I posted last Friday
Is There “Missing” Heat In The Climate System? My Comments On This NCAR Press Release
along with several e-mail exchanges with Kevin Trenberth. I have repeated this set of exchanges here, as well as new ones from an invitation to Josh Willis for his comments. The e-mails (presented with their permissions) are ordered from first to most recent [with spelling typos corrected]. I want to thank both Kevin and Josh for a constructive discussion which I look forward to continuing. This discussion can only further inform us of the issues associated with accurately monitoring the climate system.
On On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Kevin
I posted a comment on the NCAR press release this morning;
https://pielkeclimatesci.wordpress.com/
Would you be willing to write a guest post in response?
Roger
On Fri, 16 Apr 2010, Kevin Trenberth wrote:
Dear Roger
I do not agree with your comments. We are well aware that there are well over a dozen estimates of ocean heat content and they are all different yet based on the same data. There are clearly problems in the analysis phase and I don’t believe any are correct. There is a nice analysis of ocean heat content down to 2000 m by von Schuckmann, K., F. Gaillard, and P.-Y. Le Traon 2009: Global hydrographic variability patterns during 2003–2008, /J. Geophys. Res.,/*114*, C09007, doi:10.1029/2008JC005237. but even those estimates are likely conservative. The deep ocean is not well monitored and nor is the Arctic below sea ice. That said, there is a paper in press (embargoed) that performs an error analysis of ocean heat content.
Our article highlights the discrepancies that should be resolved with better data and analysis, and improved observations must play a key role.
Kevin
On 4/16/2010 Roger A Pielke Sr wrote:
Hi Kevin
Thank you for your response. I am aware of the debate on the quality of the ocean data, and have blogged on the von Schuckman et al paper. Since 2005, however, the data from 700m to the surface seems robust spatially (except under the arctic sea ice as you note). An example of the coming to
agreement among the studies is Figure 2 inLeuliette, E. W., and L. Miller (2009), Closing the sea level rise budget
with altimetry, Argo, and GRACE, Geophys. Res. Lett., 36, L04608,
doi:10.1029/2008GL036010.We both agree on the need for further data and better analyses. I have posted on this issue; e.g. see
However, I do not see how such large amounts of heat could have transited to depths below 700m since 2005 without being detected.
I am very supportive, however, of your recognition that it is heat in Joules that we should be monitoring as a primary metric to monitor global warming. Our research has shown significant biases in the use of the global average surface temperature for this purpose; e.g.
Pielke Sr., R.A., C. Davey, D. Niyogi, S. Fall, J. Steinweg-Woods, K. Hubbard, X. Lin, M. Cai, Y.-K. Lim, H. Li, J. Nielsen-Gammon, K. Gallo, R. Hale, R. Mahmood, S. Foster, R.T. McNider, and P. Blanken, 2007: Unresolved issues with the assessment of multi-decadal global land surface temperature trends. J. Geophys. Res., 112, D24S08,
doi:10.1029/2006JD008229. https://pielkeclimatesci.files.wordpress.com/2009/10/r-321.pdfKlotzbach, P.J., R.A. Pielke Sr., R.A. Pielke Jr., J.R. Christy, and R.T. McNider, 2009: An alternative explanation for differential temperature trends at the surface and in the lower troposphere. J. Geophys. Res., 114,
D21102, doi:10.1029/2009JD011841.https://pielkeclimatesci.files.wordpress.com/2009/11/r-345.pdfWould you permit me to post your reply below along with my response on my weblog.
Best Regards
Roger
On Fri, 16 Apr 2010, Kevin Trenberth wrote:
Roger you may post my comments. The V.s paper shows quite a lot of heat below 700 m.
Kevin
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Kevin
Thanks! On the V.s et al paper, lets assume their values since 2005 deeper than 700m are correct [which I question since I agree with you on the data quality and coverage at the deeper depths]. However, if they are correct, how much of this heat explains the “missing” heat?
It would be useful (actually quite so) if you would provide what is the missing heat in Joules.
Roger
START OF NEW E-MAILS
On Fri, 16 Apr 2010, Kevin Trenberth wrote:
The difference is a few tenths W m-2.
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Kevin
This is a flux. What is the missing accumulated heat in Joules?
Roger
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Josh
Can you provide me with feedback I can post (or off the record if you prefer) on the NCAR press release and my comments that I posted today my weblog?
Roger
On Fri, 16 Apr 2010 Josh Willis wrote:
Roger,
I think that it is still premature to make claims about the Earth’s energy imbalance based on satellite observations and ocean heat content data over ANY period. As with the satellite observations, the ocean heat content data continue to undergo refinement and removal of systematic errors. Since the satellite data are insensitive to the absolute value of the imbalance, they rely on ocean heat content data to estimate it. However, I personally belive that there is not a long enough common period between the satellite observations and the RELIABLE ocean heat content record to make any strong claims about the energy budget. Many people are working on both data sets, however, and I hope that a more reliable comparison will become available soon.
Josh
On Fri, 16 Apr 2010, Kevin Trenberth wrote:
I agree with Josh and that is part of the message of our article, plus that it is important to be able to do this better.
Kevin
On Fri, 16 Apr 2010, Josh Willis wrote:
Cheers to that! :)
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Josh
Thanks for the quick feedback. I assume, as usual, I can post your comments. Kevin has already given me permission to post his.
On whether the ocean data is good enough to diagnose the radiative imbalance, do you now disagree with the error bars in the data you provided me that I presented in the paper
Pielke Sr., R.A., 2008: A broader view of the role of humans in the climate system. Physics Today, 61, Vol. 11, 54-55. https://pielkeclimatesci.files.wordpress.com/2009/10/r-334.pdf?
as well as the plot of the data in Figure 2 of
Leuliette, E. W., and L. Miller (2009), Closing the sea level rise budget with altimetry, Argo, and GRACE, Geophys. Res. Lett., 36, L04608, doi:10.1029/2008GL036010.
This is what you wrote to me and I posted with your permission in the post https://pielkeclimatesci.wordpress.com/2009/12/23/further-comments-on-the-inadequate-epa-response-to-reviewer-comments-on-ocean-heat-content/
“The Agro Science Team continues to improve the accuracy of the float pressure data for the entire historical Argo dataset. As it strives to achieve the array-averaged accuracy of 1-2 db that is necessary for estimates of global sea level and ocean heat content, small but significant revisions in estimates based on Argo should be expected, particularly in the early years of the array prior to 2005.”
Your comment in response to my e-mail conflicts with that data analysis. Have you now also found significant errors in the analysis from 2005 to the present?
Best Regards
Roger
On Fri, 16 Apr 2010, Josh Willis wrote:
Hi Kevin and Roger,
Incidentally, you two might be interested in this recent paper by Sarah Purkey and Greg Johnson:
http://www.pmel.noaa.gov/people/gjohnson/Recent_AABW_Warming_v1.pdf
They looked at the prospect of deep warming on decadal time scales using the sparse, but highly accurate repeat hydrographic sections and found that below 3000 m in the global oceans, and below 1000 m in the southern ocean, the ocean is taking up an energy equivalent of about a 0.1 W/m^2 energy imbalance at the top of the atmosphere. So while this is significant, it suggests to me at least that the deep ocean is probably not taking up a bunch of heat really rapidly and the traditional idea that most of the action is in the upper several hundred meters is probably going to hold up. (did I get that right, Greg?)
Cheers, Josh
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Josh
If the ocean heat data is robust since 2005, this means the global average radiative imbalance can be computed from that metric (i.e. the change of heat in Joules over time). I discussed this
Pielke Sr., R.A., 2003: Heat storage within the Earth system. Bull. Amer. Meteor. Soc., 84, 331-335.
https://pielkeclimatesci.files.wordpress.com/2009/10/r-247.pdfDo you disagree with this now?
The conclusion of my article was that we do not need the satellite measurements of the radiative imbalance, if we allow the climate system (predominately the oceans) to provide the actual heat change in Joules.
I agree with Kevin that we need better data, and we must seek to obtain more precise measurements both of satellite derived radiances and of ocean measurements.
However, from 2005 onwards, as you confirm today in your e-mails, the upper ocean is well sampled since then and there is relatively little accumulation of heat over this time period deeper in the ocean.
Where do you conclude that the “missing heat” could be, if it is missing?
Best RegardsRoger
On Fri, 16 Apr 2010, Josh Willis wrote:
Hi Roger,
No, I think the analysis from 2005 to the present is roughly right. What we have not done yet, however, is compare these recent estimates of ocean heat content with the satellite observations of the radiation budget. Until such time as that analysis is done, I think we should use some caution about saying what the radiative imbalance is and has been.
I am happy with you posting any of these remarks.
All the best,
Josh
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Josh
In diagnosing the year to year accumulation of heat, this does not tell us anything on attribution. However, for the years where there is little accumulation, there must be a near zero radiative imbalance. Prior to 2003, the models (at least the GISS model) were in good agreement, but it has not been more recently [at least through mid-2009 which is the last time period I have seen; please let us know if there is more recent information].
I do agree with you and Kevin on the need to use the satellite data, as you outlined. However, it is not needed to diagnose the radiative imbalance from the ocean heat content as long as the data is sufficiently spatially and temporally accurate.
Since you agree that the data is robust since 2005, it should be a high priority to update this data in as close to real time as possible. Also, since the global average surface temperature is used as the “gold standard” to communicate to policymakers, the upper ocean heat content changes should be presented also together with the temperature trend and anomaly.
One very positive aspect to Kevin’s article may be a call to do this. While I disagree with him on the implications of this “missing” heat, I completely support his recommendation for better data and for the use of heat as a priority climate metric.
Roger
On Fri, 16 Apr 2010, Josh Willis wrote:
Hi Roger,
I think that’s right, that the ocean heat content is relatively small over the last few years. However, as I’ve mentioned in the past, it is not uncommon for coupled models to have several year periods with little or no ocean heat accumulation as part of their natural internal variability. In fact, I saw an excellent talk on this at the most recent Ocean Sciences meeting in Portland by Caroline Katsman.
I also think that the satellite data is still very important because it provides further information such as changes in albedo, OLR, and incoming radiation that cannot be retrieved from ocean heat content data alone. So, if the ocean has not accumulated a large amount of excess heat over the last several years, the satellite observations should help explain why. For this reason, I think it is very important to reconcile the ocean heat content data with the satellite observations.
On Fri, 16 Apr 2010 Roger A Pielke Sr wrote:
Hi Kevin
The radiative imbalance in watts per meter squared can be directly converted to Joules per time period. I have just been sent your very clear figure in your Science paper and it does present quantitative values for the radiative imbalance (the “net radiation:). Jim Hansen has done this for the GISS model results.
I will estimate it from your figure for a post on your paper I am doing for Monday, but prefer your value since it is your paper.
On Fri, 16 Apr 2010, Kevin Trenberth wrote:
Roger I don’t believe any of the current dozen or so estimates of ocean heat content are correct. The TOA estimates are probably closer to being correct but they too have problems. The data may be robust since 2005 but the analysis methods are not. Kevin
On Fri, 16 Apr 2010, Kevin Trenberth wrote:
Roger you seem not to be aware of all of the other publications we have on this topic. Please see my website. Wrt the Science article, some further detail is in
Trenberth, K. E., 2009: An imperative for climate change planning: tracking Earth’s global energy. Current Opinion in Environmental Sustainability, 1, 19-27, doi:10.1016/j.cosust.2009.06.001. [PDF] Click here for the Science Direct link including a video presentation of the abstract.
The leveling off of ocean heat content since 2003 is only in the upper 700 m. von Schuckmann et al have 0.54 W m-2 globally for 2003 to 2008 for depths to 2000m. There is good reason to believe this is an underestimate because of the way they analyzed the data.
On Fri, 16 Apr 2010, Josh Willis wrote:
Hi Kevin,
I’m not sure why you think that the analysis methods of recent ocean heat content estimates are not robust. Since about 2005, most any analysis method that makes use of the Argo data should get approximately the same answer, which is that there is little net warming over this period.
In fact, I have verified that my estimate compares well with Karina’s between 2005 and the end of 2008, even though she integrates to 2000 m and I only integrate 900 m. In the early part, however, the von Schuckmann analysis is problematic because it definitely contains some Argo data that still had pressure biases, and because they relied on a climatological background field that was probably too cold. Because the early part of the Argo record (pre-2005) has large gaps, their analysis relaxes toward the cooler climatology in the early part of the record. This has the potential to make the global trend appear larger than it may actually be.
Without cleaned up Argo and/or XBT data with reduced biases, I do not belive that we can actually say that the satellite data in recent years are more reliable than the ocean heat content estimates.
On Sat, 17 Apr 2010 Roger A Pielke Sr wrote:
Hi Kevin
I was not aware of your paper below and will read. We are in agreement that better data quality is needed. I still, however, can not understand how heating can occur below 700m without it being seen transiting through that upper level.
In terms of the robustness of the analysis methods, for the period since 2005, I interpret Josh’s comment as disagreeing with you on this.
Regardless, this exchange of viewpoints has been very constructive.
Roger
On Sat, 17 Apr 2010, Kevin Trenberth wrote:
Well of course any deep warming goes thru the surface layers but that requires detailed measurements far beyond Argo to track that flux. It can relate to MOC, convection, ENSO etc.
Kevin
On Sat, 17 Apr 2010 Roger A Pielke Sr wrote:
Hi Kevin
I agree measuring the fluxes is a challenge. However, this is not needed in order to see heat transfer to lower levels. We just need to monitor the Joules in the layers from the sfc to say 700m to see the propagation of heat to lower levels. Unless the transfer were faster than the time interval of the measurements, this should be resolved by the available data.
Moreover, with respect to heating deeper in the ocean, even when this occurs, it would be diffused out spatially over time. It is hard to see how it could reemerge in large amount of Joules into the upper ocean and atmosphere in a short time frame. I did mention this possibility in my paper
Pielke Sr., R.A., 2003: Heat storage within the Earth system. Bull. Amer. Meteor. Soc., 84, 331-335. https://pielkeclimatesci.files.wordpress.com/2009/10/r-247.pdf
where I wrote
“An assessment of the heat storage within the earth.s climate system offers a unique perspective on global change. If the heat actually remains within the earth system in the deeper ocean, for example, while the heat content of the remainder of the heat reservoirs in the earth system remains unchanged, sudden transfers of the heat between components of the system (from the ocean into the atmosphere) could produce rapid, unanticipated changes in global weather.”
However, I am now convinced that such a transfer cannot be sudden. I am open, of course, to examples of how this could occur.
Roger
On Sat, 17 Apr 2010, Kevin Trenberth wrote:
Josh
It is simply that there are many OHC estimates and although they all seem to have a slow down in warming in recent years they are all different. Indeed not accounting for a global warming signal in the analysis is a general flaw (assuming stationarity) and how one interpolates in space and time. Regardless of the details of vs et al. they do show a signal from below 700 m, right?
Kevin
On Sat 17 Apr 2010, Josh Willis wrote:
Actually, Kevin, I do not think that they are all very different. It sounds like you are familiar with Lyman’s upcoming Nature paper and in both Figure 1 and Figure 2 of that paper, you can see that the estimates after 2005 all have the same basic variability. In fact, if these estimates did not include the XBT data, they would be even more similar in terms of their variability. I have personally verified that my own estimtes of OHC variability are very similar to those made by von Schuckmann, Eric Leuliette and another Argo-only analysis by Dean Roemmich for the post 2005 period. During this period, the technique and statistics used to interpolate the data is really not that important because the data coverage is very good. In fact, the same is true for the period of the 1990s. One of the main points of the Lyman Nature paper is that the data biases are by far the most important remaining error–much larger than the differences caused by different interpolation techniques or differences in the assumed statistics.
Karina’s paper does show signals below 700 m in the zonal averages, which is not surprising, but she does not separate the 0-700m layer and the 700m – 2000 m layer in the global average. Dean Roemmich looked at this in his 2009 Progress in Oceaongraphy paper, and found that there were clear decadal differences between deep argo and earlier work, but I don’t think he saw a very big year to year signal in the > 700 m globally averaged layers.
You should also note that Karina’s paper suffered from errors in the altimeter data that were still not corrected at the time of her paper. These errors tended to make the altimeter time series show too much global sea level rise, and after correcting them the trend in globally averaged sea level since 2004 or 2005 is significanly lower.
Finally, I do not think that any of the techniques used by various groups should be supressing the global warming signal in the data over the period from 2005 to the present. As I mentioned above, the Argo data coverage during this period is such that any reasonable interpolation technique should do. Capturing the trend over 50 years, however, is another story.
Cheers,
Josh
Climate Science: Roger Pielke Sr. 