Hi guys, I have a small problem when trying to calculate NSIDC september average myself: If I use daily values from nrt file, I get 3.58 mio km2 instead of 3.61 mio km2. What am I doing wrong ? Any help would be well appreciated.

@Peter Ellis: I'm getting increasingly depressed by the trend of commentors on this blog to ignore and denigrate published science and cling to whatever idea-of-the-moment seems most likely to prove catastrophic. I agree with you on this one. However, it is sometimes difficult to grasp 1 or 2 deg C change in temperature, as temperatures vary by much more with weather. But in reality, it is dangerous. For me, the following picture works fine: - This year we had about for heat waves here in Slovenia, nothing special, nothing dramatic, although there has been some drought reports (it was large drought but as far as I can recall, not the worst one). Let's say that during one of these heat waves the daily max temperatures were 32, 33, 34, 34, 35, 36 and 36 deg C. Now, this is a little bit uncomfortable, but not unbearable. Now, I just added two degrees of predicted warming for each day for same heatwave and tried to imagine, how this same heat wave would look like - the temps would be 34, 35, 36, 36, 37, 38 and 38. Now, this is much more uncomfortable - you can live without air conditioner when temps are bellow 35, but above 35, it is bad (and yes, plants do not have air conditioner). I think people do not realize that weather, which was a little bit extreme yesterday, it will become normal in the future. I think this is what people must somehow understand.

JAXA has reported about the record value: http://www.eorc.jaxa.jp/en/imgdata/topics/2012/tp120825.html

An interesting somehow funny, but sad comment on msnbc news about hitting record low: nteresting feedback loop rarely considered.... Fossil fuel usage increases CO2 levels in atmosphere, CO2 increases cause increased world wide temperatures, "Global Warming," Global warming causes Arctic ice cap to retreat, exposing region to more "Drill Baby Drill" opportunities, Fossil fuel usage increases CO2 levels...

Which is actually wrong, as it should be display current year and lowest year before current.

And if I saw it correctly, Arctic-Roos sea ice extent has also reached record low.

After revision for 24th, the record still holds. New provisional is at 4.09, so it may take just couple of day to reach bellow 4 mio. And I guess that average NSIDC extent may fall bellow 5 mio km2. Currently, it is 5.14 mio km2, and if I repeat last value of 4.19 mio for the rest of august, it is at 4.89 mio km2.

Geoff, your first link is broken, correct is http://www.bbc.co.uk/news/science-environment-19348194

Perhaps, at least for the fall, the effect of increased cloudiness can step in. I mean, with much more open water and thus warmer temperatures, cloudiness can increase substantially and can even block more outgoing radiation. This could be some additional positive feedback for ice loss.

It is the best if you just remove file name and use directory. There are 3 files in there: - Current data (nrt.csv) file - Climatology data (climatology_1979-2000.csv) - Historical data (final.csv)

Neven, pre 2011 values are also available at: ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/north/daily/data/NH_seaice_extent_final.csv

I have a problem with value of Bo Andersen. E.g. his value in pdf http://www.arcus.org/files/search/sea-ice-outlook/2012/08/pdf/pan-arctic/andersen.pdf he refers to his final values (He used Arctic Roos data) as 4.1 for Area and 5.6 for Extent. In text, he mentiones 2010 values. However, in prediction, it is mentioned that he predicts 4.1 for average extent. I am a little bit confused about that.

Dabize, I can agree with Lodger, it would be very interesting to us too. And you can get many "Nice work Dabize calls" for that :), let's say this is second one (after Lodger's)

A little bit off topic, but I think there seem to be large algae bloom visible north of Norway: http://rapidfire.sci.gsfc.nasa.gov/imagery/subsets/?subset=Arctic_r02c05.2012216.terra

I think quite a lot. If you switch to low resolution (4 km) on area 4x4 pixels (roughly the size of sensors), there are only 3 pixels of white enough to count this as an icy cell. But it depends on microwave image and there can be completely different picture based on that.

Tamino has posted some animations too. The story is very visible: http://tamino.wordpress.com/2012/08/03/arctic-sea-ice/

@Kevin: Am I correct that Maslowski's prediction for nearly ice free summer Arctic sea is 2016+/-3 years ? That would mean he is correct if Arctic is summer ice free as 'late' 2019. This is actually quite close to predictions for 2025 and actually, very close to "linear rate" decrease from PIOMAS. On the other hand, I personally think that even prediction for ice free Arctic in 2040 or even 2060 or even the "outdated" 2100 is extremely alarming. We must be aware of this.

Also, Ron Lindsay from PSC did his own prediction based on PIOMAS model for 4.06 +/- 0.42 million square kilometers: http://psc.apl.washington.edu/lindsay/Prediction/2012/September_ice_extent_2012.html

@A4R: I think it has been discussed here that last day IJIS data is not very usable due to huge corrections. In fact, it is better to skip 2 days for this reason.

Or maybe I just forgot about albedo being pretty high for the moment right ther.

@clouds keep things warmer right now by blocking out-going radiation: Well Neven, I am not sure about this being true. According to Insolation calculator, current insolation @75 N is 411.15 W/m2, which is I think more than which should be around 300 W/m2 at 0 deg C (black body radiation and it is colder than that now).

Hans: That same reasoning was heard in 2010, when extent at some point in April reached "almost normal" values. But, as it showed up in 2010, the extent quickly decreased that year. Late freeze up usually means early melt down. And in April, there is much more variability. As Bob has explained to you: Extent does not show you everything. Pretty normal would be reached when throughout the couple (meaning at least five or even more) of years, we would measure normal extent/area. Which I seriously doubt it will happen.

Sorry Neven, a sort of off topic, but I couldn't help myself sharing this here: http://www.youtube.com/watch?v=g7rZTZBOrqQ

Hat tip to you Neven. You can even write faster than I can read. Great, great, great :). Now, I will take my time to slowly digest huge amount of information you provided.

Well, about the threshold, it is not completely arbitrary though, it is constrained to values where you can reliable detect whether pixel has some ice within or not (so, you cannot put your threshold to low, neither too high, but you can vary it a little bit around some value). And note that sensors are capable of detecting only percentage of ice cover within pixels, but not very accurate percentage. So, measurement cannot tell you if there is 1 or 2 meters thick ice within pixel. And you should note that ice is not really randomly dispersed around Arctic, you will find quite solid ice pack north of CA/Greenland and towards Siberian cost and areas of really open water in southern seas north of Siberia/Alaska and somewhere in between, there will be some ice which is really dispersed (check slush puppie on this site). So all in all, extent will give you good image, how large area of Arctic ocean is covered by ice. So, although there is thickness ommited from measurements and some errors given, the value measured will give you relatively good picture of what is going on with ice as the distribution of ice is averaged over many years. Of course, and you are right here, volume would be preffered measure. But it is really, really difficult to measure it, as you have to measure it on max few cm level from distance of more than 700 km and so far, we had only limited measurement of it (ICESat from 2003-2007) and now we are waiting for Cryosat team to finish their calibration. In the mean time, PIOMAS model has been the best tool to estimate thickness and thus volume and it looks like we are now way below 2007. But it is not measurement. The only thing we are measure directly, on daily basis on whole Arctic is extent (and area). And people are convinced that ice is in worse state than in 2007 because of measuring extent and sporadic measurement of thickness in many places (although limited comparing to whole Arctic), checking PIOMAS model and also by comparing visual images of ice in many, many areas.