and coverage bias in the observations have been identified as potential causes. However, the explanations of the so-called ‘warming hiatus’ remain fragmented and the implications for long-term temperature projections are unclear. Here we estimate the contribution of internal variability associated with the El Niño/Southern Oscillation (ENSO) using segments of unforced climate model control simulations that match the observed climate variability. We find that ENSO variability analogous to that between 1997 or 1998 and 2012 leads to a cooling trend of about −0.06 °C. In addition, updated solar and stratospheric aerosol forcings from observations explain a cooling trend of similar magnitude (−0.07 °C). Accounting for these adjusted trends we show that a climate model of reduced complexity with a transient climate response of about 1.8 °C is consistent with the temperature record of the past 15 years, as is the ensemble mean of the models in the Coupled Model Intercomparison Project Phase 5 (CMIP5). We conclude that there is little evidence for a systematic overestimation of the temperature response to increasing atmospheric CO2 concentrations in the CMIP5 ensemble."

 

 

See, this is actually a problem for me (and others who argue the flaws in the models)...they are claiming that we basically did have a 2 sigma event...or that we didn't know what aerosols did in the past 10-15 years until recently. Yet, we assume we know the aerosol forcing from 50 years ago or 100 years ago. These are big assumptions that on the surface can be defended, but when you look closer, the past gets swept under the rug a bit. We make all these fine-tuning to the models in the past 15 years to account for the hiatus, but then assume the previous 100 were fine.

 

It certainly is an explanation and is plausible. I just find it less likely than the other explanation of GCMs having a TCR that is too high.

 

 

The good news is that it won't take very long for us to find out if that explanation is correct. Another 15-20 years will make it more obvious.

[chubbs]

See, this is actually a problem for me (and others who argue the flaws in the models)...they are claiming that we basically did have a 2 sigma event...or that we didn't know what aerosols did in the past 10-15 years until recently. Yet, we assume we know the aerosol forcing from 50 years ago or 100 years ago. These are big assumptions that on the surface can be defended, but when you look closer, the past gets swept under the rug a bit. We make all these fine-tuning to the models in the past 15 years to account for the hiatus, but then assume the previous 100 were fine.

 

It certainly is an explanation and is plausible. I just find it less likely than the other explanation of GCMs having a TCR that is too high.

 

 

The good news is that it won't take very long for us to find out if that explanation is correct. Another 15-20 years will make it more obvious.

The literature may be giving the wrong impression about model performance by not carefully aligning model emission scenarios, model forcings, and real world conditions. Certainly the blogospere is full of misleading model and data comparisions. Increases in the  NOAA GHG index since 1979 would support yearly temperature increases  of roughly 0.016C if the TCR was 2.0. The math is simple. man-made GHG increased by roughly 1% of pre-industrial CO2 per year between 1979 and 2013.   GHG impacts need to be adjusted to account for other man-made effects. The AR5 value from Lewis and Curry is 0.81. The following estimate then results: 0.016C T increase per year = 1% of pre-ind CO2 per year x 0.81 x 2C TCR. This simple excersize indicates that the models are probably in the right ballpark and calls into question the lowest of the published TCR estimates.

[ORH_wxman]

The literature may be giving the wrong impression about model performance by not carefully aligning model emission scenarios, model forcings, and real world conditions. Certainly the blogospere is full of misleading model and data comparisions. Increases in the  NOAA GHG index since 1979 would support yearly temperature increases  of roughly 0.016C if the TCR was 2.0. The math is simple. man-made GHG increased by roughly 1% of pre-industrial CO2 per year between 1979 and 2013.   GHG impacts need to be adjusted to account for other man-made effects. The AR5 value from Lewis and Curry is 0.81. The following estimate then results: 0.016C T increase per year = 1% of pre-ind CO2 per year x 0.81 x 2C TCR. This simple excersize indicates that the models are probably in the right ballpark and calls into question the lowest of the published TCR estimates.

 

 

Your calculation makes the dubious assumption that all the warming was anthropogenic since 1979. This is strongly contested in the literature.

 

Using a longer timeline creates a lower TCR.

[chubbs]

Your calculation makes the dubious assumption that all the warming was anthropogenic since 1979. This is strongly contested in the literature.

 

Using a longer timeline creates a lower TCR.

 I doubt there has been much natural contribution to warming between 1979 and 2013 when you include impacts of the most recent hiatus years.  My expectation is that we roughly maintain the 0.016 per year trend of the past 35 years going forward. As you say we will find out soon whether the models can climb back from the hiatus induced move to the lower end of the uncertainty band. 

 

You are right that the TCR that matches the observed temperature rise decreases as you extend the record back further in time. Using this same method a TCR of 1.7 is a good match to the entire GISS record from 1880 to 2013 but 2.0 is a better match for the  past 50 years. The apparent TCR increase with time is not surprising since a pipeline of future warming has build up gradually with time.