This is Rob Dekker's Typepad Profile.
Join Typepad and start following Rob Dekker's activity
Join Now!
Already a member? Sign In
Rob Dekker
California
Recent Activity
wayne said : But John and Rob, way off with respect to minima, namely current sea ice status worse than 2012! Not only in extent, but with the presence of dark seas more prominent, and unusually long Gyre current reinforcement as with respect to anticyclone hanging about for a month and a half. I don't know about the "long Gyre current reinforcement". Can you quantify that so we can run a regression analysis over it as a predictor for Sept SIE ? But I DO know about " the presence of dark seas more prominent". It is part of my estimation formula in the form of sea ice area. In general, the smaller the 'area' the lower the albedo, which correlates with lower September SIE in my method. So are melting ponds and boundary ice (extent minus area). Snow thickness (over ice or over land) does not affect albedo, so there is no physical reason to include it in the regression formula. It would just introduce a new variable which will increase the risk of (statistical) "over-fitting". Remember Von Neumann's famous saying : "With four parameters I can fit an elephant, and with five I can make him wiggle his trunk." So we want to include enough variables to make a valid prediction, but not a single variable more. I use three variables in my prediction method (that all affect albedo and thus absorbed solar energy in the Arctic) : sea ice extent, area, and land snow cover.
Toggle Commented Jun 8, 2018 on PIOMAS May 2018 at Arctic Sea Ice
Tealight at the ASIF informed me that NSIDC changed their definition of ice extent and area : https://forum.arctic-sea-ice.net/index.php/topic,103.msg157655.html#msg157655 So I re-ran my regressions using the new V3 data and ended up a bit lower than with the old V2 data : Prediction of September sea ice extent, based on sea ice area and extent, as land snow cover is now 4.65 M km^2. Standard deviation over the residuals is 470 k km^2, which is still showing significantly better skill than linear decline.
Toggle Commented Jun 8, 2018 on PIOMAS May 2018 at Arctic Sea Ice
I will submit my prediction to the SIPN this year again : https://www.arcus.org/sipn
Toggle Commented Jun 6, 2018 on PIOMAS May 2018 at Arctic Sea Ice
Darn. The Rutgers snow lab picture is just cut-off for the last year(s). Please open in a separate tab, and notice that snow cover in May was quite high when compared to the past 10 years. Or look at the source directly : https://climate.rutgers.edu/snowcover/chart_anom.php?ui_set=1&ui_region=nhland&ui_month=5
Toggle Commented Jun 6, 2018 on PIOMAS May 2018 at Arctic Sea Ice
Most of you know that I am using Northern Hemisphere snow cover in spring and summer as a predictor for the September minimum. Here is a guest post on ASIB of my method : http://neven1.typepad.com/blog/2013/07/problematic-predictions-2.html This method of prediction works really well for June data, but even for May data it has some skill. Now that Rutgers Snow lab published the Northern Hemisphere snow cover for May, I ran my regression formula on the 1992-2015 training period, using May data for snow cover, ice concentration and ice area, and arrived at a prediction for September sea ice extent minimum of 4.84 M km2. This number makes sense, since snow cover in May was still fairly high (compared to the previous 10 years), and even though sea ice extent in May was at a record low, sea ice area is actually just 3rd or 4rth lowest. This means the ice pack is still fairly 'compact' which reduces the amount of heat the ice pack will absorb from the ever higher sun in the Arctic. Standard deviation over the prediction is 460 k km2, which is substantially better than a linear decline as a predictor (which has a standard deviation of about 550 k km2). So the method has some skill with May data, but for an accurate prediction, please wait for the start of July, when the June data is in, since that has real skill with about 300 k km2 standard deviation. Until then, the prediction of 4.84 M km2 stands.
Toggle Commented Jun 6, 2018 on PIOMAS May 2018 at Arctic Sea Ice
To explain the difference between Zwally's assessment and GRACE observations, it is suggested here that Zwally assumed the increased snowfall over Eastern Antarctica to have the density of ice, while in reality of course the density of snow is about 1/3rd : https://www.scientificamerican.com/article/what-to-believe-in-antarctica-rsquo-s-great-ice-debate/ That would explain the difference between Zwally's assessment and actual observations (by GRACE) of the mass of the Antarctic continent ice shelf.
Toggle Commented Apr 21, 2018 on PIOMAS April 2018 at Arctic Sea Ice
Regarding Antarctic ice loss, the GRACE satellites recorded the following : Research based on observations from NASA’s twin NASA/German Aerospace Center’s twin Gravity Recovery and Climate Experiment (GRACE) satellites indicates that between 2002 and 2016, Antarctica shed approximately 125 gigatons of ice per year, causing global sea level to rise by 0.35 millimeters per year. https://svs.gsfc.nasa.gov/30880
Toggle Commented Apr 21, 2018 on PIOMAS April 2018 at Arctic Sea Ice
Thanks for the update Neven. It is indeed quite spectacular what is happening in the Bering. It would be interesting to keep an eye on how this anomaly works its way into the ice cover of the Chukchi sea going into the melting season. P.S. The Arctic Sea Ice Forum seems to be down. I get this message : Table './arcticse2/smf_messages' is marked as crashed and last (automatic?) repair failed
Toggle Commented Apr 1, 2018 on Bering goes extreme at Arctic Sea Ice
John, my prediction method, using NH snow cover, sea ice extent, and sea ice area as predictors, is useful with data from the end of May, but statistically significant only with data from the end of June. Sorry, I still have not found any variables that make a statistically sound prediction earlier than that.
Sorry. That was ASIF.
Thanks Neven ! I for one am glad that the melting season started. It will put the focus back to climate and science after a winter of arguing politics on the ASIK :o)
Very nice update, Neven, thank you ! Regarding that PIOMAS-CryoSat divergence issue, Your theory makes a lot of sense ; that thick snow cover may fool the Cryosat freeboard measurements into believing there is thicker ice (and thus more volume) than there really. If so, PIOMAS is likely more accurate than Cryosat for years when there is a lot of snow on the ice. However, just yesterday I read this paper : http://onlinelibrary.wiley.com/doi/10.1002/2017GL075494/full They note that is snow load on the ice is so large that the weight of the snow pushes the ice under the water level, then sea water will creep into the snow and form what they call snow-ice. The snow layer does not get thinker after that, since for every kg of snow that falls, an almost equal amount of snow-ice will be formed. They note that this effect of snow-ice formation occurs with heavy snowfall, and is actually the main reason of ice formation over second-year ice, If the PIOMAS does not account for that effect (of snow-ice formation under snow), then it may actually be Crysat that is more accurate. Something to keep in mind.
Toggle Commented Mar 12, 2018 on PIOMAS March 2018 at Arctic Sea Ice
Jim Hunt wrote Whilst I do appreciate that people publicly peddling porky pies do need to be publicly corrected, do you think it might be possible to try and stick to the Arctic/Antarctic theme in here? I'll try to restrain myself, but when disinformation pops up, that's hard. Either way there is a good thread on the forum that discusses more details, if (the other Jim) would like to continue the conversation : "Cars, cars and more cars. And trucks, and...." https://forum.arctic-sea-ice.net/index.php/topic,438.msg136297.html#msg136297
Toggle Commented Dec 15, 2017 on PIOMAS December 2017 at Arctic Sea Ice
Jim said : The idea that electric cars will save us from CO2 is a pet peeve of mine. Rather then rant about the issues here is a very good blog post that describes the problem: http://www.aviewfromthecyclepath.com/2017/09/business-as-usual-by-driving-green-car.html When comparing carbon intensity of ICE cars to EV, if they come out about even, you ALWAYS have to check the facts. As in this case too. Your link tells about the original Mini : Burning a litre of petrol results in 2.3 kg of CO2 being released, hence a gallon results in 10.4 kg of CO2 or as it is usually expressed these days, 162.5 g/km. and about the Leaf : The Leaf requires 12.4 kWh of electricity to cover 64 km, and that results in 155 g/km of CO2 if the car is charged from the Dutch electricity network. Now, we can argue if the original Mini and the Leaf are comparable automobiles (one could argue that the Mini is a Mr.Bean car where you can only comfortably drive it if you are sitting on the back seat). And we can argue if the 12.4 kWh for 64km is comparable, since EVs have a much better efficiency in city traffic than ICE vehicles. But even with all that assumed equal, there is still a BIG problem with this story : That 12.4 kWh causing 155 g/km data is not referenced anywhere. So I did some digging myself and found this scientific paper : https://www.sciencedirect.com/science/article/pii/S1361920916307933 That one clearly states that on the Dutch grid, the carbon intensity of supplied electricity is about 547 g/kWh. For 12.4 kWh that is 6.782 kg CO2, spread over 64 km is 106 g/km. That is some 40% better than the Mini on petrol. And since the Netherlands is phasing out coal and increasing wind/solar, that number will only get better over time. Fact checking rules.
Toggle Commented Dec 14, 2017 on PIOMAS December 2017 at Arctic Sea Ice
DCS, I hope you are right.
Toggle Commented Nov 30, 2017 on PIOMAS November 2017 at Arctic Sea Ice
The question is especially important, since so far the statistical evidence suggests that 'volume' loss is still accelerating : https://sites.google.com/site/arctischepinguin/home/piomas
Toggle Commented Nov 28, 2017 on PIOMAS November 2017 at Arctic Sea Ice
DCS said : In light of the content of some of my comments under the previous article, I’m projecting a significant deceleration in the decline of Arctic sea ice volume, if that hasn’t already happened. You seem to advertise the theory that summer heat melts only 'area', not 'volume'. Do you have any evidence that that "already happened" ?
Toggle Commented Nov 28, 2017 on PIOMAS November 2017 at Arctic Sea Ice
That's good work DCS. Thank you ! You even made both A''(t) and T''(t) negative. It's still a mathematical exercise though. If this scenario will pan out in real time or not will depend on the bigger question : Will summer heat melt 'volume' or will it melt 'area'. If it melts area than you will be right and volume will decelerate. But if it melt volume then volume will just go down (linearly) and at some point, as Wadhams stated before : "In the end, it will just melt away quite suddenly". Area that is.
Toggle Commented Nov 26, 2017 on PIOMAS October 2017 at Arctic Sea Ice
DCS said Proof: Assume that A(t) and T(t) are each continuous and twice differentiable. Note that V = A T. Then V'(t) = A(t) T'(t) + A'(t) T(t), and V''(t) = A(t) T''(t) + 2 A'(t) T'(t) + A''(t) T(t). Assume that A(t) > 0, A'(t) < 0, A''(t) ≥ 0, T(t) > 0, T'(t) < 0, and T''(t) ≥ 0. Then A(t) T''(t) ≥ 0, 2 A'(t) T'(t) > 0, and A''(t) T(t) ≥ 0. It follows that V''(t) > 0 and that the decline in V is decelerating. You don't know if A''(t) ≥ 0 nor do you know if T''(t) ≥ 0. You are just assuming that. So what remains is your statement : "If Arctic sea ice volume were to decelerate significantly (which I'm expecting, maybe that is wishful thinking) "
Toggle Commented Nov 25, 2017 on PIOMAS October 2017 at Arctic Sea Ice
D_C_S, you did not read that article all the way to the end.
Toggle Commented Nov 14, 2017 on PIOMAS October 2017 at Arctic Sea Ice
Excellent post by Tamino : for anyone who thought that Arctic sea ice decline slowed down after 2007 : https://tamino.wordpress.com/2017/11/02/cycle-change-arctic-sea-ice-edition/ I’d like to congratulate all the readers who have read the last two posts. You now know a helluva lot more about Arctic sea ice and how it has changed, than most people. You know: - It has declined dramatically - The decline has not recently stopped or even slowed. - The decline isn’t just a summer/fall annual minimum thing; it has declined throughout the entire year. - The annual minimum decline dropped steeply in 2007. - The geometry of the continents (of land areas in general) affects the relationship between the sea ice edge latitude and its extent. I’d guess that you are now prepared to discuss the issue intelligently. I’ll opine that you are far better prepared to discuss it intelligently than any of those who have recently been claiming it isn’t still declining.
Toggle Commented Nov 11, 2017 on PIOMAS October 2017 at Arctic Sea Ice
Hans said about my (5.4) prediction of the average NSIDC September extent : Let's say we're determining if a prediction is accurate by way of using the Sept. average, but since that is more forgiving, let's say the range of accuracy needs to be within .25 million square kilometers, then .5 is double that standard and thus inaccurate. You are right that my prediction is not accurate. In fact, it will probably be more in the range of 4.8 - 4.9. We will see at the end of September. Note though that this was a June prediction, and the standard deviation of that prediction is about 350 k km^2. So 4.8 - 4.9 is still well within the 95% uncertainty margin. Also, there is an identifiable reason why my prediction was too high : My method does not include ice thickness, which was low this year due to the anomalously warm winter. Lewis did a good piece on that : http://neven1.typepad.com/blog/2017/08/guest-blog-arctic-prediction-models.html Either way, I do think my prediction method is not good enough yet. And I'm looking for better early indicators that can improve the model.
Toggle Commented Sep 26, 2017 on PIOMAS September 2017 at Arctic Sea Ice
Hans, About this polynya, I presented a different theory, that it may be caused by the Coriolis forces driving the Atlantic warm water upward while making a U-turn at that very location against the Gekkal Ridge : http://forum.arctic-sea-ice.net/index.php/topic,1834.msg125728.html#msg125728
As Wipneus reports on the ASIF : According to NSIDC NT sea ice concentration, extent in 2017 is currently the 8th lowest, area is 7th. Looks like we are not going to see any records broken this year.
Lewis, thanks. One point of clarification : I include land snow cover in my model since it clearly affects albedo and thus the heat budget over the Northern Hemisphere. Snow on ice affects the albedo much less and thus in my opinion is not that important for the heat budget.