Welcome to American Weather
Sign in to follow this  
Followers 0
skierinvermont

Statistical model for removing ENSO, TSI, and Volcanoes

63 posts in this topic

I thought this deserved its own thread so that I can make modifications and so that the discussion of this doesn't get mixed up in several different threads.

I finally got around to statistically removing ENSO, TSI, and Volcano variation from the temperature series 1980-2010. I used the skierinvermontTM temperature index (Average of HadCRUT/GISS 60S-60N with UAH at the poles) which I believe to be one of the most accurate ways to estimate surface temperature. Yearly temperature anomalies were adjusted by -.105C for each 1C of ONI (3 month lag), and by the proposed -.18C for each W/m2 anomaly of TSI found in Camp and Tung 2007. The years 1991, 1992, 1993, and 1994 were adjusted by +.05, +.35 +.22 and +.1C respectively to adjust for Pinatubo. The result of removing the ENSO, TSI, and volcanic variation is a much more consistent temperature series that shows much more steady increase. Obviously not all variation can be removed because of unexplained variation and the interaction of factors, but much of it can be removed by using accepted relationships between ENSO, TSI, volcanoes and temperature which constitute the 3 strongest, most consistent, and predictable short-term effects on temperature.

post-480-0-31834800-1302543564.png

The following graph is ENSO, TSI, and Volcano adjusted (blue line) along with a 5 year smoother (red line).

post-480-0-85592100-1302543618.png

Share this post


Link to post
Share on other sites

1) Where do you get the numbers/formulas to remove these indices? As in, please post peer reviewed paper(s) explaining that indices can indeed be removed, and by how much.

I'm not going into the GISS debate right now, but I'm sure the world knows my feelings on that.

Share this post


Link to post
Share on other sites

1) Where do you get the numbers/formulas to remove these indices? As in, please post peer reviewed paper(s) explaining that indices can indeed be removed, and by how much.

This question is answered in the original post and has been answered a half dozen times already.

The effect of TSI variation comes from Camp and Tung 2007. The effect of ENSO comes from a statistical correlation between ONI and temperature which I performed. The effect of Pinatubo comes from numerous studies which have been posted elsewhere in this forum.

Also, I did not use GISS. I used an average of GISS and HadCRUT between 60S and 60N (where they are nearly identical anyways) and then used UAH for the poles. The reason GISS runs warmer than HadCRUT recently is primarily the poles. Since many people object to GISS's extrapolations at the poles, I replaced them with UAH. Numerous people including tacoman and zucker have agreed this is a reasonable method for estimating surface temperature trends. You are the only person on this forum that continues to object.

Share this post


Link to post
Share on other sites

As numerous people including zucker and I have explained to you the +ENSO tendency caused by AGW is purely speculative and is quite small compared to the short-term ENSO tendencies which we witness from year to year or decade to decade. Removing short-term sharp large ENSO spikes and dips is not the same thing as removing the long term hypothesized +ENSO tendency which is much much smaller and may not even exist.

Share this post


Link to post
Share on other sites

As numerous people including zucker and I have explained to you the +ENSO tendency caused by AGW is purely speculative and is quite small compared to the short-term ENSO tendencies which we witness from year to year or decade to decade.

:arrowhead: Read.

I'm not sure how you cannot understand my post....and then nitpick a small section and turn it into something else. Its about energy balance...its that simple. If the climate system has a positive feedback, it will act on everything.

Heres a skeptical paper on the issue, just for jabs.

http://wattsupwithth...eenhouse-gases/

Also remember the effects of Cloud Cover (GCC), GLAAM in locations of HLB & strength, Global SST.........The BASIS of the ENSO events, the Strength of the PDO....ALL will have effects that will alter the trendline.

As we all know, IPCC states the Earth is loaded with Positive Feedbacks to Forcing Agents......OK, be that as it may, if it were true, any assumption on "Rapid Equilibrium" regarding Solar Activity/TSI is in significant Error, because it would take longer for Earth to come to equilibrium from TSI if the planet was positive feedback prone.

As in.....Yes, 11yr TSI variations may be short term 1-2yr lags, but the overall heat Content coming to equilibrium should take Much Longer if the IPCC is correct about Positive feedbacks within the Climate system.......In regards to TSI, yes, we've seen a leveling off since 1980, however, the 1950-2006 time period is still the Most Active TSI period in our recorded History. So, who says we've come to overall Equilibrium? If the IPCC is correct in regards to positive feedbacks dominating the Climate system, then the TSI overall equilibrium would take quite awhile to be Reached as record TSI for 50 years.

Just Like CO2...Even if we were to stop emitting it now, supposedly, the Earth Would Continue to warm. Since positive feedbacks HAVE to act on the entire climate system, it would have to apply to TSI as well...Only in this Case...TSI Would Have to Return to Pre-Industrial Levels to match the trend of (in relation) "Stopping CO2"!!!

So...when much of that HEAT built up in the Oceans is released during the +PDO phase from 1976-2006, would you expect the excess heat in the atmosphere to be released in torrents? Absolutely!

Share this post


Link to post
Share on other sites

I'm not sure how you cannot understand my post....and then nitpick a small section and turn it into something else.

I could not even begin to respond to the substance of your post until the glaring errors were corrected. The glaring errors were

1) The long-term +ENSO tendency is purely speculative and is much smaller than the extreme year to year and decade to decade fluctuations that we witness. As zucker and I both have explained statistically removing ENSO is a valid approximation.

2) I was not using GISS.

If you are willing to concede these two corrections, then I can respond to the substance of your post.

The substance of your post concerns the response to the TSI fluctuations. You are correct there probably is some long-term response due to the lag caused by the oceans. However, the majority of the response occurs quite quickly. My model cannot remove the additional long-term response caused by the ocean-lag but it does remove the majority of the TSI response which is short-term.

Share this post


Link to post
Share on other sites

As it turns out, Tamino has performed a very similar analysis on his blog:

http://tamino.wordpr...-earth-warming/

His is better than mine because it uses a single regression to attribute variation in temperature to the MEI, sunspots, volcanism, or the long-term trend. Whereas I only performed a regression for ENSO, and then used other studies to select the solar and volcano effect.

It is also quite interesting because he performs the analysis for each of the five major temperature source. This reveals that the ENSO and TSI response is larger for the satellites than for the surface. This explains much of the short term decade-to-decade discrepancies between the satellite sources and surface sources. As seen below:

adj1yr.jpg?w=500&h=325

Because the effect of ENSO and TSI is larger on the satellites than the surface, and TSI has displayed a negative tendency over the 1980-2010 period, the adjustments bring the satellites and surface sources closer. As you can see by comparing the following tables, the discrepancies in the long term trend are smaller for the adjusted data than for the unadjusted, because the satellites are adjusted upwards more.

Raw 1980-2010 trends (C/yr):

GISS 0.0176

NCDC 0.0171

HadCRU 0.0169

RSS 0.0163

UAH 0.0141

Adjusted 1980-2010 trends (C/yr)

GISS 0.0172

NCDC 0.0172

HadCRU 0.0171

RSS 0.0183 (.002C/yr greater!)

UAH 0.0159 (.0018C/yr greater!)

Share this post


Link to post
Share on other sites

Looks like Tamino is publishing this in a peer-reviewed Journal. A similar analysis has already been published (Lean and Rind 2009 in Geophysical Research Letters).

I knew I was onto something laugh.gif

Lean and Rind basically do the exact same thing I've been doing in my posts the past few weeks. In addition they use these relationships to project temperature over the next 10 years.

http://sciences.blog...2009-2019-2.pdf

Share this post


Link to post
Share on other sites

They have avg running mean up near 1998 by 2020, I'll take the under on their predictions all day.

Share this post


Link to post
Share on other sites

I could not even begin to respond to the substance of your post until the glaring errors were corrected. The glaring errors were

1) The long-term +ENSO tendency is purely speculative and is much smaller than the extreme year to year and decade to decade fluctuations that we witness. As zucker and I both have explained statistically removing ENSO is a valid approximation.

2) I was not using GISS.

If you are willing to concede these two corrections, then I can respond to the substance of your post.

The substance of your post concerns the response to the TSI fluctuations. You are correct there probably is some long-term response due to the lag caused by the oceans. However, the majority of the response occurs quite quickly. My model cannot remove the additional long-term response caused by the ocean-lag but it does remove the majority of the TSI response which is short-term.

I will do more research on the +ENSO feedback just in case I'm mistaken.

However everything effects the trend

1)If we could figure out how exactly to adjust for IPO, QDO, PDO, AMO, Global SST, Global Cloud Cover/GCC, LLCC, M/ULCC, Volcanism, IOD, HLB, NAO/AO/AAO, QBO, NAM, Global Sea Ice, Global Snowcover, TSI, and Deep Ocean currents/speed, MJO, all little drivers.....Then we could remove for ENSO easily. Its not Just removing ENSO from the trend...its everything that infects the trend, which Tamino does not do. If we could measure GCC for example, we could probably determine alot more just on the one scale.

2) Can you please UAH for once? Maybe compare UAH & GISS alone, then infill. Your refusal to use it is astounding to me. GISS's stations may be accurate at measurements where they are located...but its the extrapolations that Kill it.

3) Agreed, and this is a huge issue now lately. The issue is the extent of Such Ocean Absorbtion. Some studies suggest smaller amounts of Impact, other suggest major impact. There are Peer Reviewed papers on Both sides.

There is still so much we don't know about our climate system, such as how the various Rays/Magnetism aspect of the Sun potentially affect the Earths internal climate processes.

Share this post


Link to post
Share on other sites

What I find interesting is how much accounting for TSI effects the trend of your graph...and yet until just the past few years, there was little mention and considerable doubt in many climate science circles that the sun could have much discernible effect on temperature trends at all. There has definitely been a bit of a shift in thinking in that regard over the past 3-5 years, no doubt in part because of the flattened temperature trend seen in conjunction with the lower solar activity.

Share this post


Link to post
Share on other sites

What I find interesting is how much accounting for TSI effects the trend of your graph...and yet until just the past few years, there was little mention and considerable doubt in many climate science circles that the sun could have much discernible effect on temperature trends at all. There has definitely been a bit of a shift in thinking in that regard over the past 3-5 years, no doubt in part because of the flattened temperature trend seen in conjunction with the lower solar activity.

That's been the premise and battle cry for a lot of AGW-heavy arguments. When temperatures rise rapidly, its due to CO2, when they don't, then "something is wrong" and there's a scramble to explain the discrepancy. The rising temps are never due to natural factors...ala the late 1970s PDO shift until the early 2000s.

You make that comment about solar/TSI, but also remember, just a few short years ago (less than a decade), most people laughed at the idea of ocean cycles affecting global temps. Now its become more mainstream knowledge.

Share this post


Link to post
Share on other sites

That's been the premise and battle cry for a lot of AGW-heavy arguments. When temperatures rise rapidly, its due to CO2, when they don't, then "something is wrong" and there's a scramble to explain the discrepancy. The rising temps are never due to natural factors...ala the late 1970s PDO shift until the early 2000s.

You make that comment about solar/TSI, but also remember, just a few short years ago (less than a decade), most people laughed at the idea of ocean cycles affecting global temps. Now its become more mainstream knowledge.

Right. And what's really interesting is that now when other factors are considered and included, climate scientists can go: "see, AGW is still happening according to plan, it's just being masked by other factors." But those factors were not considered significant until recent years...and given that many scientists did not forsee this "masking" taking place, what's to say they are understimating other factors influencing our climate in the longer term as well?

So much emphasis has been placed on CO2 as the primary driver, that I think there has been a lack of research into other climate influences. No one can tell me that 10 or 15 years ago hardly any scientists thought we would see a significant slowdown in the warming trend due to lower solar activity or the PDO, because I know that was not the case. It was clearly the mainstream view that CO2 would increasingly overwhelm all other factors. That view has not played out well so far.

Share this post


Link to post
Share on other sites

That's been the premise and battle cry for a lot of AGW-heavy arguments. When temperatures rise rapidly, its due to CO2, when they don't, then "something is wrong" and there's a scramble to explain the discrepancy. The rising temps are never due to natural factors...ala the late 1970s PDO shift until the early 2000s.

You make that comment about solar/TSI, but also remember, just a few short years ago (less than a decade), most people laughed at the idea of ocean cycles affecting global temps. Now its become more mainstream knowledge.

"Mainstream" IPCC tends to ignore ocean cycles though. Maybe their 2014 report will be improved, they stated they're looking heavily at cloud cover in this one, they way Ocean cycles may affect Cloud Cover maybe bring to light some of the issues.

Share this post


Link to post
Share on other sites

To illustrate the amount of uncertainty involved here: http://news.national...-space-science/

:lol:

So wait, lower solar activity may actually increase warming and "this is not good news for skeptics"? So what on earth has slowed warming over the past decade then, if both CO2 and lower solar activity favored increased warming?

If these SORCE guys are right, then its terrible news for IPCC more-so than skeptics, regarding the flat-lining in warming.

Although they didn't say how much it has boosted the warming.

Share this post


Link to post
Share on other sites

Right. And what's really interesting is that now when other factors are considered and included, climate scientists can go: "see, AGW is still happening according to plan, it's just being masked by other factors." But those factors were not considered significant until recent years...and given that many scientists did not forsee this "masking" taking place, what's to say they are understimating other factors influencing our climate in the longer term as well?

So much emphasis has been placed on CO2 as the primary driver, that I think there has been a lack of research into other climate influences. No one can tell me that 10 or 15 years ago hardly any scientists thought we would see a significant slowdown in the warming trend due to lower solar activity or the PDO, because I know that was not the case. It was clearly the mainstream view that CO2 would increasingly overwhelm all other factors. That view has not played out well so far.

This is the beef that most skeptics have with a lot of the AGW theory. We have a fairly solid idea of what CO2 does in a vacuum. The problem, is that the climate system is quite a bit more complex than that. CO2 contributes warming by itself. That is well understood. What has been a problem is that for years we were told that stuff like solar variation and ocean cycles were extremely minor and didn't matter. Now they will say the same thing of course now, but just stretch the time line to a century or so. But just a few short years ago, even on the decadal scale they were not suppose to be a big factor.

We still don't really have much of a handle at all on the solar stuff. Especially the longer term solar stuff (i.e. extended periods of high solar activity ala the 20th century vs extended period of low activity ala the LIA). What if solar cycle 25 is like cycle 24? How much of a longer term affect will that begin to have?

Share this post


Link to post
Share on other sites

This study illustrates the uncertainty and possible disconnect between theoretical estimations of solar forcing and empirical results.

http://www.acrim.com...e%20warming.pdf

That paper contains a number of serious methodological errors. Benestad and Schmidt take Scafetta and West to the woodshed. They perform a very thorough analyses of the Scafetta and West methodology and definitively reject it.

http://www.leif.org/...tad_Schmidt.pdf

We also repeated the analyses of Scafetta and West, together with a series of sensitivity tests to some of their arbitrary choices. These tests showed clearly that the published uncertainty in their estimates was greatly underestimated. In particular, the arbitrary assumption of their equilibrium sensitivity (Zeq) has a dramatic impact on their attribution of 20th century changes to solar forcing. We next showed that their methodologies were not able to robustly retrieve the solar contribution in GCM experiments where the answer was known a priori. In fact, we found that the presence of internal variability and additional forcings greatly confounded their method's accuracy. Even in much simpler cases, examined here using Monte Carolo simulations of synthetic climate time series, we found that their diagnostics had a very wide range in the absence of a true signal, so cannot be considered a robust metrics of solar-induced contribution.

We conclude that as with the simpler linear regression methodologies described earlier the [scafetta and West] methodology is highly sensitive to internal variability of the climate system and the presence of colinear trends in different forcings. Given the concomitant increases in greenhouse gas forcings over the 20th century, this implies that their published attributions greatly exaggerate the role of solar variations in global mean temperature trends.

Claims that a substantion fraction of post 1980 trends can be attributed to solar variations are therefore without solid foundation, and solar-related trends over the last century are unlikely to have been bigger than .1 to .2C.

Share this post


Link to post
Share on other sites

In addition to my above post on Scafetta and West...

Bernestad and Schmidt also critique the methods of Lean and Rind 2008 (who basically do the same thing I did in the original post) as well as Camp and Tung 2007 (the solar study I relied upon).

I will probably redo my model based on the valid criticisms presented by Bernestad and Schmidt.

Share this post


Link to post
Share on other sites

Legras et al. 2010 makes similar criticisms of simple regression techniques and reaffirms the rejection of Scafetta and West by Bernestad and Schmidt.

http://sciences.blog...s10climpast.pdf

"It has been demonstrated that a correlation analysis which takes only one cause into account can lead to a spurious attribution (e.g. Scafetta and West, 2006a,b, 2007, criticized by Bernestad and Schmidt 2009)."

This is probably the simplest way to explain the basic flaw in the Scafetta and West method. The essentially assume that climate variability is caused by solar and then perform a correlation to that ONE variable, which as anybody who is familiar with statistics should know, will cause them to attribute far more variability to that one variable than if we tested against, 2,3, 5, 10+ variables at once. A proper statistical test includes ALL possible variables involved, not just one. Even then there may be problems (see Bernestad and Schmidt's critique of Lean and Rind). But it's still better than the egregious error committed by Scafetta and West by only including ONE variable in their regression (solar). The Scafetta and West study should be ignored.

Share this post


Link to post
Share on other sites

The Data by SORCE Satellite, however, is not theory, it is measurement. So even though there is no proof the sun always behaves this way, an Increase in Visible Light would boost warming from the Sun, not Reduce it.

Its theories that change often, but its good to see "Skierinvermont" going against the IPCC consensus :thumbsup:

Even at risk of Further warming from the Oceans releasing more heat Content due to high TSI as recenly as the early 2000's. The slight decline in TSI since 1980 is decieving since all of those Solar Cycles were Strong enough to be considered "Hyperactivity" By NOAA. The warming impact from those could easily Impose another +.2C/decade when coupled with the +PDO and +ENSO dominated eras from 1977-2007

Not even going into Global Cloud Cover which could have caused significant warming as well! So if SORCE is correct about warming impact from the Sun from 2004-2007, there is more to this.

If we could figure out how exactly to adjust for IPO, QDO, PDO, AMO, Global SST, Global Cloud Cover/GCC, LLCC, M/ULCC, Volcanism, IOD, HLB, NAO/AO/AAO, QBO, NAM, Global Sea Ice, Global Snowcover, TSI, and Deep Ocean currents/speed, MJO, all little drivers.....Then we could remove for ENSO easily. Its not Just removing ENSO from the trend...its everything that infects the trend, which Tamino does not do. If we could measure GCC for example, we could probably determine alot more just on the one scale.

Share this post


Link to post
Share on other sites

That paper contains a number of serious methodological errors. Benestad and Schmidt take Scafetta and West to the woodshed. They perform a very thorough analyses of the Scafetta and West methodology and definitively reject it.

http://www.leif.org/...tad_Schmidt.pdf

We also repeated the analyses of Scafetta and West, together with a series of sensitivity tests to some of their arbitrary choices. These tests showed clearly that the published uncertainty in their estimates was greatly underestimated. In particular, the arbitrary assumption of their equilibrium sensitivity (Zeq) has a dramatic impact on their attribution of 20th century changes to solar forcing. We next showed that their methodologies were not able to robustly retrieve the solar contribution in GCM experiments where the answer was known a priori. In fact, we found that the presence of internal variability and additional forcings greatly confounded their method's accuracy. Even in much simpler cases, examined here using Monte Carolo simulations of synthetic climate time series, we found that their diagnostics had a very wide range in the absence of a true signal, so cannot be considered a robust metrics of solar-induced contribution.

We conclude that as with the simpler linear regression methodologies described earlier the [scafetta and West] methodology is highly sensitive to internal variability of the climate system and the presence of colinear trends in different forcings. Given the concomitant increases in greenhouse gas forcings over the 20th century, this implies that their published attributions greatly exaggerate the role of solar variations in global mean temperature trends.

Claims that a substantion fraction of post 1980 trends can be attributed to solar variations are therefore without solid foundation, and solar-related trends over the last century are unlikely to have been bigger than .1 to .2C.

So this proves that two scientists disagree with the methodology of two other scientists.

Share this post


Link to post
Share on other sites

Its theories that change often, but its good to see "Skierinvermont" going against the IPCC consensus :thumbsup:

Not that I haven't, but I don't know what you are specifically referring to.

Even at risk of Further warming from the Oceans releasing more heat Content due to high TSI as recenly as the early 2000's. The slight decline in TSI since 1980 is decieving since all of those Solar Cycles were Strong enough to be considered "Hyperactivity" By NOAA. The warming impact from those could easily Impose another +.2C/decade when coupled with the +PDO and +ENSO dominated eras from 1977-2007

This is physically impossible, as explained by numerous attribution studies including the Bernestad and Schmidt study I posted earlier. There is no physical mechanism whereby once solar activity stops rising, temperatures would continue rising rapidly, aside from a slight lag due to the oceans which would be a relatively minor factor. The atmosphere is fairly tansparent to energy and will rapidly lose or gain energy based on the radiative balance. Unless the incoming energy is increasing, or the outgoing energy decreasing, there aren't going to be large changes in temperature. The atmosphere rapidly warms or cools to equilibrium in response to external forcings. Once solar stops increasing, which it did mid-century, then so does the warming effect.

If we could figure out how exactly to adjust for IPO, QDO, PDO, AMO, Global SST, Global Cloud Cover/GCC, LLCC, M/ULCC, Volcanism, IOD, HLB, NAO/AO/AAO, QBO, NAM, Global Sea Ice, Global Snowcover, TSI, and Deep Ocean currents/speed, MJO, all little drivers.....Then we could remove for ENSO easily. Its not Just removing ENSO from the trend...its everything that infects the trend, which Tamino does not do. If we could measure GCC for example, we could probably determine alot more just on the one scale.

As numerous people have told you multiple times, including zucker and myself, the ENSO response is very consistent and predictable and one can arrive at a decent approximation by removing it. It's quite clear that we warm in Ninos, and cool in Ninas, and that most of this effect can be removed so as to approximate the ENSO-neutral trend.

Share this post


Link to post
Share on other sites

So this proves that two scientists disagree with the methodology of two other scientists.

And if you read what they say, it's quite obvious that their criticism is valid. Scafetta and West assume the climate to be driven by solar a priori and then conduct a regression against that ONE variable. This will ALWAYS tend to overestimate the influence of that variable. A regression MUST include all possible related variables.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0

  • Recently Browsing   0 members

    No registered users viewing this page.