Register to reply

Ocean Heat Storage

by joelupchurch
Tags: heat, ocean, storage
Share this thread:
joelupchurch
#1
May5-09, 01:55 PM
P: 149
This paper "Earth's energy imbalance: Confirmation and implications. Science, 308, 1431-1435 Hansen et al" pdf available http://pubs.giss.nasa.gov/abstracts/...en_etal_1.html seems to have been very influential in research on climate change.

The basic argument is that very little of the earth's heat is stored in the atmosphere and that the heat stored in the first 2.5 meters of the ocean is equivalent to the whole atmosphere. Therefore ocean heat storage is a more reliable tool to measure the radiative imbalances in our climate system than surface temperature changes.

I've noticed that even researchers at the far end of climate debate from Hansen are starting to use the ocean heat storage metric in their papers.

Part of this is probably driven by deployment of the Argo float system that has given us unprecedented capacity to measure changes in ocean temperatures down to 2000 meters.
http://www.argo.ucsd.edu/ . There are currently 3325 active Argo floats providing real time information about ocean conditions.
Phys.Org News Partner Earth sciences news on Phys.org
chriscolose
#2
May5-09, 05:38 PM
P: 59
The recent Levitus et al 2009 paper may be a good read
ftp://ftp.nodc.noaa.gov/pub/data.nod.../grlheat08.pdf
sylas
#3
May5-09, 06:09 PM
Sci Advisor
sylas's Avatar
P: 1,750
Quote Quote by joelupchurch View Post
This paper "Earth's energy imbalance: Confirmation and implications. Science, 308, 1431-1435 Hansen et al" pdf available http://pubs.giss.nasa.gov/abstracts/...en_etal_1.html seems to have been very influential in research on climate change.

The basic argument is that very little of the earth's heat is stored in the atmosphere and that the heat stored in the first 2.5 meters of the ocean is equivalent to the whole atmosphere. Therefore ocean heat storage is a more reliable tool to measure the radiative imbalances in our climate system than surface temperature changes.

I've noticed that even researchers at the far end of climate debate from Hansen are starting to use the ocean heat storage metric in their papers.

Part of this is probably driven by deployment of the Argo float system that has given us unprecedented capacity to measure changes in ocean temperatures down to 2000 meters.
http://www.argo.ucsd.edu/ . There are currently 3325 active Argo floats providing real time information about ocean conditions.
This is really interesting stuff… and with lots of unanswered questions which adds to the interest!

It's long been known that the ocean's heat capacity is important; the problem has always been the computation to take it into account.

The equivalence of heat storage for the atmosphere to about 2.5 meters of ocean is a back of the envelope calculation, using century old physics. Heat capacity of air is about 1000 kJ/kg; and for water is about 4000 kJ/kg. (Single digit accuracy) The atmosphere has a pressure of 1000 hPa, under gravity of 10 m/s^2; hence there is about 10,000 kg of atmosphere for each square meter of Earth's surface. That has the heat capacity of 2,500 kg of water. You get 2,500 kg of water per square meter with a depth of 2.5 m.

Strictly speaking, this isn't a comparison with the whole ocean, but with atmosphere and ocean over a single square meter. To consider the whole Earth, you'd have to note that about 30% of the surface isn't ocean. Taking this into account, the heat capacity of the entire atmosphere is about the same as the heat capacity of the top 3.4m of the world ocean.

That only tells you that the ocean is important, which has long been understood. The modeling of climate and weather has progressed steadily thanks to more detailed measurement of factors involved, better mathematical descriptions of the physics, and faster computers to actually handle the computations. A really helpful background historical summary of the progress in quantified understanding of weather and climate over the last century is at General Circulation Models of Climate, a chapter out of Spencer Weart's online book on the discovery of global warming.

The paper you have cited by Hansen et. al. has been influential, not so much for persuading people to consider the ocean, but for a quantified estimate of it's impact in terms of the net energy flow currently going into the ocean.

However, in my opinion the estimate given there is a substantial over estimate. They quote 0.85 W/m^2 +/- 0.15 net imbalance for the whole planet. But if you read the paper, it's clear this number is not a measurement, but based on the models. The actual measurements do indicate a positive imbalance, but substantially less. Climate models have been an enormous help in this whole field, but modeling of ocean circulation still has a ways to go before I'd put too much weight on this specific number.

The research Chris has cited is a better guide, I think. The Argo floats you mention are a key part of this – although much of the issue there is sorting out teething problems with systematic measurement errors.

Bottom line: the Leviticus et al paper estimate increasing heat content of the world ocean at roughly 4e21 J/year over recent decades, and this fits broadly with a range of empirical estimates.

Crunch the numbers: a year is 3.15e7 seconds, and the Earth's surface is 5.1e14 m^2, and this works out to about 0.25 W/m^2 as the energy imbalance.

I think this is a much more plausible value than the 0.85 used in Hansen et al (2005).

Cheers -- sylas

joelupchurch
#4
May5-09, 11:48 PM
P: 149
Ocean Heat Storage

Quote Quote by chriscolose View Post
The recent Levitus et al 2009 paper may be a good read
ftp://ftp.nodc.noaa.gov/pub/data.nod.../grlheat08.pdf
I'm not comfortable with how they mix in the lower quality pre argo data. There seems to be a sharp change in the trendline around 2004. Is that real or just because they are mixing apples and oranges?
sylas
#5
May6-09, 01:03 AM
Sci Advisor
sylas's Avatar
P: 1,750
Quote Quote by joelupchurch View Post
I'm not comfortable with how they mix in the lower quality pre argo data. There seems to be a sharp change in the trendline around 2004. Is that real or just because they are mixing apples and oranges?
I'm not sure what you refer to. I've got the paper; you can refer to a figure or a page. I've not read it in great detail. I know some of the issues involved in this, but only from comparatively superficial reading of the associated literature. I'm not an expert.

One of the problems has been systematic errors in Argo measurements. Most of that is probably ironed out, but it's an ongoing issue to sort out problems with Argo. In fact, I see that just yesterday there was a recall notice put out at the Argo site wanting to stop float deployment and bring in floats for repair to pressure sensors. Pressure sensors seem to have been something of a problem all along.

This is a great experiment, but they are still sorting it all out as far as I can see. It's got great potential to help sort out ocean measurement.

A part of this paper (Levitus et al 2009) has been to consider the impact of corrected Argo data from earlier systematic errors. They also seem to be working on identifying systematic errors in older "bathythermograph" measurements that I don't know about, but which are a crucial part of the longer instrument record.

Argo is still new, so it can't tell you much about long term trend by itself. There certainly is real short term variation; there is also short term sampling errors. When you are dealing with incomplete data the short term variation is often hard to pin down. Furthermore, it's really the long term trend people are most interested in, at this point.

As a general rule in ANY measurement of a trend of anything, you can't draw conclusions from a short sample. So sure, there is likely to be some real changes from year to year. The idea of a trend is to get a longer term picture that is not distorted by short term natural oscillations and variations.

The other important recent reference is:That is an analysis of the instrument record up to 2003, but it is by the Argo floats research group and does use this new source of information. They obtain a trend in ocean heat content involving 16 +/- 3 *1022 J from 1961 to 2003; and that corresponds to a planetary energy imbalance of 0.24 W/m2 – very similar indeed to Levitus et al (2009).

Cheers -- sylas
joelupchurch
#6
May6-09, 07:25 PM
P: 149
Quote Quote by sylas View Post
I'm not sure what you refer to. I've got the paper; you can refer to a figure or a page. I've not read it in great detail. I know some of the issues involved in this, but only from comparatively superficial reading of the associated literature. I'm not an expert.
I'm referring to figure 1 on page 2. From 2000 to 2004 it increases from 7x1022 to 13x1022 joules and it then flattens out.

I just did a back of the envelope calculation based on the ocean storing 1.5x1022 joules a year. I figure that the human race uses about 5x1020 joules per year, so the ocean was storing 30 times as much energy as the human race was consuming.

http://tonto.eia.doe.gov/ask/generalenergy_faqs.asp. There are 1055 joules in a BTU.
Xnn
#7
May6-09, 07:55 PM
P: 555
— The global mean concentration of CO2 in 2005 was 379
ppm, leading to an RF of +1.66 [±0.17] W m–2.
The combined anthropogenic RF is estimated to be +1.6
[–1.0, +0.8] W m–2, indicating that, since 1750, it is extremely
likely that humans have exerted a substantial warming
influence on climate.
Increasing concentrations of the long-lived greenhouse
gases (carbon dioxide (CO2), methane (CH4), nitrous oxide
(N2O), halocarbons and sulphur hexafl uoride (SF6); hereinafter
LLGHGs) have led to a combined RF of +2.63 [±0.26] W m–2.
From page 131. RF = Radiative Forcing

http://www.ipcc.ch/pdf/assessment-re...1-chapter2.pdf
joelupchurch
#8
May6-09, 08:45 PM
P: 149
The other important recent reference is:That is an analysis of the instrument record up to 2003, but it is by the Argo floats research group and does use this new source of information. They obtain a trend in ocean heat content involving 16 +/- 3 *1022 J from 1961 to 2003; and that corresponds to a planetary energy imbalance of 0.24 W/m2 – very similar indeed to Levitus et al (2009).
[/QUOTE]

I found a free copy of the paper here http://www.astepback.com/GEP/Nature%...LR%20rates.pdf.

I'm not sure how much to read into their agreement with Levitus. It looks like they are working from the same datasets and may be using the same corrections.

I don't know why they used the pre 1970 data. The 1 standard deviation band is so wide, that they might as well insert random numbers. I'm also a little concerned that they cut off their graph at 2003 when Leviticus includes data through 2008.
joelupchurch
#9
May6-09, 08:53 PM
P: 149
Quote Quote by sylas View Post
The other important recent reference is:That is an analysis of the instrument record up to 2003, but it is by the Argo floats research group and does use this new source of information. They obtain a trend in ocean heat content involving 16 +/- 3 *1022 J from 1961 to 2003; and that corresponds to a planetary energy imbalance of 0.24 W/m2 – very similar indeed to Levitus et al (2009).
Sorry I messed up. The part above was from a post by sylas not me.
sylas
#10
May6-09, 09:06 PM
Sci Advisor
sylas's Avatar
P: 1,750
Quote Quote by joelupchurch View Post
I'm referring to figure 1 on page 2. From 2000 to 2004 it increases from 7x1022 to 13x1022 joules and it then flattens out.
OK. I see what you mean. Here's the same data as presented at the NOAA page for this research: Global Ocean Heat Content:

The data being plotted here and in figure 1 of Levitus et al (2009) is in the "World Ocean" column of the file: h22-w0-700m1-3.dat (ftp, NOAA site). The rise you mean is from the data point for 2001 (5.073e22 J) to 2004 (12.154e22 J). You can also get at slightly more detailed 3-monthly data from the same pages.

This rise around 2003 is a part of the Argo data; it is not something that comes from trying to merge older data with new "better" data. I don't think you can draw any definite conclusions about a short term effect like this; it could be almost anything, from a calibration error to a real heat content jump to a measurement artifact of how floats are distributed with changing currents. The overall picture of increasing heat content, however, is pretty unambiguous. There are various short term oscillations in ocean currents and temperature distributions, like the ENSO oscillations with El Nino and La Nina. That's a real effect – not simply as a change in total heat content, but signal that can confound short term measures of heat content.

In this data the whole trend is a strong positive, and there's shorter noise and up and down all along the timeseries. In the Domingues et al paper (2008) that I cited, you can see their figure 1 also showing a similar up surge. Since it is the same data being analysed, this isn't enough to tell how much is an artifact of measurement, and how much is a real difference in shift. The data files allow you to look at North and South Hemisphere separately, and at a 3-month series rather than annual data, and it looks a bit like there was at least some role for a fortuitous case of simultaneous upswings in both NH and SH throughout the year 2003 and a bit each side… but either NH or SH in isolation looks more like ongoing noisy variation and an increasing trend.


I just did a back of the envelope calculation based on the ocean storing 1.5x1022 joules a year. I figure that the human race uses about 5x1020 joules per year, so the ocean was storing 30 times as much energy as the human race was consuming.

http://tonto.eia.doe.gov/ask/generalenergy_faqs.asp. There are 1055 joules in a BTU.
That sounds about right. The contribution of human activity by direct energy production is just about negligible. As far as the Earth is concerned, the only source of energy that matters is the Sun. That's why the greenhouse effect is so important. This isn't a source of energy – our actual energy input is tiny by comparison. The greenhouse effect has such a big impact because it alters how the Earth responds to the solar energy input.

Cheers -- sylas
sylas
#11
May6-09, 09:29 PM
Sci Advisor
sylas's Avatar
P: 1,750
Quote Quote by joelupchurch View Post
I'm not sure how much to read into their agreement with Levitus. It looks like they are working from the same datasets and may be using the same corrections.
Of course. The thing is that this IS Argo data being used.

I don't know why they used the pre 1970 data. The 1 standard deviation band is so wide, that they might as well insert random numbers. I'm also a little concerned that they cut off their graph at 2003 when Leviticus includes data through 2008.
It takes a lot of time and work to do these analyses. You can't just read off a value from Argo data. A global heat content figure is obtained by a lot of analysis of thousands of individual floats, along with all the issues of calibration and instrument drift and known errors with pressure sensors. Hence a paper submitted in 2007 is using data from some time prior to that.

There's a slightly different focus in the two papers Levitus (2009) is particularly concerned with correcting errors in XBTs, whereas Domingues (2008) is particularly concerned with fixing problems with recently discovered problems with the Argo floats. The Levitus paper is a repeat of an older analysis, and has been updated to take into account the new Argo corrections amongst other things; and I guess they rely upon the Argo team for the corrections to systematic Argo errors. It appears so from the acknowledgements of the papers.

I don't know how you conclude pre 1970 data is that bad. It certainly does not appear so to my analysis. The natural variation in measurement is very similar all along the series. You do have less trend prior to 1970, which is unsurprising. All the related data, on sea surface measures and sea level and so on, all indicate that the ocean has been warming substantially more since 1970 or so than prior. That's pretty definite.

We'd like to know how much... and so pre-1970 data is of considerable interest. But the figure does not show any greater "randomness" before 1970 than after, and it fits well with other related data, like the SST record. I also ran a quick check of my own using confidence bounds based on the data only and a simple regression analysis. There's a little bit of extra amplitude to the short term variations through the 1980s, but only very minor and probably not significant. There's no indication at all of additional randomness prior to 1970.

Cheers -- sylas
joelupchurch
#12
May6-09, 11:31 PM
P: 149
Quote Quote by sylas View Post
Of course. The thing is that this IS Argo data being used.
I don't know how you conclude pre 1970 data is that bad. It certainly does not appear so to my analysis. The natural variation in measurement is very similar all along the series. You do have less trend prior to 1970, which is unsurprising. All the related data, on sea surface measures and sea level and so on, all indicate that the ocean has been warming substantially more since 1970 or so than prior. That's pretty definite.

Cheers -- sylas
Ignore the data points in figure 1 and look at the gray band marking the 1 std dev. If they are using it the way I'm familiar with, then the line is the measured value and the band represents where the is a 75% chance that actual value is. The pre 1970 data band is almost 3 times wider than the post 1970. You can fit any trend you want inside that band.

This assumes the data actually following a Gaussian distribution.
sylas
#13
May7-09, 04:36 AM
Sci Advisor
sylas's Avatar
P: 1,750
Quote Quote by joelupchurch View Post
Ignore the data points in figure 1 and look at the gray band marking the 1 std dev. If they are using it the way I'm familiar with, then the line is the measured value and the band represents where the is a 75% chance that actual value is. The pre 1970 data band is almost 3 times wider than the post 1970. You can fit any trend you want inside that band.

This assumes the data actually following a Gaussian distribution.
Ah! My apologies… I was still looking at figure 1 in Levitus et al (2009). In the Levitus analysis the error bars are not actually plotted, but they can be found at the download site.

We need to bear in mind that everything in these figures is a calculation; including the error bars. There is no such thing as a direct measurement of the ocean's total heat content – it has to be inferred on the basis of many thousands of measurements and a whole heap of analysis. Both Levitus (2009) and Domingues (2008) are papers that are clearing up systematic errors in earlier measurements.

Here is a copy of Figure 1 from Domingues (2008):
(image source: realclimate blog).
What this shows are two older calculations and the new revision that Domingues et. al. have obtained, in black. The error bars shown here represent the one standard deviation of the Domingues et. al. analysis, and this does indicate that the analysis used is degraded for trends prior to 1970.

One of the lines there in the Domingues paper, given in red, is from Levitus (2005).

The first paper we’ve been using in this thread is Levitus (2009), introduced by Chris Colose. This is mainly concerned with cleaning up errors in the XBT datasets, and updating their previous and erroneous 2005 analysis. The figure 1 in Levitus (2009) does not include the error bars, but it does give a side-by-side comparison with their 2005 analysis, as shown in red in Domingues (2008). There are error bars in the datasets that I linked to previously. Hence I can plot them for you. This plot shows the annual OHC for the Levitus (2009) analysis, along with hi and lo limits corresponding to one standard error. The baseline is a bit different to the published figure… this is simply a somewhat arbitrary choice for a reference zero point.

I've also plotted here the width of the standard error calculation independently above the main data plot, which gives a better idea of how the standard error develops over time.

There are different error bars here in Levitus (2009) because they are different datasets and different calculations. Levitus is working mainly with XBT data. This is reasonably seen as a reduction in errors associated with the older dataset; but it is also likely that the error bars represent something a bit different in this case… the method error without explicit additional measurement error. That's fair enough in a downloaded dataset; it's probably method error that's more useful for other analysis. However, it remains the case that the two analyses by Levitus and Domingues are different analyses with their own associated error estimates. The 2009 analysis from Levitus et al is mostly somewhere in between their 2005 analysis and the Domingues analysis.

I don't have direct access to the Domingues et al data -- it is available, but not as a direct download, and I haven't put in an order for it by email. But it looks like the big hump from the Levitus 2005 analysis is mostly gone in 2009. What's left of that hump still represents a difference between the two analyses.

The Domingues 2008 error bars most probably include a larger measurement error component, not explicitly addressed in Levitus 2009. That's how I read it, in any case.

A great summary of this ongoing work is available at the NASA earth observatory: Correcting Ocean Cooling. It tells the story of how the errors in the Argo floats were found and fixed, and gives a good insight into how difficult it all is. Although the article focuses on the Argo team, the final word of the article goes to Syd Levitus:
"Models are not perfect," says Syd Levitus. "Data are not perfect. Theory isn't perfect. We shouldn't expect them to be. It's the combination of models, data, and theory that lead to improvements in our science, in our understanding of phenomena."
Cheers -- sylas
Saul
#14
May7-09, 09:39 PM
P: 272
It appears the oceans have stopped warming. I am not sure that is cause for celebration.


http://www.ingentaconnect.com/conten...nlpdhr7c.alice

Ocean heat content data from 2003 to 2008 (4.5 years) were evaluated for trend. A trend plus periodic (annual cycle) model fit with R2 = 0.85. The linear component of the model showed a trend of −0.35 (±0.2) × 1022 Joules per year. The result is consistent with other data showing a lack of warming over the past few years.
Author: Loehle, Craig

Source: Energy & Environment, Volume 20, Numbers 1-2, January 2009 , pp. 101-104(4)

Cooling of the global ocean since 2003

http://www.npr.org/templates/story/s...oryId=88520025
joelupchurch
#15
May8-09, 09:34 AM
P: 149
Quote Quote by Saul View Post
It appears the oceans have stopped warming. I am not sure that is cause for celebration.


http://www.ingentaconnect.com/conten...nlpdhr7c.alice



Author: Loehle, Craig

Source: Energy & Environment, Volume 20, Numbers 1-2, January 2009 , pp. 101-104(4)

Cooling of the global ocean since 2003

http://www.npr.org/templates/story/s...oryId=88520025
I found a copy of the paper here:

http://www.ncasi.org/publications/Detail.aspx?id=3152

As far as I can tell, there is no data in the paper that isn't in the Leviticus paper we have been discussing. Frankly, I think, my earlier characterization of the data as essentially flat is more accurate. The trend they refer to overwhelmed by seasonal variations.

I didn't think that was remarkable, since it was consistent with the surface temperature record with some lag.
Saul
#16
May8-09, 10:46 AM
P: 272
Quote Quote by joelupchurch View Post
I found a copy of the paper here:

http://www.ncasi.org/publications/Detail.aspx?id=3152

As far as I can tell, there is no data in the paper that isn't in the Leviticus paper we have been discussing. Frankly, I think, my earlier characterization of the data as essentially flat is more accurate. The trend they refer to overwhelmed by seasonal variations.

I didn't think that was remarkable, since it was consistent with the surface temperature record with some lag.
A trend of minus -0.35 x 10^22 Joules/year for the period 2003 to 2008 is not essentially flat. A negative trend in the planet’s ocean temperature shows the planet is cooling. A clearer description than "a lack of warming of over the last few years" is the planet was been cooling from the period 2003 to present.

Ocean heat content data from 2003 to 2008 (4.5 years) were evaluated for trend. A trend plus periodic (annual cycle) model fit with R2 = 0.85. The linear component of the model showed a trend of -0.35 (±0.2) x 10^22 Joules per year. The result is consistent with other data showing a lack of warming over the past few years.
Both the atmosphere and the ocean appear to be cooling. (Which makes sense as sea ice is also increasing.)

What is causing a significant five year cooling of the planet?

Comment:
The upper stratospheric temperature paper that was published in 2008 only used data up until 2003.



sylas
#17
May8-09, 07:13 PM
Sci Advisor
sylas's Avatar
P: 1,750
Careful guys; this is descending into crackpottery and irrelevance. There's a reason the Earth Science forum requires peer-reviewed references for support of controvertial claims – there's a lot of flagrant nonsense out there.

This thread is about the topic of Ocean Heat content. This is an excellent topic, and there are some genuine scientific puzzles and open questions here. Unfortunately, the thread risks being derailed into consideration of low grade popular caricatures of genuine science investigation.

Specifically.
  • Loehle 2009 is not a legitimate peer-reviewed science paper. The available data is much better examined using the legitimate scientific papers already cited rather than distracting into fringe oddities.
  • Bizarre claims are being made with no reference. The claim that "sea ice is increasing", for example, is symptomatic of the extreme flight from reality that shows up so often in these debates.
  • Irrelevant distractions are showing up. For example, we have large and badly produced graphs of tropospheric temperature. That's not the ocean. (Not to mention that the data actually shows more warming than in the ocean. More below.)
  • There is no "significant five year cooling". Five years is a short variation, almost by definition a five year "trend" is not significant!

More detail on each point.

References from Energy and Environment

Energy and Environment is not a science journal. It was set up by an English climate skeptic with no profession background in the topic. It doesn't show up in the recognized ranking systems and impact factors for scientific publications, anymore than OMNI or Readers Digest. The founder and main editor is Sonja Boehmer-Christiansen, an academic with background in geography, politics and policy – not science at all. The magazine has become a clearing house for easy publication of low grade material that can't get into a real science journal. Indeed, that was the aim. Boehmer-Christiansen apparently thinks there a bias against publishing work running counter to the mainstream science; but she's got it backwards. The mainstream is what it is because it has a bias against shoddy methodology and pseudoscience.

This "journal" has a lot of currency OUTSIDE science, where it is eagerly lapped up by people who don't accept the basic findings of conventional climatologists. It's cited in blogs, pundits, congressional submissions, etc, etc… but it has very little circulation in university libraries.

The paper by Craig Loehle (2009) would be okay as an undergraduate project, but the low level engaged is very obvious if you look at it side by side with a real science paper, such as Levitus (2009) or Domingues (2008) that has been used in the thread. Loehle's actual conclusions are rather weaker than his abstract suggests. The conclusion of the article is as follows:
While the current study takes advantage of a globally consistent data source, a 4.5-year period of ocean cooling is not unexpected in terms of natural fluctuations. The problem of instrumental drift and bias is quite complicated, however, (Domingues et al. 2008; Gouretski and Koltermann 2007; Wijffels et al. 2008; Willis et al. 2004, 2008a) and it remains possible that the result of the present analysis is an artifact.
Exactly so. The result is "not unexpected" in terms of natural fluctuations, and even worse – it remains possible that the result is an "artifact".

The analysis in the paper is trite. It's nothing but an exercise in curve fitting, using a distinctly odd method for finding a trend. He fits a sine curve to the data as a way of handling the annual cycles, and then fits for an underlying linear slope, and also a linear change in amplitude. That pretty much ensures he's going to get results with limited meaning. It makes sense to have a linear trend over a long period of time, but a linear damping?!? Seriously, forget the Loehle paper. It's not going to be a part of the actual ongoing scientific work, and with good reason. Stick with the science journals. It's all the same data being used, and the science journals do it much better.

Tropospheric temperature.

We've got two huge graphs above, badly produced, talking about atmospheric trends.

This brings up a whole new topic. Fundamentally, the atmosphere shows MUCH more variation than the surface or the ocean, which puts bigger uncertainty on the trends. But in fact, the graphs Saul have provided do show an overall warming trend, along with a lot of up and down that is larger than at the surface.

The regression trend for his RSS data is 0.155 C/decade warming, with 95% confidence bounds of 0.135 to 0.175. The trend in the UAH data is 0.128 C/decade, with 95% confidence bounds of 0.107 to 0.149. There's a pretty good working relationship between the RSS and UAH research groups. The scientific disagreements are the name of the game. Each group continues to work with the other in helping to find problems. The RSS website has a really nice little tool that allows you to compare all the various datasets involved. See Comparing RSS and UAH Atmospheric Temperatures to Adjusted Radiosonde Data. But take care to read the associated text; you get different trends when data is weighted by sampling to fit radiosonde data, probably because of the increased warming over the land.

For what it is worth, the sea surface temperature trend over this same period is about 0.133 C/decade, with 95% confidence limits of about 0.121 to 0.144, using the standard HadSST2 dataset: here. The deeper ocean warms more slowly than this.

This thread is about Ocean heat content.

The measurement of tropospheric temperature is an ongoing active research question. As matters stand, this is mostly about measurements. There are some theoretical issues here about refining dynamic models of the atmosphere generally, but as matters stand the models are consistent with the wide range of plausible measurement. Furthermore, this all stands distinct from questions of what is causing the rising temperatures we observe. The primary modeling issues here would be the same whether warming was driven by greenhouse, or insolation, or albedo. Theoretical predictions are within the range of the available measurements, and ongoing work to resolve discrepancies between different data sets makes for lots of interesting and contrasting work in the scientific literature. But it's almost all about trying to nail down difficult measurements.

The recent 5 year trend

Anyone who talks about a five year climate trend doesn't understand trends.

Over five years, the natural variation of climate and weather means you are necessarily looking at local variation. This is not a measure of "trend" at all, but of small scale local change. This is certainly interesting in its own right, but it isn't trend. A trend is something that extends over the time, and for climate and weather, the five year scale is dominated by natural oscillations. The long term trends show up only over longer periods.

Nor is it particularly surprising that a number of climate indices show lower values recently. The major contributing factor for this is the ENSO oscillations -- El Nino, La Nina -- which have just been through a cool point after a high point about five years ago or so. A good discussion of this is
Sea ice

The sea ice trends are strongly downwards in the Arctic, and about level in the Antarctic. Saul says:
Quote Quote by Saul View Post
Both the atmosphere and the ocean appear to be cooling. (Which makes sense as sea ice is also increasing.)
The above statements just don't make any sense. The actual trend for ocean and for atmosphere shows warming. Saul has mixed up a short term swing with a trend. That's just wrong, and fails to understand what "trend" means in the context of data with natural up and down variation. The oscillations are not trend, and the whole idea of trend is to abstract away from the short term variations. Trend is, by definition, not something you can find within spans of time dominated by oscillations.

But the sea ice thing takes wrong to a whole new level.

The sea ice talking point is symptomatic of the low level to which popular debate has sunk. What we actually have with sea ice is no significant trend in the Antarctic, and a very strong and unambiguous drop in the Arctic. You get the same thing with any actual scientific source of data, this is not ambiguous at all. There are good diagnostic timeseries for HadISST at the Hadley Centre.

Recently, however, Washington Post columnist George Will produced an incredible piece of nonsense mentioning increasing sea ice. The fallout from this was interesting and IMO symptomatic of the low grade of popular debate. The Washington Post has since published a couple of much more sensible op-eds on this, specifically noting that George Will's conclusions contradict the actual scientific data. Some readers might think that this represents two sides of a scientific debate. It doesn't. It represents error and denial on one side and science on the other.

Cheers -- sylas
Saul
#18
May8-09, 11:42 PM
P: 272
Quote Quote by sylas View Post
Careful guys; this is descending into crackpottery and irrelevance. There's a reason the Earth Science forum requires peer-reviewed references for support of controvertial claims – there's a lot of flagrant nonsense out there.

Cheers -- sylas
Sylas,

Name calling of the author, the publisher, and of others in the forum does not prove your point.

Data indicates the planet (both hemispheres) was been cooling for the last 5 years. Ignoring the data does not change the data.

I provided a paper that shows the planet's oceans (Total heat content.) have been cooling for the last 5 years.

http://www.ingentaconnect.com/conten...nlpdhr7c.alice

Ocean heat content data from 2003 to 2008 (4.5 years) were evaluated for trend. A trend plus periodic (annual cycle) model fit with R2 = 0.85. The linear component of the model showed a trend of −0.35 (±0.2) × 10^22 Joules per year. The result is consistent with other data showing a lack of warming over the past few years.
http://www.npr.org/templates/story/s...oryId=88520025

It is quite reasonable if the continents cool the oceans will also cool. Or alternatively if the oceans are cooling there will be an increase in sea ice and colder temperatures on land.


Register to reply

Related Discussions
Heat and Internal Energy and Heat and Temperature Change: Specific Heat Capacity Advanced Physics Homework 4
2 spheres in an ocean Introductory Physics Homework 3
Echo in the ocean Introductory Physics Homework 2
Ocean Heat Capacity Introductory Physics Homework 5
Europan Ocean q's Astronomy & Astrophysics 4