Sure, ask a billion questions: but why ask what needs to be asked? What, indeed, needs to be asked? To quote myself (the rest of this piece will be aired tonight): When will the ice be gone and THE BOMB revealed? That is very much the question. In fact, you don't need ice to be gone in the dark polar night of the winter in order to release this bomb. It's enough that the ice is gone when the Sun returns in April. That will of course happen even earlier than the first year of no ice at all, even no ice in winter. What I did in November in order to find the answer to WHEN, was to ask my apps a different question. Instead of asking the apps when we last saw a similar collapse rate in the PAST, I now asked them how soon in the FUTURE we will lose all ice with that current collapse rate.

If you believe your extrapolations and that the Arctic will be ice-free year-round by 2021 or 2023 or 2025, you also believe that civilisation will collapse (and humanity will probably go extinct under such a scenario), because that's what such a rapid change would mean. No two ways about it. And if you believe that Essentially the exact same question as Thursday, so I'll provide my Wednesday answer for the 3rd time. My app may of course be wrong, but with apps for extent and volume consistently outputting 2021–24 for either daily, weekly or monthly, since November 1st, I would say these 4 years constitute The Ballpark for what we must expect. Things could unravel faster than linear, however, and then the 365 ice–free will be here 2020 or earlier. Or the other way around, there could be mechanisms that slow things down, and we're looking at 2025+. My apologies to everyone for having to repeat this so many times, but demand seems to be virtually endless.

Neven, you seem to have a very hard time understanding what an average is, what an ad–hominem attack is, what Wishful Thinking is, what testable scientific method is, and also why people are concerned about Arctic sea ice collapse. Not being the most naive person on the blog, I have no illusion that any of these things can be explained in any way so that you will get them. People are what they are and they particularly don't appreciate having to change the way they see things. The Arctic will change very very slowly until BANG! it doesn't change slowly anymore. Kicking out all the nay–sayers from your forum and blog won't stop the bomb going off, but I'm probably wasting my time already trying to explain this to someone like you. You never make fun of yourself when you launch ad–hominem attacks against me, Neven. They're always against me, not against yourself, you are smarter than that. We all know you're not really trying to kick yourself out of your own blog.

I just find it one of the puzzle pieces that is harder to grasp (especially in the way you present it), but I'm probably just too stupid. Neither ad–hominem attacks nor Wishful Thinking will make Arctic winter white again, Neven. I'm just a messenger, so you no other choice but attacking me!

Neil, I found your explanation today of the uselessness of IPCC–style climate scientists to be more than just a little bit enlightening: Over and over again, during the debates, Gavin Schmidt has been seen to say that these annual variations must be ignored if we are to see the larger picture. And he's right, the larger picture is 100 to 1,000 years. So what they do is bury the annual variations in decadal averages and then bury them again in multi decadal averages. Truth be told, if you take the 30 year running mean, we're pretty much on target. When you look at the 2000's averaged out. Even the 2010's, when the decade is done, will be averaged with the previous two decades to create the 30 year running mean. The problem with this methodology, which is used by all climate scientists when they report to the IPCC, is that it fails to anticipate, or even detect, step changes when they happen. In fact it's designed to do exactly that, remove them. The major problem with that approach is that what is happening to the Arctic is massively driven by annual variations and those variations are getting larger as every decade goes by. By the time that the 2010's annual variations are released from the 1990's, it will already be blindingly obvious to everyone that they are out of touch. Also the model will mitigate to tone down even those effects. In reality the 30 year running mean has been a wonderful ruler for measuring future change over the last 5 decades. It' was extremely useful in the denialist rantings in the aftermath of the 97/98 nino and the return to the norm which happened there. It forced the denialists to take out the 97/98 as a baseline and then their entire assertions fell apart. So, I think, when railing at the "Scientists" for not predicting what we are seeing now, I respectfully submit that their models and their projections are specifically designed to ignore it. Because, so far, by ignoring it, they have been more right than wrong. Honestly I feel that an ice free arctic in 2022 will force them to reassess that. Because the possible forcings created by the black swan event are enough to overwhelm the 30 year running mean and to continue with it would be foolish. They would need to create a new baseline and then run a parallel comparison and draw conclusions that way. Getting scientists to throw away long held and very good baselines will take an extreme act. The same extreme act we see evolving before us in an unprecedentedly warm winter with unprecedentedly low ice volume and extent. What I'm saying is "Don't allude motives to the Climate scientists just because they are not monitoring the same thing you are". Because, in the end, these people have been in the firing line for a long time and the vast majority of them are both honerable and extremely thick skinned. But, believe me, they have feelings too. I'm guessing they will have to go directly from 30–year means to monthly, quarterly, half–yearly and yearly drops in Annual Average Volume of Arctic sea ice when that elusive Black Swan starts to sing. And that it will be a rough time for the milieu that started hating annual averages the moment they first heard of them, but that they will adapt and mitigate once people with vested interests in climate change start using them.

But you can't just put a straight table like that and say "see no such cycle exists". LOL, Neil, call me a Cycle Skeptic, as I think it's all falling apart, including this alleged Cycle itself. One of the reasons for that which I forgot to mention is this: "Going lowest in 2017 might be followed by a 2019–20–21–22 rush of new records for minimum volume" is not entirely accurate, as we could see Blue Ocean Event uncharted waters as soon as 2017, and a number of math–savvy commenters have pointed out earlier that you cannot go lower than ice–free (although temperatures may be able to exceed 0C water, with some help from our star). In a nutshell, I think we're collapsing way too fast for any such patterns to be able to repeat, or carry out a new Cycle of 5 years, sans collapse. Collapse itself is collapsing: When this very blog ceases to emit a certain level of (unfounded, IMO) optimism, it will cease to be.

Sam, Greg, J_b_t, Neven (et al): Since the January 2021 shock comment, at the latest update and blog post we have a record slow zero ice prognosis of only minus 87 thousand km2 per month, which translates to April 2025. Gives everyone a bit more time to think about what we've done as a species/collective. (However there is a growing gap between volume based and extent based estimates, echoing some people's statements that JAXA extent may be a suboptimal measure of sea ice collapse.) Neil, After the 2007 extent record low, only 2012 set a new record (Year 1 in new Cycle): $elow[1] = 3177455; // 2012 $elow[2] = 4065739; // 2007 $elow[3] = 4257003; // 2015 $elow[4] = 4269199; // 2011 $elow[5] = 4500623; // 2008 $elow[6] = 4622092; // 2010 $elow[7] = 4809288; // 2013 $elow[8] = 4884120; // 2014 $elow[9] = 5054055; // 2009 $elow[10]=5179300; // 2005 Emerging details/trends within the Pattern also suggest that as 2013 AAE went on to go lower than 2012, without 2008 AAE going lower than 2007, then both 2018 and 2019 AAE may go to new all–time lows (even lower than the record low 2016 AAE, and, perhaps the new low in 2017), breaking down the so–called Cycle quite a bit. Another break–ball is the amazing October–December plunge of 2016, that is only matched by 2007 and not 2012, which was less dramatic. In Cycle lore, this plunge was supposed to have waited till autumn 2017, but here we are.... After the 2007 volume record low, 2010–11–12 all set new records (Year 4 + 5 in Cycle, + of course 1 in new Cycle). $vlow[1] = 3673; // 2012 $vlow[2] = 4302; // 2011 $vlow[3] = 4582; // 2010 $vlow[4] = 5392; // 2013 $vlow[5] = 5670; // 2015 $vlow[6] = 6458; // 2007 $vlow[7] = 6810; // 2014 $vlow[8] = 6832; // 2009 $vlow[9] = 7072; // 2008 As we know volume is the real deal, volume is perhaps what should be focused on for sea ice collapse. With the increased severity and abruptness, going lowest in 2017 might be followed by a 2019–20–21–22 rush of new records for minimum volume.

Obviously we'll have to see how the spring melt goes before having a better view, but, if you subscribe to the model, then it would be easy to predict that 2016 was not going to break any minima records and that 2017 would. Neil, I don't subscribe to that nor any other 'model'. What has been recurring in 2006, 2011 and 2016 is that all of them go lowest ever (so far) in annual average extent. After each of 2006 and 2011 the next year went even lower for AAE, and that year also broke the 1–day September low record. I think 2017 will be lowest ever both in terms of JAXA AAE and 1–day Sep low. Not because of the pattern, but because of how this winter looks so far.

Either way, I hope that year-round ice-free conditions are well into the future. Or else, why would we even bother doing anything? Year-round ice-free conditions in 2021 would mean civilisational collapse, as the rate of change would inherently be much faster than it already is. That makes an ice-free September look rather benign. Is that the idea? Neven, I don't know how to respond to that. Final ice loss will be a rough time for everyone and every living being on this planet.

[Rob Dekker is possibly the last person here who needs to be told how to do statistics or make a graph. No need for condescension; N.] Maybe so, Neven. I was merely showing him how he and everyone else could get what he wanted. I totally accept his request, however, and I've already put it on my (very) long–term plan for further graph developments for the Arctic sea ice collapse.

Doesn't mean it's wrong, but chances are that it's a crap shoot. "PIOMAS Annual Average Volume is down 7.37 km³ per day — If sustained, the first ice–free 365 days would commence in January 2021" isn't wrong, Neven. If collapse rates are sustained, there will be no Arctic sea ice starting in January 2021. You may say that you don't believe they will be sustained, in which case you and I would agree. Here's what I (already) stated about that very likely eventuality: My app may of course be wrong, but with apps for extent and volume consistently outputting 2021–24 for either daily, weekly or monthly, since November 1st, I would say these 4 years constitute The Ballpark for what we must expect. Things could unravel faster than linear, however, and then the 365 ice–free will be here 2020 or earlier. Or the other way around, there could be mechanisms that slow things down, and we're looking at 2025+.

I get your anxiety, hombres. So if I make the "prediction" that next year (2016+1) will be 2017, that's also a model, right? Because I use math. Oh well, my bad. I keep forgetting how unpopular that may be in some places. Rob, you may indeed. Anyone is urged to dive into the matter of annual averages and all their applications for predicting the near and far future. I find it highly interesting, yet my choice has been to zoom in to see interesting things, rather than zoom way out and lose all the detail. [Rob Dekker is possibly the last person here who needs to be told how to do statistics or make a graph. No need for condescension; N.]

13800/7.37 isn't a model but a simple division, Neven. And let's not split hairs here, even though that's more pleasant than focusing on the message. Josh, you may indeed. Anyone is urged to dive into the matter of annual averages and all their applications for predicting the near and far future. I find it highly interesting, yet my choice has been to zoom in to see interesting things, rather than zoom way out and lose all the detail. Here's how you do it: Download the full dataset from either PIOMAS (sea ice volume) or JAXA (sea ice extent), import to a spreadsheet and tell that sheet to give you an average for 365 days of ice. Copy or repeat and tell the app to draw you a graph. Best of luck! :)

Vid, if you subscribe to the 5 year cycle like Hans and I do, your model gets a 2 year kicking every 5 years. Neil, I don't subscribe to any cycle, and I absolutely do not have a model. If volume keeps collapsing at the current pace or even faster than this linear decline, there's simply no ice left to last through 2017–2021. Climate is changing in the North, so we shouldn't expect it to stay the same just at slightly lower ice levels. Uncharted waters means we haven't been here before and we don't know what we should expect. I know it's unsettling, but clinging to what used to be just isn't the right way to stay safe these days.

But a year ice-free at the north pole (thankfully!) requires a lot more CO2 than we've yet released. It's just not a physically possible scenario under current conditions. Greg, as J_b_t pointed out already, there are a vast amount of feedback loops, and CO2 isn't the only show in town. My app may of course be wrong, but with apps for extent and volume consistently outputting 2021–24 for either daily, weekly or monthly, since November 1st, I would say these 4 years constitute The Ballpark for what we must expect. Things could unravel faster than linear, however, and then the 365 ice–free will be here 2020 or earlier. Or the other way around, there could be mechanisms that slow things down, and we're looking at 2025+. Also, we're not talking about 'current conditions'. Here's Wikipedia on that last element of 'climate change': The process or end result of becoming different

Suffice to say that if the current trend continues and the ice doesn't get a chance to thicken sufficiently this winter, things will get even more worrisome than they already are. That's putting it mildly, IMO. With the new PIOMAS data I ran the Annual Average Volume script again, and it suggested at current pace all ice all year round would be gone starting January 2021. Good luck with your sabbatical, but we've stepped outside of the cliff now, and we've even looked down and realised our mistake, thus we begin the fall.... Click graph for full blog post, or here for the youtube teaser.

Say you inherit €9 million and carelessly spend €100,000 a month, you will eventually run out of currency. Your calc app says 90 months which is 7 ½ years. In the Arctic we have 9.77 million km² annual average extent (AAE) of sea ice, and we're "spending" about a 100,000 km² of this "inheritance" each month. We can spend or lose AAE all year round because it's an annual average, and an annual average removes seasonal variation. Therefore a rapid collapse with monthly 100k losses can last from July to October (2007), or August to October (2012), or indeed from October to December (2016). Going from melt to refreeze during such a lasting plunge (2007 & 2012) is NOT problematic for the Arctic (only problematic for debaters if they did not still learn the basics of an annual average). In the Arctic we got 9.77 million km² annual average extent (AAE) of sea ice, and we're "spending" about a 100,000 km² of this "inheritance" each month. Calc app says we've spent it all after 97.7 months. This is 8.1 years, and from today that takes us to 2025. What this calculation means is AAE would be 0 km² in 2025. But because AAE for a specific date is the average for the 365 days ending with that specific date, the start of that very first 365–day period of no sea ice at all would be a year earlier, in 2024. This is why the blue column for November stops at 2024. The loss per quarter (red) is slower and leaves winter ice lingering until 2028. Loss per half year (orange) lets ice last to 2034, and per year (green) to 2042. [For volume the outlook is even darker: 2021, 2022, 2023 & 2024 for monthly, quarterly, halfyearly & yearly, respectively. Extent collapse trends are merely catching up with ice volume collapse.] What happens when we've spent all the ice we inherited? Click graph for full blog post, or here for the youtube teaser.

As for "We will get the Arctic ice back"... It isn't really so hard as you would believe. We just have to reduce our net emissions. [...] We just have to slow our emissions, and things will begin to correct themselves. Glenn, Nature also has a huge amount of very strong positive feedback loops, making the 'Make Arctic Winter White Again' campaign an uphill battle. We've tried Wishful Thinking and oligarch controlled fake democracy for half a century. It didn't work. Sadly, humans aren't all that creative on a species level, and we're out of options when those two methods didn't work.

"Make Arctic winter white again!" I'd vote for that president in 8 years.

Neil, I must say I too like the parallel of the 'slow moving truck' very much. I just don't believe in it, because the data say otherwise. Hence the 100 m left of the fall against the rocks, and the clicking of 'likes' on the iPhone for either Donald or Hillary.

I might start a book on just how long the sabbatical will last. ;-) I shall avoid the topic of how long, but must admit I find the concept very literary. But then everything is when you write. I have a protagonist in my cli–fi who's writing a cli–fi novel but worries mostly about his readers — a complete no–no for any author, but I guess natural for a cli–fi writer — and about getting his book out there in time for their to be living beings with reading abilities. Another figure, maybe his brother, may be trying to take a break from the ice, like the swells crushing it and the dipoles flushing it out the Fram he tries to flush all of it out of his mind. Perhaps only to find he spends more time trying to not think of ice, than he use to spend thinking of ice and abrupt climate change.

The correct way to deal with them is to say that the truck is moving at 1 mph and it will hurt if you don't move away from its path. Yes. I would use a slightly different setting: You forgot you parachute and have already fallen 900 of the 1000 meters from your plane, you can see the Rocky Mountains approaching fast. Holding your smartphone you can either like a status from the Blue Party or the Red Party. Before crashing into the rocks. Every similar or unsimilar form of politics at this point — including trying to explain anything at all to (fake) skeptics — is as pointless as that final like. Having the best available estimate for when the Falling Man will hit the rocks — or SHTF — is slightly less pointless. At least in our case we can choose to use that information to move to higher ground or more rural areas. In general, I think way too many people worry way too much about 'the debate', the debate is insignificant. At least at this point it is, ref the 900 meters already fallen. If we can just stop worrying about the debate, we perhaps focus on a better estimate for the inevitable?

John, wrong on all counts, I'm afraid: - A steep drop in AAE is more easily attained in Aug/Sept, where the remaining ice is very vulnerable to wind action - The steep drop for 2007 AAE continuing in October only happened due to high SIE increase in Oct. 2006 - Since SIE currently is higher than 2007, and 2015 SIE did not catch up with 2006 SIE before Nov., there is virtually zero possibility for the 2016 AAE drop to mirror that seen in 2007. And yet, mirroring of 2007 plunge is *exactly* what we see, or at least no other year than 2007 can match the devastating 2016 fall–to–winter plunge. Click for full size.

I think the extreme alarmism view, a justifiably scary no sea ice at all scenario, is overrated, it makes good pr against AGW, but it will take too long before it happens. Wayne, I can see where you're coming from, but 1) I don't think you have enough data to back up the 'alarmism' claim, and 2) I don't think you have that data arranged in easy–to–play–around–with variable arrays. If you did, you would see that the monthly, quarterly, half yearly and yearly drops for annual extent are quickly catching up with the same for annual ice volume, and just taking a linear view of those collapse rates gets us to zero ice shockingly fast. That's in the data itself, and I also believe no–one can guarantee us it won't collapse even faster than linear. 'Alarmism' of course means causing needless worry. I don't think you really mean to say there's no reason to be worried.

and other cyclemongers For the record, Philip, I don't personally reckon we have enough ice anymore to go through another 'cycle', as you call it, presuming you are referring to the 5–year cycle apparent in the recent decade or so of sea ice data? I also personally don't agree to the notion that everyone should now lose interest in ice 100%. One enthusiast is taking a sabbatical, or trying to, that's about it. From the data itself this is the most interesting time in the history of this blog (and its adjacent Forum), so for it to be not about ice at all at this exact point would be novelle, indeed!