This is Al Rodger's Typepad Profile.
Join Typepad and start following Al Rodger's activity
Join Now!
Already a member? Sign In
Al Rodger
Recent Activity
I have gained a bit of learning about the NSIDC monthly SIE data. This resulted from some unwarranted criticism from me towards the NSIDC Arctic Sea Ice News bulletin covering the December SIE numbers. (Likely some, maybe many, are already familiar with the issue I now explain, so my apologies for this lengthy whitter if you are already in the know.) The daily trace of SIE (eg see ChaArctic with 'Show all') shows December 2016 continually tracking below all other years. So when the monthly figure for December was posted by NSIDC as a smidgen higher that the 2010 value (which remained 'lowest December on record') I assumed this had to be a mistake. It turns out, after some enquiries with NSIDC that the December SIE figure was actually correct. The NSIDC communique explained how a set of lower daily values could yields a larger monthly SIE with the following example (condensed down & editted a bit by me):- Imagine one 10 x 10 km grid cell (100 sq km). [Call this 'cell X'] Over a 30-day month, it is covered by 100% on 5 days and the other 25 days it is ice free. AVERAGING DAILY DATA SIE 'cell X' = (5*100 sq km + 25*0)/30 = 16.67 sq km AVERAGING EACH CELL OVER FULL MONTH 16.67 sq km > 15% x 100 sq km ∴ SIE 'cell X' = 100% = 100 sq km Thus for 'cell X' (and many others similarly), SIE can be 'boosted' by up to 85 sq km. Of course, another way of seeing this is that daily sampling by using thirty-times more averaged data points is more accurate than the monthly cell sampling version, so the result will be closer to the actual Sea Ice Area. We are happy accepting SIA measures being smaller than SIE so an SIE calculated using daily data which will be more accurate, which will be more SIA-like, should yield a smaller value. And using the daily rather than the a monthly data does almost always yield values smaller than the monthly cell SIE value. Examining the size of the 'boost' in SIE through using monthly cell data rather than averaging daily figures (and ignoring 1987 which has what looks like artifacts of the change in data sets) the size of the 'boost' does vary quite a lot through the year, smaller boosts during max & min ice, larger boosts during the melt/freeze seasons, particularly the freeze, and particularly October. There is also a lot of year-to-year noise which is of a similar order of magnitude as the boost. As a result of these 'boosts', the monthly cell SIE for December 2016 was boosted 620,000 sq km above the daily averages value while back in 2010, SIE was only boosted by 179,000 sq km, allowing 2010 to remain the Lowest December SIE using monthly cell data even though the daily trace for December 2016 was significantly lower than the 2010 daily trace for the entire month. Within all the noise, there are also trends in the size of the 'boost' evident in some months which in a few months yield statistically significant linear trends. But because of all the noise and the wobbles in the daily SIE trace, the 'boost' doesn't exctly jump out at you. Thus, even the month with 'boosts' far larger than all other (October) which also has the most convincing trend (the October boost in the 1980s averaged 420k sq km, in the 2010s it is averaging 750k sq km), still has a lot of noise attached, its linear trend measuring +110,000sqkm/decade (+/-70,000sqkm -2sd) For the curious, here are the full-record average boost for each month (thousands of sq km), with the average for the last 10 years in brackets to give some indication of trends. Jan 130 (120), Feb 80 (60), Mar 90 (80), Apr 80 (70), May 100 (140), Jun 140 (210), Jul 250 (260), Aug 40 (90), Sep 90 (80), Oct 600 (770), Nov 280 (320), Dec 220 (270). ANNUAL 175 (210) (This 'boost' may well partly explain why my trace of 12-month average JAXA SIE graphed at MARCLIMATEGRAPHS sits 600,000sqkm below the NSIDC 12-month average – the JAXA trace being calculated from daily JAXA data and the NSIDC calculated by NSIDC from cell data averaged over the month.) NOTE – The size etc of this 'boost' his can be compared with what we could call the SIA-to-SIE 'boost', the difference between SIA & SIE. The SIA-SIE 'boost' is far bigger, just as noisy, is more consistent through the year and has a stronger trend but in the opposite direction (ie reducing the difference -90,000sqkm/decade (+/-30,000sqkm -2sd)) with a visible acceleration in that reduction of annual average over the period. For comparison, here is the table above but calculated for SIE-SIA:- Jan 1,800 (1,600), Feb 1,900(1,600), Mar 1,900 (1,500), Apr 1,700 (1,500), May 1,800 (1,700), Jun 2,300 (2,300), Jul 2,800 (2,500), Aug 2,100 (1,800), Sep 1,600(1,200), Oct 2,000 (2,100), Nov 1,600 (1,400), Dec 1,800 (1,600). ANNUAL 1,900 (1,700) And with all that, NSIDC inform me:- “We have received similar questions in the recent past about our December numbers, and the science leads have decided to switch the way in which the averaging is completed. The current method is really just a legacy way of doing things as the dataset's original intended purpose was to simply produce coarse resolution figures (c.a. 2007) on a monthly interval for our site. The dataset is now clearly the most popular product we have due to our blog-style publication and thus changes will be made after considering any impact to the community.“ So the 'boost' to SIE monthly numbers due to averaging grids over the month rather than averaging daily SIE, that boost may soon be a thing of the past for the headline monthly SIE.
Toggle Commented Apr 7, 2017 on PIOMAS April 2017 at Arctic Sea Ice
Rob Dekker, Ding et al (2017) does address a worry that imposes itself on polar climatology. I think it presents a useful result but that result is presented in such a way that it can be used to suggest that natural variability is responsible for perhaps half of the loss of Arctic SIE recorded over the last four decades - a suggestion which is patently false. Importantly, it is not the absolute loss of SIE in 'millions sq km' which Ding et al address. It is the trend in SIE decline in 'millions sq km/decade'.Sadly the paper does not assist very well in preventing its findings being misrepresented. If you plot (as Rob Dekker suggests) temperature against Sept Arctic SIE, what do you see? For temperature, HadCRUT4 global is probably the best choice. It doesn't have a lot of Arctic weighting so we are surely comparing AGW against SIE without picking up any natural Arctic temperature wobbles. The resulting plot shows a strong linear relationship except for the period 2007-12. It is that wobble (which is now over) that impacts the calculation of trend in SIE decline. The trend in SIE if the wobble years are ignored (and using 1979-2014 as per Ding et al) is something like 0.6M sq km/decade. If the wobble-years are reinstated, the trend increases to 0.9M sq km/decade, yielding a 33% reduction due to the wobble which is a result not greatly different to Ding et al's 43%. But importantly, that wobble is now over and as of 2016 there is no significant natural wobble impacting the levels of summer SIE.
Toggle Commented Mar 22, 2017 on Lowest maximum on record (again) at Arctic Sea Ice
Golocyte Golo, I've not read the paper below myself yet, but it looks like it might make interesting reading. It was poorly referenced in a book and so didn't spring to hand when I first went looking for it. Qian et al (2016) 'Climatic anomalous patterns associated with the Arctic and Polar cell strength variations' (PDF)
Toggle Commented Mar 15, 2017 on PIOMAS March 2017 at Arctic Sea Ice
Golocyte Golo, I'm no expert on this but I do see a bit of a hole in the descriptions of atmospheric circulation. The theory of the three cells is described well enough in many places but I don't know of a good description of the actual reality resulting from the likes of land masses and summer&winter. I would suggest that it is helpful to consider the cell system of the atmosphere as driven by the tropical solar heating and that it is a lot weaker over the poles. Indeed, would it be better to call it 2½ cells rather than 3? The one piece of learning I found helpful in understanding the atmospheric circulation is that the jetstreams are a sort of 'release-valve' equalisng the East-West corriolis effect. That sort of takes the mystery out of jetstreams. Specific to your question, the magic word here is "Arctic Oscillation". When the AO is positive (perhaps confusingly) there is negative pressure over the pole. The index of AO from the NOAA shows it can persist in a positive state (more negative pressure) for some months. But even those positive phases do usually show a residual high pressure zone within the cell. Events as extreme as early January 1993 (illustrated here) are the rare exception. The CCI Climate Reanalyser shows Global NH Mean-Sea-Level-Pressure allows a useful view of how the pressure systems over the Arctic vary each month and annually. Over the year, pressures average out to show the positive pressure over the pole. The usual discussion of the impact of AGW on atmospheric circulation is the expanding Tropical cells. The wobbly Arctic cell usually comes under discussion of the AO. I see a Review article from 2014 (I don't remember reading it) which may useful place to start on the literature.
Toggle Commented Mar 13, 2017 on PIOMAS March 2017 at Arctic Sea Ice
Hans Gunnstaddar, The CO2 increase over the previous 12 months is still boosted by El Nino and will only get back to a more normal level of annual increase when the northern growing season kicks, so that is from May onward. I have been plotting out the annual rise 2015-17 against the rise back in 1997-99 - see here (usually 2 clicks to 'download your attachment') - and there is nothing so far to suggest that the annual CO2 rise will not settle down to 2.25ppm or so, as it was pre-El Nino.
Toggle Commented Feb 13, 2017 on PIOMAS February 2017 at Arctic Sea Ice
David Nemerson, I think your reference must have dropped a decimal point shift somewhere in their calculations. I find the bottom paragraph of the PIOMAS page is always very useful in remembering the energy required to melt sea ice. It's about 3,300 cu km per zettajoule (zetta=10^21). So 11,000 cu km would require something like 3.3 x 10^21 j. It doesn't affect the conclusion but it is a significant correction.
Toggle Commented Feb 3, 2017 on A new Arctic feedback (?) at Arctic Sea Ice
So the storm intensity is as forecast (959 hPa atm). JAXA is showing a lot less ice than ever recorded for the time of year, 0.75M sq km below the 2002-15 average, an anomaly much bigger than any seen before in winter. How far will this storm impact that figure?
Rob Dekker, The image is viewable in all its glory simply by a right-click & View Image. Where I object to these annual DMI graphs is their lack of information on the variability of pervious years. Past variability is shown on this graphic from NSIDC's Andrew Slater which more than compensates for the this-year trace not being entirely continuous. (Hopefully I can remember the code to link an image into the comment - I haven't posted an image with HTML for ages.) This image looks like it will also require right-clicking & View Image as the size command doesn't do naff-all here. The Andrew Slater website with graphs for previous years etc is here.
Sam, As a check on your calculations, you may find this CDIAC webpage useful. It is updated with 2015 GHG concentrations & the various GHGs listed total to 527ppm CO2(eq). I should repeat that CO2(eq) values are a poor substitute for more detailed modelling of the consequences of our GHG emissions.
Toggle Commented Nov 20, 2016 on PIOMAS November 2016 at Arctic Sea Ice
Sam, The mechanisms being described as "slow feedbacks" by Hansen et al (2013) are the result of changes in albedo & GHGs, both resulting from melting of the chryosphere & vagitation change. As the name says, they are slow and they are feedbacks and will precipitate their own fast feedbacks (when/if they were to occur). Hansen et al (2013) cites papers suggesting the present level of global temperatures give a 30%-50% increase on warming due to slow feedbacks. This increases to 100% if warming moves to an ice-free world as per 35 million years ago. I interpret the message being given as saying that we are prodding a beast that can be very dangerous and that we do not entirely understand. Our continued CO2 emissions are an extreme folly. Were this not Neven's Arcic blog, I would spend a little more time considering the content of this comment but I feel we are now addressing quite broad subject-matter that is not cryospherical in nature.
Toggle Commented Nov 20, 2016 on PIOMAS November 2016 at Arctic Sea Ice
Wayne Kernochan, Just sticking with the CO2 growth.... Of course, time will tell what levels of CO2 growth will drop to soon enough. However, a view that CO2 growth will remain a 3.5ppm.pa until March 2017 is a long way from my own assement. Do note that the 0.93ppm/yr figure given by NOAA MLO data for 1999 is the CO2 growth over the year (Ave.CO2 Nov99-Feb00 minus Ave.CO2 Nove98-Feb99) and so strongly impacted by the La Nina conditions that had appeared immediately afrer the 1997/98 El Nino. (This time round, the La Nina is not anything like as strong, but a short and weak La Nina that will be replaced by neutral conditions by early next year.) The weekly data provided by NOAA shows annual CO2 growth remained above 2ppm.pa until week 20 of 1999. I do recomment using the weekly data to those assessing CO2 growth.
Toggle Commented Nov 20, 2016 on PIOMAS November 2016 at Arctic Sea Ice
Wayne Kernochan, Concerning your second comment. I am surprised by what you say. I too follow MLO CO2. The one exemplar we have for a strong El Nino affecting CO2 levels is the 1997/98 event. The weekly data if smoothed by say averaging over a 5-week period show clearly that the increase in CO2 levels over the 1997/98 El Nino peaked in early October 1998 and dropped down from that peak level over the following nine months or so. The decline is not smooth but bumpy. This 2015/26 El Nino saw CO2 rises peak in June, earlier than in 1998. The peak was a higher level (roughly +0.3ppm). We perhaps should expect a slighly larger margin with emissions being in excess of 2Gt(C)pa higher than in 1998 & an atmospheric fraction of 0.4. The rise in annual CO2 levels is now reducing bumpily, down from June's 4.0ppm.pa to 3.3ppm.pa in October. Give it time. Let the bumps come and go. By mid-2017 we will see a more representative rate of increase. And concerning your third comment, the concept CO2(eq) is really designed to allow decision-making within climate change mitigation, allowing the effort to reduce one GHG to be gauged against the effort to reduce another. Do not use CO2(eq) to calculate resulting global temperature rises, certainly not in the manner you (& Sam) describe. The pre-industrial levels of non-CO2 GHGs would be solely used to calculate the forcing these non-CO2 GHGs add, forcings which allow a straighforward calculation of the resulting warming. CO2(eq) is the CO2 level that would cause such a forcing. The pre-industrial CO2(eq) is simply not part of the calculation.
Toggle Commented Nov 20, 2016 on PIOMAS November 2016 at Arctic Sea Ice
Jeff Lemieux, I think you would describe Margaret Davidson as a technical/scientific civil servant rather than a scientist and her un-referenced statement is not well made. I'm sure that if there were some "OMG" field data out of West Antarctica, even preliminary data, it would not remain secret in the manner described. So I would suggest that Davidson is referring to something in plain sight. Perhaps she refewrs to Hanson et al (2016). That paper is based on a certain level of evidence but is more a discussion document rather than a focused piece of science. It does point to "West Antarctica and Wilkes Basin in East Antarctica (having) potential to cause rapid sea level rise" which sort of fits the bill although the paper does not provide the "roughly 3 meters by 2050-2060".
Toggle Commented Apr 15, 2016 on Greenland under early pressure too at Arctic Sea Ice
A final cautionary tale on the PIOMAS postings. In the past I (& others) have encountered at the NASA sites what I shall call "sticky caching." That is, you link to a web page and it shows you what it looked like last time you linked to it - updates since then do not register. I now hit 'refresh' when I'm expecting NASA updates & they haven't appeared (although I think the problem is passed at NASA). The relevance to PIOMAS is that I have just encountered "sticky caching" on the official PIOMAS page and that monthly page - the March update that I knew should have been there only appeared when I hit 'refresh'.
Toggle Commented Apr 5, 2016 on Winter analysis addendum at Arctic Sea Ice
Apologies, one and all. Neven saying it looked like the value for March 1st jogged my poor fuddled brain, and memories returned of an extraneous value appearing last month on that monthly page at the same time as February was posted. I commented at the time. It is that same number. Sorry for the false alarm.
Toggle Commented Apr 4, 2016 on Winter analysis addendum at Arctic Sea Ice
I know no more of the provenance of the monthly volume page I link to above. I only happened upon it a few months back. The numbers have in the past matched the daily data pointed to by the official PIOMAS page. Whether the monthly posting starts life as the final total or starts life as a provisional value, I know not? That the posted value isn't so far from any expected value wouldn't help in answering that. GWPF? Gentlemen Who Prefer Fantasy!! The quote from NSIDC's Ted Scambos that the 'gentlemen' leave off their version of the blog is “The Arctic is in crisis. Year by year, it’s slipping into a new state, and it’s hard to see how that won’t have an effect on weather throughout the Northern Hemisphere” Homewood, their blogger, calls this "no longer a case of science" which is probably correct. Once the science is settled, the scientific findings can be freely adopted beyond science. That Homewood chooses to use anomaly time-series to cloak the "'slippage' into that new state" simply demonstrates he is an untrustworthy source of information. So he and the 'gentlemen' are probably well made for each other.
Toggle Commented Apr 4, 2016 on Winter analysis addendum at Arctic Sea Ice
The PIOMAS monthly volume page is showing March's volume at 20.621k - 770 cu km below the previous lowest (2011). With that, it did occur to me that with the unprecedented Arctic temperatures for the first part of 2016, the PIOMAS model will be pushing into new territory. Interesting times for model-based data.
Toggle Commented Apr 4, 2016 on Winter analysis addendum at Arctic Sea Ice
Bill Fothergill, 170 Zj to melt 5.09 x 10^17Kg ice? I find the bottom paragraph on the PIOMAS page is a good memory jogger if I ever forget the ~3,000 Gt(ice)/Zj. (They actually say 16,400 cu km requires "about" 5 Zj.) Working that through, it works out on the nail at 5.1 x 10^17 kg ice. You have to say, it's a mighty fine web page, that PIOMAS one is!!
Toggle Commented Mar 29, 2016 on 2015/2016 Winter analysis at Arctic Sea Ice
Colorado Bob, Northern snow cover is the Cinderella of the Arctic AGW. We are already experiencing spring & summer anomalies that were projected by the IPCC to happen decades into the future. This dramatic snow loss sits alongside a smaller increase in the (far far more variable) snow cover over the autumn & winter. Mind, we are this year seeing large negative winter anomalies although these are not unprecedented (even in recent years). With the large temperature anomalies presently over NH land, these negative winter anomalies may (and this could all disappear under a change in the weather in coming weeks) stick around into April (but no signs of them going so far) and that may give the spring snow cover a run at a stonkingly-big new record. My snow graphs are numbers on Vc and Vd on this site. (I'm conscious that I'm plotting decadal values with annual values on Vd which is a bit wrong given the winter wobbles. I'm thinking of adding my annual graph to rectify that.) (I should perhaps also add given comment up-thread that in my view NH snow cover and global temperature both are no more scary than they were last year, or the year before.)
Toggle Commented Mar 15, 2016 on PIOMAS March 2016 at Arctic Sea Ice
Looking at the PIOMAS year-on-year graph, the 20,621k sq km figure posted on the monthly data page is almost certainly the last-day-of-February figure (which scales to 20,613k) posted by mistake. Jim, thanks for that. Pressing SEND worked. I do wonder if my problem was that when I first encountered that page it wasn't up & running properly. This was a long time ago now, but having tried repeatedly submitting my e-mail, name, shoe size, etc & pressing SEND I never once got a link come up. So I haven't bothered pressing SEND since, not until just now.
Toggle Commented Mar 9, 2016 on PIOMAS February 2016 at Arctic Sea Ice
I've not been monitoring it so I don't know how long PIOMAS has been showing February's value - possibly not long. What I find mystifying is the monthly data has been updated to March. And more, it is giving a March figure 800k below the March record set in 2011. (As I lost the link to the daily data which doesn't appear for me on the PIOMAS site, I can't see what has happened there.)
Toggle Commented Mar 9, 2016 on PIOMAS February 2016 at Arctic Sea Ice
With 2016 a pip away from the 2006 global SIA record, we are later in the year than the minimums of 2006 & 2011. Perhaps more telling, the Arctic SIA sits at the lowest level for time of year and there will need to be +800,000sq km added to SIA to prevent 2016 snatching 2011's lowest Arctic winter SIA record.
Tor Bejnar, You whetted my appetite so I went looking. I soon was finding talk (eg "How the Isthmus of Panama Put Ice in the Arctic" by Haug & Keigwin) that the Panama gap had closed as a significant feature on the climate by 4.2 million years ago but it required something else to kick the Arctic into proper icy conditions a million years later. The orbital obliquity which is one of the theories mentioned as a trigger(or the lack of orbital obliquity stopping the trigger 4.6-3.1My) is plotted in this earlier Haug & Tiedemann & paper.
Toggle Commented Apr 13, 2015 on The Ns are calling the maximum at Arctic Sea Ice
Tor Bejnar, Timing-wise, the appearance of the Panama Isthmus in terms of shutting off ocean currents can be different to the isolation of Pacific from Atlantic eco-systems. In terms of climate, Drakes Passage between Antarctica & S. America wasn't simply open or shut. And it too may have played a part in mid-ploicene climate change. It had been open for some millions of years but was still widening in the mid-pliocene, having suffered a bit of a squeeze in the early Miocene (c25-15my), according to Lagabrielle et al (2009). For ocean currents, it seems size matters.
Toggle Commented Apr 12, 2015 on The Ns are calling the maximum at Arctic Sea Ice
Bill Fothergill, Paul Whitehouse, David Whitehouse? Thank you for correcting my mistake. Names is not my strong point. Chris Reynolds, Regarding wobbles, I think we look in different places. I have been conscious that the annual cycle and trends in that annual cycle can create wobbles with some forms of analysis. So the daily anomaly trace shows wobbles because the summer melt is recently far bigger than the average over the anomaly base period. One approach is to consider different times of the year in isolation, as you have done. Another is to average over a full 12 months, as I have done. One use of rolling 12-month averages was to divide the record into four, carry out a linear regression and subtract the trend (so any change in long-term slope is addresses), then quantify the wobbles by calculating the standard deviation. The results were (from memory, units?) 1.2, 1.5, 0.9, 1.8. Another use of rolling 12-month averages was to calculate the SD for short periods (24-month) as a way of addressing changes in long-term trend. The graphical result is posted here. Again it shows recent times more wobbly but not that greatly so. If you iron out the wobbles to give, say, a set of decadal values (actually 8-year values, ave SD, M sq km) they are 1980-88 0.093, 1989-96 0.119, 1997-2004 0.076, 2005-13 0.140.
Toggle Commented Apr 8, 2015 on PIOMAS April 2015 at Arctic Sea Ice