...

[Jonger]

I'm going to put a graphic up, but when you see how global sea temperatures were taken just 10 years ago and how sparse the data is, you might not feel so comfortable with that statement.

Pay attention to the last 10-15 years.

0 - 700M just goes FLAT after 2003, look at the buoy deployment starting in 2003 onward.

You have seen the 0-2000m OHC graph from the OHC thread that friv started?

I have seen it.... 0-700 has been flat and 700-2000 has had a steep rise, but assuming only one figure flat lined doesn't remove skepticism. If anything, it brings questions over the accuracy of the entire pre 2003 measurement network. What are we comparing the trend line to, what really is the base and also, do the 2000m temps regularly rise every certain number of years.

True, the 2000m temps have steadily risen, but that has to be questioned after the other trendline changed after a more complete measurement network presented itself.

[Jonger]

With the ARGO network in place and adjusted, we should get a more clear picture as time goes by.

[f2tornado]

You have seen the 0-2000m OHC graph from the OHC thread that friv started?

 

What about depth >2000m? We don't even sample half the ocean volume. There's a lot of water below 2000m. The research into OHC is in its infancy and scientists are still, to some extent, figuring out the best way to resolve errors with the old XBT sensors.

[Snow_Miser]

Right, but most hindcasts start at 1900, which allows us to model the PDO/ENSO long enough in that it should have very little impact on the global temp trend. In a vacuum, where 1975-2005 is modeled entirely anthropogenic, I could see the GCMs way overdoing the sensitivity. Generally in most papers with hindcasts, that is not the case.

 

The hindcasts do a very poor job at simulating the early-20th Century rate of warming. This could be either because the models underestimate the importance of the early 20th Century increase in solar forcing, or that they don't simulate the PDO/ENSO very well. Or it could be a combination of both.

 

[nflwxman]

What about depth >2000m? We don't even sample half the ocean volume. There's a lot of water below 2000m. The research into OHC is in its infancy and scientists are still, to some extent, figuring out the best way to resolve errors with the old XBT sensors.

I think I read a Hanson paper siting that heat under 2000m is trivial to the process (it might have been another paper). I suppose anything is possible, we know very little about the really deep oceans.

[nflwxman]

The hindcasts do a very poor job at simulating the early-20th Century rate of warming. This could be either because the models underestimate the importance of the early 20th Century increase in solar forcing, or that they don't simulate the PDO/ENSO very well. Or it could be a combination of both.

It looks like the models perform the worst around the big climate "shifts" associated with the PDO. 1943-1948 and 1973-1977 are both periods where the models did not "see" the dramatic temperature changes. Perhaps that could explain what's happening now in surface temps. After 2005, we had a dramatic drop in the PDO. In general though, when you get a decade away from the major shifts- the models track very well. This is just a qualitative observation. Ill try and make some graphs later to see I'd I can back up my obs.

[tacoman25]

I gotta be honest. I don't believe this discussion will bear fruit. I think people should hold judgement on alarmists and deniers (etc) and who wins and losses. This last decade is certainly not enough data to make sweeping statements and change our models. For all we know, we could warm double the rate next time the PDO flips and surpass surface temp modeled projections

We are literally debating over 3 % of climate change's additional energy. We have not talked about how the PDO affects sea level rise (happening faster than "modeled," ocean heat content, sea ice (melting faster than "modeled"), and ice sheets.

 

You haven't proven to be particularly open-minded in your responses to people, so you may be right. You seem to be convinced of your position regardless of new evidence presented to you.

 

It seems as though you are going through the same thing as climate science in general: trying to reconcile the reality that everything is not exactly going according to expectations. It can be a bit unsettling, especially if you placed too much faith in climate scientists to begin with.

[tacoman25]

I really think climate scientists have know about this longer than many believe. There are several studies prior to this that show essentially the same thing (with quantifying it).

Do you think sea level rise will be faster since more of that energy is staying in the ocean for longer periods?

 

How much had you read from climate scientists in the 1990s and early 2000s? Honest question.

[nflwxman]

You haven't proven to be particularly open-minded in your responses to people, so you may be right. You seem to be convinced of your position regardless of new evidence presented to you.

It seems as though you are going through the same thing as climate science in general: trying to reconcile the reality that everything is not exactly going according to expectations. It can be a bit unsettling, especially if you placed too much faith in climate scientists to begin with.

I'm sorry I put my faith in the experts.

All my point was is that everyone needs to withhold judgement on surface temp models as they were not meant for short term validation. It's not a hard concept to grasp. I don't believe anyone has put forth a good argument to scrap long term projections as we now know it. I just don't buy we need to revise our 2050, 2100 projections due to a few years of model to observation divergence. Study the PDO? Yes. Maybe understand how it impacted past temperatures? Sure. But declare well validated climate models failures? No, sorry. Not yet.

10-20 years down the line? Perhaps

I'd like to hear your thoughts about why the models seemed to underestimate sea level rise the last 8 years? I have a feeling it's related to the PDO as well.

[f2tornado]

I'm sorry I put my faith in the experts.

All my point was is that everyone needs to withhold judgement on surface temp models as they were not meant for short term validation. It's not a hard concept to grasp. I don't believe anyone has put forth a good argument to scrap long term projections as we now know it. I just don't buy we need to revise our 2050, 2100 projections due to a few years of model to observation divergence. Study the PDO? Yes. Maybe understand how it impacted past temperatures? Sure. But declare well validated climate models failures? No, sorry. Not yet.

10-20 years down the line? Perhaps

I'd like to hear your thoughts about why the models seemed to underestimate sea level rise the last 8 years? I have a feeling it's related to the PDO as well.

 

You are putting faith in the experts but primarily in selective experts only. You continuously regurgitate stuff from John Cook's site, hook, line, and sinker. While I agree on your point the GCMs are not ideal short range forecast tools, there are apparent larger feedback problems associated with them discussed in peer reviewed literature and summarized in the the recent water vapor feedback thread and elsewhere in this forum. http://www.americanwx.com/bb/index.php/topic/40960-on-the-water-vapor-feedback-comments-on-desser-2008-and-soden-et-al-2002/ This is not a symptom of "well validated climate models",  To borrow the words of John Christy, I believe the evidence is clear that the departures are real and significant which then exposes serious problems for the models as long-range forecasting tools. What you're proposing is knowingly buying a used Chevy that does not have a sound transmission. The vehicle is shiny on the surface, is functional, and you're 95% confident it will get you across town for a while but eventually the thing could, and likely will, break down. Wouldn't it make sense to get a new transmission before taking your friends on a road trip?

 

I did a simple Google search for "models underestimate sea level rise" and came up with this very recent paper. http://phys.org/news/2013-05-volcanic-underestimate-sea.html

 

[Snow_Miser]

It looks like the models perform the worst around the big climate "shifts" associated with the PDO. 1943-1948 and 1973-1977 are both periods where the models did not "see" the dramatic temperature changes. Perhaps that could explain what's happening now in surface temps. After 2005, we had a dramatic drop in the PDO. In general though, when you get a decade away from the major shifts- the models track very well. This is just a qualitative observation. Ill try and make some graphs later to see I'd I can back up my obs.

 

It's not just the climate shifts from one PDO phase to another, it's the rate at which we warmed from around 1916-1943. The models underestimated that rate by over a factor of 4. Clearly, the radiative forcings that the hindcasts have in the early-20th Century cannot explain the early-20th Century warm period. That may mean that there is a much larger solar forcing (ie an amplification mechanism like from Cloud Cover, resulting in a larger forcing than the TSI forcing in the hindcasts, ultimately resulting in a less sensitive system than the sensitivity in the hindcasts) or the models simulate multidecadal variability from the PDO extremely poorly. I personally think that it may be both.

[Snow_Miser]

We did have a massive volcano eruption in 1991 (Pinatubo) that kept the early 1990s temperatures depressed. A chunk of the 1990s warming is a rebound from that along with the Super El Nino at the end of the decade.

 

Even including the rapid 1990s warming period, the models still simulate too much warming by a factor of two over the last 20 years. The paper also notes that there has not been any statistically significant warming since 1998, or in 15 years, and that the modeled trend was off by over a factor of 4 for this timeframe.

 

http://www.nature.com/nclimate/journal/v3/n9/full/nclimate1972.html?WT.ec_id=NCLIMATE-201309

 

[f2tornado]

Even including the rapid 1990s warming period, the models still simulate too much warming by a factor of two over the last 20 years. The paper also notes that there has not been any statistically significant warming since 1998, or in 15 years, and that the modeled trend was off by over a factor of 4 for this timeframe.

 

http://www.nature.com/nclimate/journal/v3/n9/full/nclimate1972.html?WT.ec_id=NCLIMATE-201309

 

 

 

 

It's a paywall but I got some of the text from Curry... http://judithcurry.com/2013/08/28/overestimated-global-warming-over-the-past-20-years/

 

The evidence, therefore, indicates that the current generation of climate models (when run as a group, with the CMIP5 prescribed forcings) do not reproduce the observed global warming over the past 20 years, or the slowdown in global warming over the past fifteen years. uch an inconsistency is only expected to occur by chance once in 500 years, if 20-year periods are considered statistically independent. Similar results apply to trends for 1998–2012.

[Snow_Miser]

 

 

It's a paywall but I got some of the text from Curry... http://judithcurry.com/2013/08/28/overestimated-global-warming-over-the-past-20-years/

 

The evidence, therefore, indicates that the current generation of climate models (when run as a group, with the CMIP5 prescribed forcings) do not reproduce the observed global warming over the past 20 years, or the slowdown in global warming over the past fifteen years. uch an inconsistency is only expected to occur by chance once in 500 years, if 20-year periods are considered statistically independent. Similar results apply to trends for 1998–2012.

 

Dr. Curry hit the nail on the head with regard to why the 20 year trends documented by the paper are interesting:

 

"The selection of 20 years is interesting for several reasons.  It gets away from the ‘cherry picking’ criticism of starting with 1997 or 1998.  Also it includes the big jump from 1993-1998."

 

Even including the large 1990s warming rate, the observations indicate a warming rate that is less than half of the modeled trend. The science is not settled.

[ORH_wxman]

This is another paper published last year that shows a lower TCR and thus lower end warming projected for the 21st century:

 

http://onlinelibrary.wiley.com/doi/10.1029/2011GL050226/abstract

 

 

They hypothesize in the paper that the models may be too sensitive to changes in aerosols and GHGs due to over-cooling vs obs in volcanic eruptions and over-warming during the late 20th and early 21st century. Thus also conclude that they possibly do not capture internal variability well.

[chubbs]

5yr running average GISS surface temperature, MEI and PDO. Not sure PDO adds much beyond the MEI running average. In any case approaching the bottem of a PDO/MEI cooling phase at present.

[Snow_Miser]

There was another paper published on ENSO's role in the pause and in the recent warming trend.

 

"Update of the Chronology of Natural Signals in the Near-Surface Mean Global Temperature Record and the Southern Oscillation Index"

 

The abstract reads:

 

"With the potential controversy arising over a particular statistical analysis removed, the findings indicate that El Nino-Southern Oscillation exercises a major influence on mean global temperature. The results show the potential of natural forcing mechanisms to account for mean global temperature variation, although the extent of the influence is difficult to quantify from among the variability of short-term influences."

 

From the conclusions:

 

"All other things being equal, a period dominated by a high frequency of El Niño-like conditions will result in global warming, whereas a period dominated by a high frequency of La Niña-like conditions will result in global cooling. Overall, the results imply that natural climate forcing associated with ENSO is a major contributor to temperature variability and per- haps a major control knob governing Earth’s temperature."

 

[JBG]

I agree. It is refreshing too see some scientists attempt to scientifically tackle reasonable points brought up by skeptics out there rather than draw the sword. This paper ironically reasserts skeptics' views as Judith Curry is more pumped than I've ever seen. http://judithcurry.com/2013/08/28/pause-tied-to-equatorial-pacific-surface-cooling/

Some scientists are rightly afraid of ridicule or opprobrium if they buck the hysteria. And some "scientists", as we saw at East Anglia, will destroy data that doesn't support their government-funded world.

[f2tornado]

Some scientists are rightly afraid of ridicule or opprobrium if they buck the hysteria. And some "scientists", as we saw at East Anglia, will destroy data that doesn't support their government-funded world.

 

I didn't put much salt in the East Anglia e-mail controversy but there are a couple gems that really show what a circle jerk the place was/is. In that respect, your suggestion is a reasonable one. I still find it baffling nearly 100% of scientists agree on anthropogenic climate change yet are in a huge pissing match. Funding, ideology, prestige, group think, perhaps all of these contribute to the problem.

 

"There shouldn't be someone else at UEA with different views [from "recent extreme weather is due to global warming"] – at least not a climatologist."

-Phil Jones, UEA

 

"What if climate change appears to be just mainly a multidecadal natural fluctuation? They'll kill us probably ..."

-Tommy Wills, Swansea University

 

I wonder what the hurricane alarmists think of this... 

Model projections of atmospheric steering of Sandy-like superstorms

http://www.pnas.org/content/early/2013/08/28/1308732110.abstract?sid=64a8d333-0e49-4595-a084-2f858e346c55

 

[ORH_wxman]

Foster comments in on the recent paper:

 

http://tamino.wordpress.com/

 

 

The agreement is outstanding, including over the last decade and more, with correlation between the models and observations of 0.93. Using just the data since 1970, when — according to the authors — sea surface temperature data is more accurate, the correlation is an impressive 0.97. This is powerful evidence that the recent slowdown in global temperature increase is not a slowdown of “global warming,” i.e. man-made climate change, it’s simply partial cancellation of global warming by the natural fluctuations due to ENSO.

 

 

Gammon adds his thoughts:

 

http://blog.chron.com/climateabyss/2013/08/learning-from-the-hiatus/

 

Summary

 

This blog post describes two recent papers on the relatively slow increase of global temperatures over the past decade and a half.  It focuses on a paper by Kosaka and Xie, and further analyzes the data from Kosaka and Xie to explore issues of model accuracy and climate sensitivity.  The evidence points toward transient climate response being slightly weaker than the CMIP5 model average.  Natural variability appears to have caused the recent hiatus but appears not to have contributed significantly to the previous period of rapid warming.

 

 

Note: I don’t like that term, because it’s only the atmospheric part of the globe that is enjoying a hiatus from global warming.  The oceans continue to take up lots of extra heat, and the glaciers continue to melt.  Call it an “atmospheric warming hiatus”, and I’ll be much happier.

I’m one of those scientists.  In a blog post last year, I noted that El Niño has a large influence on year-to-year climate changes, but that if you sort each year according to whether it’s El Niño, neutral, or La Niña, and exclude years of major volcanic influence, there’s a steady warming trend all the way from the 1970s to the present day, with no sign of a hiatus.  My analysis was motivated by Foster and Rahmsdorf (2011), who looked more closely at both natural variations and variations in forcing and came to a similar conclusion.  Later analysis by Troy Masters questioned whether one could say for sure that there was no hiatus, and the issue remained open.

 

 

Bad analysis by Tamino (Foster) in accusing Judith Curry of cherry picking, but then he ignored the natural rise in the late 1970s-2000 timeframe according to the model in the study...even if you were to correct Curry using linear regression, its still a positive influence in the trend that is obvious when looking at the graphs.

 

But then again, his own paper still had an ENSO residual in it too, so I'm not surprised that he tends to gloss over it so much. But that is the weakness of so many of these models that try to extrapolate ENSO influence.

[Snow_Miser]

There was another paper published on the PDO and it's impact on the long term temperature trend by Courtillot et al.

 

From the paper:

 

"Thus, the Earth’s climate may have entered a new mode (a new ~30-yr episode) near the turn of the 20th century: no further temperature increase, a dominantly negative PDO index and a decreasing AMO index might be expected for the next decade or two."

 

[ORH_wxman]

There was another paper published on the PDO and it's impact on the long term temperature trend by Courtillot et al.

 

From the paper:

 

"Thus, the Earth’s climate may have entered a new mode (a new ~30-yr episode) near the turn of the 20th century: no further temperature increase, a dominantly negative PDO index and a decreasing AMO index might be expected for the next decade or two."

 

 

 

 

I don't know why they are harping over the difference between HadCRUT3 and CRUTem4 at the end of their discussion...don't they know that CRUTem4 is just land only? Its like they used that instead of HadCRUT4 when comparing to HadCRUT3.

[skierinvermont]

I don't know why they are harping over the difference between HadCRUT3 and CRUTem4 at the end of their discussion...don't they know that CRUTem4 is just land only? Its like they used that instead of HadCRUT4 when comparing to HadCRUT3.

 

Pretty embarrassing and revealing commentary if you ask me. Just goes to show that the paper, in an obscure journal, was not carefully peer-reviewed. The aim of the journal is "rapid publication" within 2-4 weeks of submission. 

 

They don't seem to understand that HadCRUT3 is a land-ocean analysis that does not include the arctic, while CRUTem4 is a land-only analysis that does include the arctic. They state that they are puzzled by the increase in monthly variability by a factor of 2 and the fact that the post-1975 temperature anomaly in CRUTem4 is double that of HadCRuT3. This is obviously due to the differences above and the fact that it is an apples to oranges comparison.

 

They are also puzzled by the findings of BEST that there is a discrepancy between HadCRUT, GISS, NOAA, and BEST post- 1980. Of course, if they actually read the BEST paper which they reference instead of just looking at the picture, BEST says that these differences are probably due to differences in how the various sources define "land-only." For example, they believe that GISS defines "land-only" as using data only from land stations, but some of this data is extrapolated over water giving a higher weighting in their analysis to coastal stations which have warmed more slowly. On the other hand, BEST does not extrapolate over water in their land analysis and so their analysis is truly a land-only methodology, and finds more post 1980-temperature increase. 

 

The paper tells us nothing that a quick glance global T timeseries would not tell you. There was warming 1900-1940.. flatline 1940-1970... warming 1970-2000.. and then slow warming after 2000. Fitting a curve to this offers little predictive power without figuring out causation. 

[ORH_wxman]

Pretty embarrassing and revealing commentary if you ask me. Just goes to show that the paper, in an obscure journal, was not carefully peer-reviewed. The aim of the journal is "rapid publication" within 2-4 weeks of submission. 

 

They don't seem to understand that HadCRUT3 is a land-ocean analysis that does not include the arctic, while CRUTem4 is a land-only analysis that does include the arctic. They state that they are puzzled by the increase in monthly variability by a factor of 2 and the fact that the post-1975 temperature anomaly in CRUTem4 is double that of HadCRuT3. This is obviously due to the differences above and the fact that it is an apples to oranges comparison.

 

They are also puzzled by the findings of BEST that there is a discrepancy between HadCRUT, GISS, NOAA, and BEST post- 1980. Of course, if they actually read the BEST paper which they reference instead of just looking at the picture, BEST says that these differences are probably due to differences in how the various sources define "land-only." For example, they believe that GISS defines "land-only" as using data only from land stations, but some of this data is extrapolated over water giving a higher weighting in their analysis to coastal stations which have warmed more slowly. On the other hand, BEST does not extrapolate over water in their land analysis and so their analysis is truly a land-only methodology, and finds more post 1980-temperature increase. 

 

The paper tells us nothing that a quick glance global T timeseries would not tell you. There was warming 1900-1940.. flatline 1940-1970... warming 1970-2000.. and then slow warming after 2000. Fitting a curve to this offers little predictive power without figuring out causation. 

 

 

I agree...its very sloppy in its description/comparison of datasets. Its as if they do not realize that Hadcrut vs Best/CRUTem4 is not an apples to apples comparison. Of course the land-only has more variance.

[Snow_Miser]

Its like they used that instead of HadCRUT4 when comparing to HadCRUT3.

 

I would have to agree. They definitely do think that CRUTem4 is land and ocean, when in reality it is just land. They should have compared it to HadCruT4.

[Jonger]

Funny how the AMO trends with global temps better than the PDO, but it isn't getting much love from the AGW community.

[ORH_wxman]

Funny how the AMO trends with global temps better than the PDO, but it isn't getting much love from the AGW community.

 

 

No it actually doesn't.

 

AMO has a residual ENSO component in it too, so its not 100% exclusive. AMO does much better at trending with arctic temperatures than global temps.

[tacoman25]

Pretty embarrassing and revealing commentary if you ask me. Just goes to show that the paper, in an obscure journal, was not carefully peer-reviewed. The aim of the journal is "rapid publication" within 2-4 weeks of submission. 

 

They don't seem to understand that HadCRUT3 is a land-ocean analysis that does not include the arctic, while CRUTem4 is a land-only analysis that does include the arctic. They state that they are puzzled by the increase in monthly variability by a factor of 2 and the fact that the post-1975 temperature anomaly in CRUTem4 is double that of HadCRuT3. This is obviously due to the differences above and the fact that it is an apples to oranges comparison.

 

They are also puzzled by the findings of BEST that there is a discrepancy between HadCRUT, GISS, NOAA, and BEST post- 1980. Of course, if they actually read the BEST paper which they reference instead of just looking at the picture, BEST says that these differences are probably due to differences in how the various sources define "land-only." For example, they believe that GISS defines "land-only" as using data only from land stations, but some of this data is extrapolated over water giving a higher weighting in their analysis to coastal stations which have warmed more slowly. On the other hand, BEST does not extrapolate over water in their land analysis and so their analysis is truly a land-only methodology, and finds more post 1980-temperature increase. 

 

The paper tells us nothing that a quick glance global T timeseries would not tell you. There was warming 1900-1940.. flatline 1940-1970... warming 1970-2000.. and then slow warming after 2000. Fitting a curve to this offers little predictive power without figuring out causation. 

 

Using an average of various datasets, the most accurate description would be flatlining again. At least since 2001.

 

 

 

 

 

 

[nflwxman]

I've been wondering about the PDO and it's impact on land only temperatures, so I completed an analysis comparing global tropospheric temps versus land only.  Based on the fact the land can't absorb or release energy with significant inertia (La Nina/El Nino) as the oceans can, one would expect the global warming signal to be more stable.  This is not to say the land temperature trend is not impacted by ocean oscillations and other natural factors (as hobbyists, we all know that). However, the lower atmospheric temperature over land is clearly less impacted by changes in natural ocean oscillations versus the lower atmosphere temperatures over the ocean, as shown in the data below.  If you look at the UAH land only temperature record, you will see there is truly not much of a "pause" in the warming trend that you see globally the last decade. This supports that the energy imbalance remains generally the same/higher than what we have seen in the past 2 decades.  Please note, the period between 1992-1994 seems to be dramatically impacted by the Pinatubo explosion, thus I would not start or finish any trends there. Also, UHI is not feasible here as an explanation due to the fact we are measuring the lower troposphere.

 

http://woodfortrees.org/plot/uah-land/from:1979/to:2013/mean:12

 

Land Trends:

 

1979-2005: 0.1519 C/decade

 

1979-2013: 0.181 C/decade

 

1998-2013 ("Pause one"): 0.1901 C/decade

 

2001-2013 ("Pause two"):  0.1926 C/decade

 

Ocean+Land Trends:

 

1979-2005: 0.1336 C/decade

 

1979-2013: 0.1338 C/decade

 

1998-2013 ("Pause one"): 0.0966 C/decade

 

2001-2013 ("Pause two"):  0.002 C/decade 

 

So the question to me is:

 

If more than 40% of the warming in the later half of the 20th century was caused by natural factors, wouldn't you see the land only trend significantly slow down with the global trend (since oceans are now absorbing more heat)?  It appears as if the opposite has happened.  To me, this seems like the ocean right now is the only major driver in masking the upward global trend shown in the land only dataset.

 

Disclaimer: Using trends under 30 years is always a dangerous prospect, but given that the satellite data set is so short, I had no choice.  Perhaps a similar analysis could be completed with the GISS.

[tacoman25]

Good research.

 

I think this supports the notion that the oceans are the primary drivers of global temp trends. After all, 75% of the earth is covered in water. 

 

So you can look at it like the global warming trends are currently being "masked" by the oceans...but that would also mean that a lot of the warming trend in the latter half of the 20th century was enhanced by the transition from -PDO to +PDO phases (1950-1975 -PDO, 1976-2000 +PDO). I don't think then your findings necessarily invalidate the idea that 30-40% of the warming trend 1950-2000 was due to natural factors - considering that warming trend is based on both land/ocean temps. 

 

Just as going from Nina to Nino will not give you an accurate overall trend of warming, going from -PDO to +PDO will do the same. To get an accurate overall trend, I would suggest starting from 10 years into last -PDO phase and ending 10 years into this one...so in about 5 years.

 

Or maybe look at 1946-2007...beginning of last -PDO phase to beginning of this one.