was wet during the downpour.
The bottom flooded.
There has been a lot of news this week. I thought I’d show you a little of our local news. Here’s a video. The tall brick building engulfed in the flooded region is The Devonshire/Westbury building where my Father-in-law spend the last 9 months of his life.
Lisle evacuated those at the bottom of the hill to Benedictine University which is two block from me, and takes up much of the top of the hill. The rain really was rather amazing. I’m high and dry. I did not drive to the bottom of the hill to see this live.
I’m a bit worried about my mother’s lot. She lives in Libertyville, on the banks of the Des Plaines, but is currently on vacation. She travels a lot and has arranged for a neighbor who winters his motor cycle in her garage check; he’d call if he saw a problem. Or so she tells us. . . (He’s a nice guy. And really, Mom has 3 people check. )
But I’m still a bit worried. (Libertyville high school was closed due to flooding on Monday. That school is further north from and the major flooded areas in Libertyville seems to sit between Lake Minear and Butler Lake. Still…. That dang river!!)
Ahh.. the joy of rising rivers. I experienced these growing up right on the banks of the famously flat Des Plaines. Guess why I live on the top of a hill? I hope other readers in Illinois are fine. Let me know if you have any particularly dramatic stories.
I flew into Chicago Ohare for the weekend on Thursday pm last just after the worst of the rain. It was a real treat getting over to Lombard with many of the roads closed due to flooding. Took surface streets as the toll rd was a parking lot and it was still only 3:30pm.
Rush hour must have been interesting if traffic was backed up that early.
While watching the video, an add popped up for Car Hire, and reminded me that I’d forgotten to book a car for my trip to California next week!
So thanks for the reminder!
Ed–Lombard is very near here. Lisle actually got the maximum number of inches of rain. Watching it fall was unbelievable. It was like someone in the sky just opened a firehose and let it start pouring. Fortunately, it stopped.
Then… it all ran downhill… Many places were undriveable. All the low lying bits were closed.
Dear Lucia:
I nearly missed the 200th anniversary of “The Battle of York”.
http://en.wikipedia.org/wiki/Battle_of_York
According to at least one author, this was the starting of the “The Civil War of 1812”
http://www.amazon.ca/The-Civil-War-1812-American/dp/1400042658
With love,
Your true north strong and free neighbour,
AJ
More on topic, rain deluges seems to becoming progressively worse here in Nova Scotia. Pielke jr. says heavier rain, but no worse flooding. Hard for me to reconcile.
It’s granted that living memory is short with respect to climate change. I remember snowfalls that were “up to here”. But, I haven’t experienced any since I grewed up.
Yours, AJ
AJ
You weren’t around for White Juan? They didn’t plow my road for three days…
Nope… was in Moncton at that time (thank God not Lord Monckon). Missed both Juan’s.
Lucky you – I was almost dead center for both.
I’ve missed a lot of natural events. Juan… White Juan… lived here in Halifax when Moncton got 5 feet. Missed the earthquake in northern NB, where I grew up way back when.
I’ve only experienced an ice storm that knocked out my neighbour’s power. Personally, I was unaffected.
Stay high and dry,
AJ
Perhaps an interesting chart from a river in eastern Canada.
http://www2.gnb.ca/content/dam/gnb/Departments/pa-ap/River_Watch/pdf/HistoricWaterLevels.pdf
Between the St. Lawrence and the Mississippi, the Saint John River is the second biggest.
In Fredericton, there seems to be flooding more frequently than in the past.
OT: correlation or causation?
Obviously causation. Judging by the spike in Seattle’s murder rate, it looks like MS is trying to turn the tide 🙂
http://seattletimes.com/html/localnews/2018321758_nationalstats31m.html
Steve
You have it wrong. It is a well known fact that it is ice cream that drives the murder rate.
As ice cream sales increase and decrease on an annual basis, the murder rate follows closely 🙂
http://biojournalism.com/2012/08/correlation-vs-causation/
Another example of correlation conflated with causation is Figure 5 of the new BEST publication:
Figure.
They plot temperature and log(pCO2) against each other, with the obvious suggestion that CO2 is forcing climate change over this entire period.
(The alternative, which I’ll leave unstated, is less flattering. That is, I think this is the most charitable interpretation of their figure.)
The problem wit this is, it is the sum of all radiative forcings that drives climate change.
Comparison of anthropogenic, natural and total forcings for GISS Model E.
I realize this is for a model, but this the current “state of the art” understanding of climate science. Satisfactory or otherwise, it’s the best we have.
If you just look at the contributions to anthropogenic forcings, what you see is this:
Anthropogenic forcings by category & total
So what is wrong with their figure is it erroneously suggests that log(pCO2) is driving temperature change over the entire period 1850-now. This is very much at odds with our current understanding, which is that anthropogenic forcing has only played a significant role since roughly 1970. Prior to that, the dominant driver of climate change was purely natural.
What they’ve done is no better than correlate against number of firemen, lap dogs, pirates etc. Correlating against unphysical variables (ones that have no model based expectation of producing a cause-and-effect relationship) is a pointless activity at best.
Carrick…
The GISS anomalies are in relation to the base year of 1880. According to the Potsdam Institute about 0.5 W/m^2 had already gone into the pipeline by this time.
http://www.pik-potsdam.de/~mmalte/rcps/
My guess is that about 0.1 to 0.2 W/m^2 of this pre-1880 forcing came out of the pipeline post-1880.
Re: AJ (Apr 28 00:16),
I don’t see where you get 0.5 W/m² in 1880. According to the data on the page you linked, total anthropogenic forcing in 1880 was ~0.2 W/m², while total forcing was ~0.4 W/m². The natural forcing was mostly volcanic.
AJ, since we’re discussing climate change, the baseline is irrelevant. so it’s the slope (trend) that matters, not the baseline. You’ll end up with a treatise if you include everything.
To keep this relatively brief, I’m just focus on anthropogenic forcings for this discussion, keeping in mind there’s also natural forcings and unforced changes (internal variability). I use trend of course rather than just change in temperature T2-T1, because it’s a more robust estimator.
The OLS trend anthropogenic forcing, 1910-1950, is only 0.0034 W/m^2/yr compared to its 1970-2010 value of 0.023 W/m^2/yr. The ratio of trends is 0.0034/0.023 = 0.067.
For comparison, GISTEMP temperature trend for 1910-1950 is 0.011 °C/year versus 0.017°C/year for 1970-1010. The ratio of trends is 0.011/0.017 = 0.65.
If we make attribution that the 0.17°C/decade of the post 1970 is dominantly anthropogenic (meaning I’m neglecting the other contributions), very little (roughly 10%) of the trend for 1910-1950 is explained by anthropogenic forcing.
For comparison, GISTEMP temperature trend for 1910-1950 is 0.011 °C/year versus 0.017°C/year for 1970-1010. The ratio of trends is 0.011/0.017 = 0.65.
.
.
a trend of 0.65 going back to the early Middle Ages of 1010 AD 🙂
Good catch, Ed. LOL.
co2 forcings versus all other anthro..
you guys need to update your data
http://tinypic.com/view.php?pic=wwdfs5&s=5
Rohde writes:
“It is a comparison of CO2 radiative forcing to all other anthropogenic radiative forcing using the IPCC assumptions. Or put another way, it is a different way of seeing why CO2 forcing has been a good approximation of total forcing during the last two centuries.
In the IPCC models, the other anthropogenic forcings include both positive (e.g. CH4 and N2O) and negative terms (e.g sulfate aerosols), and it just happens by coincidence that the sum of all such terms other than CO2 roughly cancel in the aggregate. Of course, there are other details such as spatial structure that one wouldn’t get from looking at the global values (e.g. sulfates are much more local than greenhouse gases), and there is no reason that the coincidental cancelling need continue in the future, but during the last two centuries anthropogenic forcing of climate is basically the story of CO2.
Anyway, just a simple striking plot I thought I’d share.”
Steven Mosher,
Here’s the version I’m using now, which is updated through 2011. There are tiny differences between this and the version I plotted, but not enough to bother updating the figures (which were current through 2010). So I think this is pretty up-to-date, as far as published literature goes.
Maybe you should link the source for your graph, and which quantities were included in the anthro portion. I’d also suggest combing them, because it’s the sum of anthropogenic forcings that gives you the relevant physical quantities.
If you don’t I will.
The way it’s plotted is visually deceptive, because it’s hard to work out how much and how large the cancellation really is.
My question is why log(pCO2)? Why not number of left-dominant giraffes?
Neither of these is the relevant variable that forces climate change, so it’s irrelevant whether either of these serve as a good proxy for the variable that actually does.
More to the point, by erroneously plotting log(pCO) and temperature on the same graph, Robert has erroneously communicated a cause and effect relationship that does’t actually exist.
Striking or not, it’s a meaningless comparison.
tiny differences between this and the version I plotted, but not enough to bother updating the figures (which were current through 2010). So I think this is pretty up-to-date, as far as published literature goes.
###############
I dunno. I just went straight for the Ar5 data.
( http://www.pik-potsdam.de/~mmalte/rcps/data/20THCENTURY_MIDYEAR_RADFORCING.xls )
hmm
look something like this
http://tinypic.com/view.php?pic=xqdd4&s=5
“The way it’s plotted is visually deceptive, because it’s hard to work out how much and how large the cancellation really is.”
The point wasnt to work out how much and how large the cancellation was. The point was to show why using C02 as a proxy happened to work. When robert just used c02 everybody wanted to see what happened if you used all the forcings.. answer.. not much.. reason.. everything else pretty much cancels..
“My question is why log(pCO2)? Why not number of left-dominant giraffes?”
Well, because
1. We tried the number of left dominant giraffes and while it was more significant than solar it still didnt explain anything.
2. Why log? we did a non log and log version. big fight.
3. why C02? well, there was theory proposed back in the late 1800s where some guy predicted that if you increase c02 the temperature would go up. So, of the millions of things one could check, it seemed like less than a lark to check C02. plus it was in the data I pointed Robert at.
basically, it went down like this. We are looking at the early part of the record before 1850 and noodling about finding some confirmation that the temperature displayed was not crazy. Zeke ( I recall) brought in A recon ( D’Argio??) and it looked interesting.
I recalled the work that lucia did with Lumpy — you all remember that ( nobody objected ) and I wondered if we could take all the forcings and do a lumpy like she did.. or look at the volcano dips. so I pointed Robert at CMIP data.. looking at all the forcings he found that a simple fit to c02 and volcanos produced a good fit to temperature. Muller went and looked at a bunch of stuff ..like giraffes and other functional forms, and various combinations of forcings etc etc. came back convinced.
So, why C02? well because the goal was to see if there was some additional data/analysis one could do to lend support to accuracy prior to 1850. And one idea was to do something akin to Lucia’s Lumpy, or to focus on the volcano dips. Anyway, Robert went off with the forcing data from cmip and then did his thing and came back with his fit and then muller went off to try to find something wrong with it. So, C02 was used because it was in the forcing data. make sense? Giraffes were not in the cmip forcings. Perhaps they overlooked something.
Now, the disagreements start when folks try to interpret what this means. For me, well, it kinda follows from what we already know, adding GHGs warm the planet. ho hum. To other guys on the team it proved nothing because it was either too simple or was just a correlation. To other folks it was proof positive. Weird. That weirdness, I think, is just part of my basic philosophy: it’s sometimes unclear how evidence applies to a theory which is why I like Quine better than Popper.
“Neither of these is the relevant variable that forces climate change, so it’s irrelevant whether either of these serve as a good proxy for the variable that actually does.”
see, there is that weirdness again. Of course one would like to know the relevant variable that causes climate change and like to see the proof of that. Personally, I didnt see putting C02 and temperature on the same chart as proving anything. We know C02 warms the planet without looking at any temperature data. At least Arrhenius knew that.. So, I see the chart of proof of nothing. its a consequence of what we know not evidence for what we know.
For other folks who start from the assumption that C02 has nothing to do with temperature, they see the chart as proving nothing..
Steven, I’ve got a copy of the AR5 CMIP5 numbers. Some of the choices are a bit weird, at least for an outsider like me.
Regarding log(pCO2), of course you use log(pCO2) because it’s a forcing, not pCO2. That doesn’t really seem like a point to debate.
Regarding left-dominant giraffes.. you’d be surprised which variables correlate with monotonically varying curves.
And anyway, left-dominant giraffes produce methane, so like pCO2 “it’s in there”. 😛 (Yes I thought of that when making up the example. And no I don’t think too much. Thinking “too much” is not a human flaw, rather the opposite.)
Anything that has a high correlation with a desired series makes a good proxy. In this case the desired series is known, so showing that a component of that series correlates well with the sum of its components isn’t useful, isn’t interesting, and is misleading.
Beyond that, I don’t think you fully get it. Newton’s second law states that the sum of the forces equals the mass times acceleration:
$latex F_{net} = F_1 + F_2 + F_3 + F_4 + \dots = m a$.
You are basically finding that $latex F_1$ correlates well with $latex F_{net}$ and are substituting $latex F_1$ for $latex F_{net}$… for no comprehendible reason, because $latex F_{net}$ is the relevant physical quantity and good estimates for it already exist.
It’s weird only in the sense of plotting appropriate variables and not doing things wrong is weird.
Again the chart is misleading. It suggests that CO2 is forcing climate change for the first half of the 20th century. The consensus view of climate science, the “orthodoxy” if you will, suggests that anthropogenically driven climate change became dominant only after 1970.
Beyond that, adding CO2 warms the planet, everything else being equal. Of course, rarely in practice is everything else equal.
If you combine anthropogenic forcings, what is found is that the amount of anthropogenic forcings is reduced for the first half of the 20th century from what it would have been if just anthropogenic CO2 were involved.
This is understandable conceptually from the fact that in the initial development of a technological society, there simply was more pollution per dollar GDP than now, so the relative contribution of anthropogenic cooling to warming was larger.
I just don’t think you can defend publishing an “interesting [but physically meaningless and even misleading] chart” in a science journal. “Personally interesting” is not the threshold for publication.
By the way, I looked at the RCP85 numbers.
Their slopes are:
0.0077 W/m^2/year (1910-1950)
0.030 W/m^2/year (1970-2010)
The ratio is about 0.25.
Again same assumptions as before, about 25% of the warming 1910-1950 would be associated with anthropogenic activity, which might be resolvable even given the very large systematic uncertainties for that period.
The clearest expression of the arithmetic that says non-CO2 forcings balance is the AR5 Fig SPM-2. The error bars show that the conclusion is not entirely robust.
Steven, I’ve got a copy of the AR5 CMIP5 numbers. Some of the choices are a bit weird, at least for an outsider like me.
########################
well, that’s easy to change. do a parametric study
“Regarding log(pCO2), of course you use log(pCO2) because it’s a forcing, not pCO2. That doesn’t really seem like a point to debate.”
well, some folks thought otherwise. go figure.
“Regarding left-dominant giraffes.. you’d be surprised which variables correlate with monotonically varying curves.
And anyway, left-dominant giraffes produce methane, so like pCO2 “it’s in thereâ€. (Yes I thought of that when making up the example. And no I don’t think too much. Thinking “too much†is not a human flaw, rather the opposite.)”
The biggest surprise was seeing that solar did nothing. I don’t think I would be surprised that correlations exist. as in duh.
“Anything that has a high correlation with a desired series makes a good proxy. In this case the desired series is known, so showing that a component of that series correlates well with the sum of its components isn’t useful, isn’t interesting, and is misleading.”
Hmm.. I dont think that C02 is the same as temperature. Now, if I were showing a correlation between say ENSO or PDO and the global series, then I can see your objection.
Beyond that, I don’t think you fully get it. Newton’s second law states that the sum of the forces equals the mass times acceleration:
.
You are basically finding that correlates well with and are substituting for … for no comprehendible reason, because is the #relevant physical quantity and good estimates for it already exist.
Well you lost me there.
“Again the chart is misleading. It suggests that CO2 is forcing climate change for the first half of the 20th century. The consensus view of climate science, the “orthodoxy†if you will, suggests that anthropogenically driven climate change became dominant only after 1970.”
Notice how the chart says nothing about what is dominant. I might have more to say when the next paper ( beyond station quality ) gets put out. Hmm, think of it as this. On time scales less than 60 years you’ll see warming explained by anthro and natural.. at 250 year time scales.. the natural cycles are less than 10% of the overall trend.. there abouts.
“I just don’t think you can defend publishing an “interesting [but physically meaningless and even misleading] chart†in a science journal. “Personally interesting†is not the threshold for publication.”
well, you think its physically meaningless. Other physicists thought it was physically meaningful. That tells me that “meaningfulness” is not determinable by simply “looking” or by simply “claiming”. That people trained in physics can have such diametrically opposed views is philosophically interesting, and also personally interesting, to me at least. I don’t know what experiment you run to show that the chart was physically meaningless, but that might be an interesting experiment. You have a theory that states adding more C02 will warm the planet. If C02 went to 2000ppm and the planet froze, you might consider putting those two on a chart. And then I might say.. putting those two on the same chart is physically meaningless.. but how? Put another way, nothing you can put on a chart is physically meaningful, because charts are just ink, so perhaps you can clarify what you mean by physically meaningful.
Put another way. I get the people who argued “we knew that”
I get the people who argued ” proves nothing”. I get the people who argued “too simple”. I dont get the chewbacca defense.
Nick, um… nobody is arguing that non-CO2 forcings balance CO2 forcings.
Steven Mosher:
Well they were wrong and should get over it. 😛
This is freshman physics.
In GISS Model E, for the period 1910-1950, solar forcings contributes about equally to anthropogenic forcings, so it does “something”.
One of the problems with correlational studies is they neglect that that different frequency components don’t necessarily remain in phase with each other. I’d suggest band-pass filtering and looking at the correlation as a function of center frequency of the band.
No, it is physically meaningless, whether I think it is or not.
This isn’t interpretive dance, this is science.
The relevant variable is net forcing, not just one component of it.
This too is freshman physics.
Random factoid: I have a flag set in my RSS reader for “Chewbacca” (case-insensitive) since I’m the one who popularized the climate blog “chewbacca defense.”
Since I’m here, I’ll contribute something to the conversation. These are the coefficients given by BEST for their linear fit (alpha + beta * log( CO2 / 277.3 ) + gamma * Volcanic):
My quick attempt to replicate their results gives:
Basically the same. It’s well within the error margins they provided (1 sigma – alpha: .05, beta: .25, gamma: .0026). However, what if I do the same fit over just the last 100 years?
The results are similar, but the volcanic forcing shifted almost outside the two-sigma error range. In fact, if one removes the manual manipulation of the volcanic series BEST used to add the El Chicon eruption to the series, the volcanic forcing does shift outside that error range.
What if I don’t use an arbitrary length for the fit? If memory serves, BEST used the period of 1960-2000 for uncertainty calculations due to it having the highest spatial coverage. This is the fit for that period:
If I remove the El Chicon addition again, gamma drops to -0.005854. And for fun, here’s what happens if I start the fit at 1960 but let it run to the end:
Removing the El Chicon addition drops this gamma to -0.006961. I find it interesting to look at how the results change when the fit is restricted to periods with higher quality data. The more restrictive we are, the less volcanic forcings matter. We could make an argument that BEST overstates the significance of volcanoes by 200%.
Incidentally, BEST used their fit to calculate a sensitivity of 3.1 ± 0.3 ºC. If we use the fit from 1960-2000, we get 3.4 ºC. If we use use 1960 on, we get 3.5 ºC. I’m not sure what to make of the fact including a period where there has been no global warming increases the calculated sensitivity.
Carrick,
“Nick, um… nobody is arguing that non-CO2 forcings balance CO2 forcings.”
Indeed. Nor I. The AR4 SPM diagram backs up Steven’s arithmetic.
How did black carbon go from being a -ve to a +ve forcing between comment 112262 and 112271?
And is land use change really a -ve forcing? I thought deforestation had been long known to be a +ve forcing. Isn’t recent reforestation being offered as one of the reasons for the 21st century flat-lining?
Carrick, You mean like solar and ENSO phasing? That does appear to be a factor because of the inconsistent timing of each, but how would you determine an impact? I notice that some time ago and it appears to be the reason the Hale cycle is more pronounced, but once I start mentioning thermal inertia eyes tend to glaze over.
Nick:
If you think so, fine with me.
dallas, I don’t know that it’s fully understood all of the mechanisms by which solar forcing couples to climate.
As an example, longer term forcings produce nonlinear effects including permanent baseline shifts that won’t be exhibited by shorter period forcings. (E.g., albedo changes come to mind as possible an important one.)
Brandon:
That makes no sense. 😉
Carrick, I kind of doubt all the intricacies will ever be completely understood. To me that is the fun part. There is though a shifting of the phase between Solar and ENSO which is a good starting point. 1955 for example, solar was close to 100% out of phase with tropical SST ENSO like conditions. In 1985, the two became in phase with the peak in 1988-89. There is a noticeable step in the NH temperature records that jive with that peak and there is a shift in diurnal temperature range starting circa 1985. I find the DTR shift the more interesting, but “surface” temperature seems to be all the rage.
I think I will look at some ways to compare the phase relationships by regions. That might be interesting.
Carrick, the change isn’t huge or anything, but it does show a problem that arises when interpeting BEST’s results. And there are tons of such problems. For example, by using both interpolation and smoothing, BEST artificially inflated its correlations (and increased the alignment of spectral frequencies).
None of the many issues I could list dramatically change their results,but the fact they exist is shameful. How can BEST publish a result based on shoddy stats like this? Did they not check the effects of various issues? That’d be incredibly lazy. Did they know volcanic forcing changes dramatically based on the period used? If so, not mentioning it is deceptive. Whatever the reason, I’d expect better from a blog post. From a paper in a scientific journal…?
And that’s just the stats. Don’t get me started on the absurdity of how they interpreted the results.
By the way, it’d be interesting to see what the results would be with other forcings (such as solar) considered if they didn’t do their fit on the whole period.
Then we could get all F&R on it…
Brandon, actually the result isn’t necessarily small, but I’ll come back to that.
The actual point I was trying to make was you can’t take one component of a series and neglect the other without strong reasons. The fact that you got a large cancellation (in GISS Model E at least) between the CO2 and other contributions,was just an interesting result so I commented on it, since it points to the folly of only including one of many contributions to total forcing.
By arguing that “the AR4 SPM diagram backs up Steven’s arithmetic”, Nick has essentially argued that because the uncertainty is large, we can assume the contribution from non-CO2 anthropenic forcings is zero. That is a wtf moment if there ever was one… the error is large so that justifies ignoring the effect!?
Back to the relative magnitude of effect, the AR4 SPM diagram Nick linked to doesn’t break it down by period, so it is actually impossible to argue as Nick tries to do that the other forcings besides CO2 can be neglected just from this diagram.
I like to show IPCC SMP Figure 4 which splits out the model runs in terms of natural and natural + anthropogenic.
This allows you to see how adding anthropogenic forcings affects the temperature history, and again what you see is a very large overlap between the forcings for natural only, and natural + anthropogenic…. until 1970. That is, until circa 1970, it is not necessary to invoke anthropogenic forcings to explain the observed variation in global mean temperature.
So I would say this result is actually remarkably robust, since this finding is the result of “58 simulations from 14 climate models”.
Anyway, not only is BEST fundamentally wrong by only plotting against log(pCO2) without any consideration of the other forcings, they are numerically wrong as well.
Another conceptual problem is in trying to estimate climate sensitivity to CO2 from a land-only temperature series. I’m not quite sure how you do that without including the SST contribution.
I think they would have been better off sticking to their primary topic, instead of engaging in this half-a$$ed excursion into attribution.
Carrick, the change was less than 5%. I think it’s fair to say it isn’t huge or anything.
Brandon:
That’s the central value not the range.
Carrick, what point are you trying to make by saying that was the central value, not the range? The range will change basically in line with the central value. The only way it could have a notable change not covered by the change in central value is if it expanded/contracted quite a bit. It didn’t.
DeWitt Payne (April 28th, 2013 at 9:26 am):
Yeah, you’re right, ~0.5 doesn’t quite make sense.
IIRC, I got this figure by taking the difference of the means in total forcing during the volcanic quiet period between 1925-1950. I neglected the fact that GISS shows a faster acceleration in forcing.
Additionally, doing a little back of the envelop calculations, even if the 0.5 figure was correct the total coming out of the pipeline post-1880 would probably have been <0.1 W/m^2.
Brandon, let’s try it this way.
BEST plotted temperature versus log(pCO2). I criticized this as an irrelevant comparison because you should use total net anthropogenic forcings instead of log(pCO2). Under certain circumstances, it’s true that you can neglect other anthropogenic forcings in favor of log(pCO2), but I contend these circumstances are not met in the AR4 SMP.2.
Moreover, I contend that prior to 1970 that one is not required to invoke anthropogenic forcings in order to explain the observed global mean temperature for that period. In the event this claim is correct, BEST’s comparison of T versus log(CO2) isn’t just misleading or inadequate, it’s completely wrong.
Further I claim that SMP2 is not relevant in testing this, because it only shows anthropogenic forcings; it does not compare anthropogenic forcings to other forcings, and it doesn’t show a breakdown for the period 1910-1950, the period I focused on above. Nobody argues that current forcings is dominated by log(pCO2), but this not relevant to my argument.
I pointed to figure SPM.5 as a relevant test of this claim, and demonstrated that this claim is largely supposed by the simulation results in that figure.
Regarding looking only at the 5% difference between total net anthropogenic forcings and log(pCO2), this is not the proper way to test the hypothesis “can I neglect other anthropogenic forcings in deference to log(pCO2).”
The range of uncertainty for total anthropogenic forcings is 0.6 to 2.4 compared to 1.66 W/m2 for CO2 forcing. That’s a range of -60% to +30% in percentile, or 2.4/0.6 = 4x.
Also, the stated uncertainty for CO2 forcing is 1.66 ± 10%.
If probability is high that the total anthropogenic forcings falls within these bounds, the assumption that the other forcings can be neglected is justified.
We can test this proposition by assuming a normally distributed uncertainty centered on the stated central value of 1.6 for total net anthropogenic forcintgs, and computing the probability that the actual value of the total anthropogenic forcings is within the ± 10% range for CO2.
This turns out to be only about 30%. Thus a 70% chance the actual value of the total net anthropogenic forcing is outside of this range, so the assumption that we can neglect the other anthropogenic forcings is not justified.
(A smaller uncertainty bounds “backs up Steven’s arithmetic” not a larger one. )
I think I see what the problem in communication is Carrick. It seems you completely misunderstood what I said. Earlier, I noted an oddity where BEST’s approach finds a higher sensitivity for 1960-now than 1960-2000. That’s peculiar because the period of 2000-now is commonly known to show a pause in warming, and thus, woudl be expected to lower the calculated sensitivity. You responded to this in a humorous fashion.
Upon seeing your response, I wanted to make sure my finding didn’t get overinterpreted. To do this, I posted a comment emphasizing the effect of my finding. I then pointed out there were several additional issues on the same level. I also made the editorial point that these issues were so obvious BEST should never have failed to address them.
None of that has anything to do with what you just said. You seem to be under the impression I was discussing other issues, ones you had brought up before I said anything, when I referred to a less than 5% change. I wasn’t. So when you say:
Please understand I didn’t say anything like that here. I’ve talked about completely different topics.
Since I’m discussing BEST’s results, I should point out I’ve been doing some work analyzing BEST’s overall reconstruction. It stalled because I couldn’t read the data files in the latest BEST release. After a long delay, I got a response from the guy responsible for R’s package for handling Matlab data files. My query to him led him to update his package, and I was able to start reading the files. I now have to figure out how to extract the data properly. It’s a pain because of the formatting, but at least it’s doable now. And quite frankly, I’m much more interested in that than BEST’s infantile curve-fitting.
I’m especially interested as my limited efforts at analyzing BEST’s results show several glaring problems. The newest release by BEST shows an inexplicable spike in uncertainty around 1970 that wasn’t present in the original release. I’m trying to work through BEST’s code to figure out why this happens. but it’s difficult. BEST changed their code between the two releases, and the original code is mostly unreleased. That makes it difficult to figure out what changed. It seems I might need to completely replicate BEST’s process.
Before I get to that though, I still have to figure out an odd puzzle. BEST’s temperature record has a notable seasonal cycle, far stronger than any other modern temperature record. This seasonal cycle exists at all levels. In fact, it seems temperature stations that don’t have a seasonal cycle sometimes have a seasonal cycle introduced by BEST’s code.
I’m not entirely certain of things because I’m testing BEST’s latest code release on its old data set, and I don’t know what changes were made between the two releases.* I’ve also never used Matlab so my attempts at replicating its code are awkward. Plus, there are several different routines in BEST’s code to handle the same thing, and I’m not entirely sure which was used.
Anyway, the point is I’m not spending much time or mental energy on the pathetic curve-fitting BEST did. I’m not even focusing on the newest papers they’ve published, despite the shoddy peer review (I have to assume the typos only slipped through because the journal is pathetic). I think there are much bigger issues to look at. With some luck, I’ll be able to pin them down. Failing that, I might get a response from Robert Rhode. I did e-mail him about a week ago.
Hrm. i still have authorship status on this blog. If I can figure out the BEST seasonality issue, maybe I should write a post about it.
*I’d love to see Mosher, the guy who went on and on about documentation of changes years back, call for BEST to live up to his standards. I suspect he won’t.
Thanks Brandon,
I misunderstood you. It happens that both effects are about 5%…
No prob Carrick. For what it’s worth, I don’t disagree with what you’ve said. I think it is fine to use CO2 as a proxy like BEST has done. We just have to consider how it affects our interpretation of the results. BEST doesn’t do so.
We can account for the effect after the fit is performed. It is (at least theoretically) possible to determine how it influences our results. It’s a complicated matter, and I don’t know exactly how it’d be done, but it is possible.
Basically, we can start from BEST’s result and try to work out a meaningful answer. It’d be complicated and messy, but it might produce something. And that result might allow us to generate some useful information. After all the “maybes” and “mights,” we might be able to get a useful answer.
The end result is BEST provided ****. It’s results might (inadvertently) lead to something, but whatever that may be, it won’t be something provided by BEST itself. It’ll be something we came up with baesd upon BEST’s results.
Looks like I have to correct my correction. Ignore my statement “I neglected the fact that GISS shows a faster acceleration in forcing.” I should have said something like “I neglected to account for different curvatures.”
Mosh – CO2 forcing will only take you so far. Beyond that, and “that” being an easily calculated number, you need additional positive feedbacks to support your gut feeling and IPCC predictions. You don’t know, and I don’t know, and in fact nobody can say with any acceptable degree of certainty what the sign of all the feedbacks is let alone the magnitude. Until that is established your forebodings are just that. Meaningless hand wringing. What we do know with a very high degree of certainty is that the feedbacks do not produce anything like what the IPCC says they should be producing and this is abundantly evident in the growing gap between reality and IPCC predictions.
It is pefectly ok to say you are wrong especially when you are wrong.