...
I think what is missing here is a discussion of the impact of increased solar irradiance.
...
Are you talking about the reduced albedo, low cloud cover, something else?
Looking at only solar irradiance I believe we are at a solar minimum right now with a lower irradiance than in 2012 and most of the rest of this decade (even then that difference is tiny).
The low solar minimum that the earth is presently experiencing is not the key issue that I am talking about. Rather, the key issue is the amount of solar energy entering the system as a result of reduced albedo in the arctic. This difference far offsets the difference in energy coming from the sun due to the solar cycle.
I broadly agree, I have also the feeling that incoming solar radiation is probably a bit an understate factor. It is of course a know fact, and a spoken one, that lower albedo implies a greater heat accumulation. But perhaps that the big train of heat ready to ram us is not fully acknowledge. By the way this is why I'm back here, at ASIF. I'm like a vulture, when I smell the good fragrance of a water bath I am here.
So to continue this discussion, I will try something (each word of this sentence is important XD
). NCEP/NCAR Reanalysis is easier to manipulate and is update with a lag of a few days only. So I already us it, but radiations data are not as good as other dataset. Good enough to say something about big trend and to beat some dead horses, but probably not good enough for details -like the exact magnitude of the June 2019 crash-. MERRA is probably better but will not be available until mid or late July for the month of June. As already state, solar heat input in June above 70°N is a big factor :
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2016JD025819With MERRA data up to June 2018 I am going to back up this claim and give some order of magnitude, and with reanalysis up to June 2019 I am going to show that this year things are worst than worse. I will update the analysis when June 2019 data from MERRA will be available in late July -when nobody will no longer care as sea ice extent will be many thousands squared kilometers below 2012 and the crash will be beyond obvious XD -. Perhaps MERRA datas are going to show a little miracle, against the reanalysis, but is not really likely.
I am going to use two different flux. With MERRA, I am using net downward flux at surface (SWGNT for short). State an other way, it is the solar input wich is really accumulating at surface. The part of the Sun wich is not absorbed can be reflected by atmosphere and surface, wich is going back to space as a outgoing shortwave radiation (RSR for Reflected Sun Radiation). And Sun can also be absorbed by atmosphere. I stick with SWGNT cause looking at RSR implies giving weight to heat absorbed by atmoshere, but I don't think this part of the flux is really important. Its variations year over year are not as important, and heat absorbed by atmosphere is probably going to be mixed all over the hemisphere in a few days (as a side note, aerosols and soots -also known as black C- implies that Sun is more easily captured by the atmosphere, wich also have implications for global warming. But, looking only at sea ice year over year, the Sun captured by the atmosphere is not looking like a big factor). So, with MERRA dataset, is is going to be SWGNT. But Reanalysis as not an easy dataset for this flux. I could be possible to mix surface albedo with the downward shortwave flux at surface, but it is starting to look a bit too shaky, given the accuracy of reanalysis. So, with this lad, I am sticking with shortwave outgoing flux at TOA (RSR in the state paper above). Of course, the higher the heat absorbed by the surface, the lesser the heat making an escape to space. As a consequences, many graphs are going to have a left hand y axis, and a reversed right hand y axis. So far, the brains already hurts XD
The first graph is September SIE and SWGNT. The latter is reversed, meaning that the more the Sun is absorbed at surface in June, the less ice survived in September. Correlation is looking quite good, so let's check this.
The two datasets are quite correlated, with a decrease of the September SIE of 1 million square kilometers if absorbed radiation increase by 7 or 8 W/m². It should be note also that extrapolating the trend brings plausible results, with a zero SIE if June heat input is up to around 150 W/m². Definitively in the realm of possibilities.
So now that we have checked we are able to replicate the results of the above study, and that sea ice is screwed if Arctic is pounded by Sun in June, let's look at what the reanalysis is saying about June 2019. Was it bad, or worse than worst ? Short answer, acording to reanalysis June 2019 is abysmal. Values from reanalysis for outgoing solar flux at TOA are correlated with values from MERRA for solar input at surface. Correlation is not so bad (R² 0.45), but 2019 is not a record low point (caramba ^^). This said, it is looking like reanalysis is not going down enough. What I'm going to do is to artificially increase the trend for USWRF. Not for the pleasure of making things looking worst, but because, without MERRA data for June yet, we can only guess what happened. And an educated guess will be that reanalysis is to shy (not a surprise here...). Correlation is improved (R² 0.55) and 2019 is to the basement. Again, this is not intended to manipulate data to prove that June 2019 is a nightmare, but it is really because it is quite probable that reanalysis is not going down enough. Correlation with September SIE is also vastly improved (R² up to 0.43, from 0.20 with bulk values). Again, it is not a surprise that adding a trend to a datset to compare it with a dataset -SIE- where the trend is overwhelming everything vastly improves the correlation -if the slope of a dataset is way higher than its variability, we can correlated it with about any dataset having also a big slope-. But I do think this as a physical meaning.
I let you also the values for SWGNT in June in W/m², and with a conversion to "how many meters of ice could be melt by such and heat input ?" to give a sense of the energy in play.
SWGNT Thickness elt
1980 107 0.90
1981 109 0.92
1982 103 0.87
1983 103 0.87
1984 108 0.92
1985 113 0.96
1986 99 0.84
1987 108 0.91
1988 109 0.92
1989 99 0.84
1990 117 1.00
1991 111 0.94
1992 102 0.87
1993 116 0.99
1994 108 0.92
1995 109 0.93
1996 100 0.85
1997 109 0.92
1998 114 0.97
1999 102 0.87
2000 112 0.95
2001 114 0.97
2002 112 0.95
2003 108 0.92
2004 107 0.91
2005 113 0.96
2006 107 0.90
2007 118 1.00
2008 117 0.99
2009 111 0.94
2010 115 0.97
2011 117 0.99
2012 118 1.00
2013 109 0.92
2014 112 0.95
2015 113 0.96
2016 113 0.96
2017 116 0.99
2018 108 0.91
So, if I am not fooling myself, if I did not make any basic calc errors, etc... June 2019 has sucked up a lot of Sun, and probably is the leading horse in this race. Put in another way : die sea ice, die ! And see you again when MERRA will update.