2010-09-08 09:19:03Refined GRACE results may produce ripple effect


Isostatic effects better estimated: 

"The innovative aspect of our method is that we simultaneously matched the current changes in the ice mass and glacial isostatic adjustment to the observations, instead of assuming that a particular glacial isostatic adjustment model is correct," says Dr Vermeersen. "For Greenland in particular, we have found a glacial isostatic adjustment model that deviates rather sharply from general assumptions. But at present there are too few data available to verify this independently. A more extensive network of GPS readings in combination with geological indicators for the local and regional changes in sea level changes around Greenland over the last 10,000 years, will possibly be able to provide conclusive evidence on this matter in the years to come." 


GRACE data is heavily cited here so this finding may require some revisions.

Of course, this does not affect observed sea level rise, creating a new "budget" hole. Premature conclusion by JPL et al regarding isostasy effect on ice sheet loss estimate? Thermal expansion bigger than previously thought? 

We'll see.  


2010-09-08 09:52:55If you want a copy of the papers
John Cook


I've got a PDF copy of the Wu paper (Simultaneous estimation of global present-day water transport and glacial isostatic adjustment) so if anyone wants it, just let me know and I'll email it over.

I haven't looked at it in detail but initially, I have two major queries about this new result. Firstly, the larger GRACE estimate is consistent with other estimates of Greenland mass balance:


Secondly, there are already difficulties in closing the sea level budget. If the Greenland and Antarctic ice sheet mass loss is less than currently thought, this makes closing the sea level budget even more difficult. So having trouble reconciling this latest result with the full body of evidence.

Welcome any other thoughts or insights (if I really wanted to be thorough, I'd contact some of the authors of the earlier GRACE papers for their reaction).

2010-09-08 10:06:06


If these ice loss estimate refinements are true, the seal level issue would seem to come down the usual: the ocean is heating, we just don't have the instrumentation to show where.

Which of course leads me to my old rant about how it is we can afford hair gel and mascara but can't take our temperature... grumble-grumble. 

2010-09-09 07:50:27Comment
Robert Way

I would consider watching out before any proclamations about this paper are made.

I wrote this at WUWT about the grace issue .... No one responded though

"Thomas Fuller,

The statement that all other methods of ice measurement besides grace show antarctica is gaining ice is absolutely and utterly a fabrication. Click my name to see a post on the matter (at skeptical science). It is very clear that Laser altimetry, radar interferometry and grace data agree well and have done so in the past. Furthermore, the only two studies to show antarctica was gaining ice were done using radar altimetry which has known biases over ice sheets (Thomas et al. 2008) that are known to have contributed at least 75 GT of false-gain over Greenland. It is quite nice to see such a clearly incorrect comment as it is easy to refute. Once again, see the graph at the bottom of the aforementioned link and prepare your correction to the original post.

Also with respect to the new GRACE estimates, notice that the new grace estimate still shows the EAIS is losing ice… and also that apparently the remainder of the worlds ice outside of a few areas is gaining according to it. Highly suspect to use such a low resolution sensor to detect changes in small regions such as svalbard, alaska, himilayas, Scandinavia, Patagonia and so on… Also note that the radar interferometry estimates for Pine Island Bay and the WAIS from Rignot et al. (2008a and b) still remain the most accurate assessments for WAIS ice losses."

This is somewhat of a summation of my thoughts on this paper. I think that the WAIS losses are being incorrectly depicted in this study. I think it is absolutely certain that there are large losses occurring in the WAIS... That being said, sure I think GRACE data should be refined and that this study perhaps can do so for Greenland where there is a good amount of GPS data but I think that for Antarctica it is an extremely large stretch to make assumptions... Also I think that it is of most interest that this study still shows EAST Antarctica losing ice... interesting...

2010-09-10 01:47:59Email
Robert Way

I responded to John also on this subject a while back in a more detailed post. See below:

It’s hard to really judge the study in just a first pass but considering it comes from the Jet propulsion laboratory where Eric Rignot works, I would say that this will be the new standard for GRACE data. That being said I think that previous studies which showed ice losses accelerating (Chen, Velicogna) are not affected by GIA correction so it’s hard to completely invalidate those studies. All that being said some important conclusions come from this paper, including that even with their conservative estimates (my wording not theirs) they find that the East Antarctic Ice Sheet is losing ice at around -23 +or- 29 Gt per year. Another interesting finding is that the West Antarctic Ice Sheet is losing only -64 + or – 32 per year. I think that estimate is wrong actually. For a lot of the radar interferometry studies they can produce extremely accurate measurements for the major flux areas (pine island bay and so on) but there’s large uncertainties. However because these regions are so important they’ve been flying depth profiles and have stations there and everything to reduce the uncertainty. To cut Rignot et al. 2008b’s estimate for that region of Antarctica in half is a tad overly optimistic I think. Another thing I can see as being an issue is that they estimate that the remaining land is actually gaining ice with their integrated method (14 gt year whereas with grace they find its losing 74 GT year). That’s a difference of about 90 GT between their integrated and just grace alone. There has to be something wrong there. Furthermore, I cannot believe that outside of those areas the sum of ice losses is positive. It can’t be happening. One of my friends actually (he used to TA a lot of my classes when I went to school at his uni) led a study on Svalbard and found it was losing ice at about 10 Gt a year (http://folk.uio.no/kaeaeb/publications/nuth_jgr.pdf
) and other estimates of the Canadian high arctic have all shown losses. The himilayan estimates show losses and so does the estimates for Patagonia (losing big time). Iceland is losing significantly and so are the swiss alps so I just can’t find it in me to trust their word that ice is gaining outside of the regions they study (Alaska, Greenland, Antarctica).  

I’m going to be absolutely honest and frank with all this. I don’t trust GRACE much in terms of actual estimates like they do here or chen or velicogna (or even cazenave) do. The data is so noisy that you have to massage it and introduce uncertainties just to get an answer.  It really just shouldn’t even be used for global estimates. Its resolution is far too low to even try at many dynamic regions such as in the Canadian high arctic or Patagonia. I think that it is ultimately a tool at our disposal for tracking accelerations in ice losses from the two ice sheets, but to do a full-blown estimate of their mass balance I think requires Altimetry (Cryo-sat 2) or Radar interferometry (such as from the Rignot et al. Studies).

Either way that’s my take on it. There’s a lack of agreement with other studies and I think there are questions to be raised but I don’t suspect they are necessarily incorrect, I just think that the scope of their study was a little further than the data could take them and that this is going to be a firestorm in the skeptic community because of it. We should really trust regional estimates done by glaciologists rather than always trying to go global. Go big or go home isn’t the model in glaciology.

Once again though i'd like to reiterate that it likely is a significant contribution to the literature on Grace and ice losses in the great ice sheets but that I suspect we will see different results from the next GRACE study which shows an update in losses...
2010-09-10 01:58:35


I had the same thought Robert did regarding the overall message from the refined GRACE results: we're still losing ice at a prodigious rate, according to GRACE. As Robert points out, lots of other indicators tell us the same thing.  

All the same, we know from past experience that "skeptics" will leap on this. It's arguably a good idea to revise articles referring to specific loss figures derived from GRACE, if this paper proves durable. Again, as Robert says, let's not be hasty, fair enough. 

2010-09-13 06:38:38Comment
Robert Way


A couple more notes from this study:

New paper coming out by renowned glaciologist Jonathan Bamber.

The sea level fingerprint of 21st century ice mass fluxes

Has somewhat different conclusions than Wu et al (2010) and does a sort of synthesis on the best available data for global sea level contributions. I think I`d be interested in doing a post on this paper if John doesn`t mind. I have asked a question about the paper with respect to Wu et al and we will see how the authors respond. Either way they use Rignot`s numbers for Antarctica and adjust them based upon studies since then (downward). I think it is a great contribution to the literature by a real expert in the field.

Another note that I just picked up on is that

For the study (Wu et al 2010) they estimated an ice loss of around -100 Gt for Alaska and Yukon

I think these numbers are incorrect based upon the new papers numbers (-60 Gt) and another papers estimates (Berthier et al. 2010) (-42 GT year up to 2006). The Berthier et al paper is very good as it actually found a problem in estimates and fixed it... he also did that here

My professor actually brought in the image from the study he fixed and had us as undergrads have to diagnose the problem. I think it might be necessary for a comprehensive article which shows glacier changes in each region. (Bamber 2010) has done most of that.

2010-09-14 21:36:18
Rob Painting


Given this is just the Greenland ice sheet (Wu 2010 red line), sure looks like more issues for the sea level budget alright.



2010-09-16 03:07:10Comment
Robert Way

I received a response from Bamber and Riva 2010 on this topic.

My Question


Their Response



We consider this paper to be an
interesting proof of concept for the approach but certainly not the "last word" on the
topic. It is important to realise that they are producing a global solution for GIA and
PDMT by finding a least squares minimisation for the GRACE, GPS and modelled OBP
data sets. The solution has not been tuned for any one location and the quality of the
solution will depend on the quality and spatial density of the data sets that went into it.
As far as we can tell, the solution is not constrained to pass through the observations.
In Greenland, there are very few GPS site with a sufficiently long record (just 3) that
could be used and none in the interior of either Antarctica or Greenland. In the future
the GPS networks set up within IPY for both ice sheets will greatly help with solving for
GIA in these areas.
2010-09-16 03:49:10I'm not sure why we are getting excited


about this.

I don't think we are planning to write up something on this right now, are we? It seems that the status is "up in the air", and the last thing we want is to have SkS tied in with a whiplash situation if it turns out that the analysis is badly flawed. We are not running a research center here, we're just trying to bring people up-to-speed on the accepted science.

2010-09-16 04:34:42Comment
Robert Way

I think that there will be a post put together at some point addressing the grace estimates and this. If Jonathan Bamber's analysis is badly flawed then I think that it would be shocking considering how well respected his work is among my peers. I think that the inclusion of the most recent estimates for Greenland ice losses is never a bad thing and when I get some time i'll piece something together.
2010-09-16 09:07:43Whiplash factor
John Cook


Neal raises a good point - often some skeptic meme gets raised and I run around all excited looking to rebut it asap - then it fades away quickly. On the other hand, some issues raise up and don't get responded to quickly enough and suddenly we find the skeptics have framed the issue in the public's mind. Think Climategate.

So it's a judgement call from issue to issue - you ignore an issue and it gets out of your control or you address it and give it more oxygen. It's a fine line to tread. If we think an issue has legs, then it's definitely worth expending some effort to put up a rebuttal that will serve as a resource for all the 'foot soldiers' out there arguing in blogs and forums. In this case, the skeptic meme is "hah, GRACE overestimated ice loss, nothing to worry about there". Will that argument have legs? I doubt it as we have multiple estimates of ice loss all giving a broadly consistent answer.  But it might be worth publishing a general rebuttal that gives an overview of the various estimates just to communicate the robustness of our understanding of ice loss - an all-purpose rebuttal for the varied attacks on ice loss data. Eg - an argument like "ice sheet mass loss is overestimated" (hmm, clunky wording but you get the idea).

2010-09-16 11:21:30Comment
Robert Way

My thinking is Ice Sheet losses are overestimated would be a good post to do and to add to the rebuttal list. This way we can have a end all resource to site back to with the Greenland is gaining ice, Antarctica is gaining ice, and It is too cold to lose ice arguments. Frankly they are all close to the same argument (with the exception of the antarctic one which uses sea ice) so perhaps if I do an advanced rebuttal to ice sheet losses are overestimated then I can have something concrete to work down from.
2010-09-16 11:22:13Comment
Robert Way

I will have to figure out how to make a graph that can show all the estimates well without being too confusing...
2010-09-16 12:03:35#130 Ice Sheet losses are overestimated
John Cook


Have added a new argument to the rebuttal list, #130 Ice Sheet losses are overestimated. I would imagine for the advanced rebuttal, you could grab a lot of content from your existing blog posts, turn it into a one-stop shop on ice loss.

Will be a good link for your WUWT comments :-)

2010-09-28 19:43:02

OK, clearly it's going to be necessary to address this GRACE issue ASAP.  The new Greenland blog post was just posted and already one commenter after another is saying that the GRACE data are overestimated.  Clearly, that argument has gotten a lot of traction!
2010-09-28 23:42:08Addressing ice loss argument
John Cook

I expected someone to bring it up but I was surprised at the amount of comments on Robert's Greenland post. Yes, obviously the Wu paper has traction. A quick response is required. Only problem is like ocean heat measurements, there's still a lot of uncertainty over GRACE data. In a sense, the Greenland post does address the central issue - we have multiple lines of data pointing to accelerating ice loss.
2010-09-29 00:21:13


The GRACE refinement should serve as an object lesson of how our understanding improves. 

What about a blog post using this as an example of forward thrust and promising revisitation? Wu is a preliminary approach itself. It's true that a response seems required but equally it is an opportunity to illustrate both the desperation of skeptics to leap with both feet on shreds of hope as well as a case study of how we improve ourselves.  

Not a reason to go on the defensive, instead use this offensively.

2010-09-29 02:09:41

Why not an "aseptic" update? We should compare the numbers before and after the adjustment (see Pelto comment), confirm the recent acceleration and point to the strength of multiple lines of evidence. And concluding with "hey, that's how science works!" :)
2010-09-29 07:38:48comment
Robert Way

Having read many of the arguments there. I should have jumped in right from the start with my commentary. I sort of dropped the ball there. My feelings on this topic are mixed. I think sure I could put together a post tomorrow tonight or tomorrow which tries to address the argument that ice sheet losses are overestimated but this is kinda a whole bag of worms that i'm not sure should be opened. Perhaps I should just posted a general blog post about Wu et al. 2010.

Just kinda starting with yeah it seems there's a lot of confusion over the new estimates, here's why and so on. I don't really know whether to address it or not. I think the bamber and riva commentary is very appropriate for this topic. Either way its a hard sell. A lot of people seemed to jump on me for not including Wu et al 2010 but yet no one even knew about Bamber and Riva (2010). Little do the people know I don't even know how to make a graph like the one in the greenland estimates... Ned? Do you by any chance know how to make the boxplot graphs or the one in my greenland posts. I have not for the life of me figured it out on excel yet.
2010-09-29 08:17:32

Geez, Robert, don't kick yourself. We've all got a life (hopefully) outside of this. 
2010-09-29 08:41:42comment
Robert Way

haha I have much less of a life than I used to... sometimes it can be therapeutic to work on things like this but other times it can be downright frustrating. I think I'm just too busy right now to be contributing very regularly and that is too bad. Perhaps I should consider taking some less controversial topics :P Either way it is too bad we will never be able to have a skeptical science mixer!

But thank you for the encouragement