From:
Slater, T., Lawrence, I. R., Otosaka, I. N., Shepherd, A., Gourmelen, N., Jakob, L., Tepes, P., Gilbert, L., and Nienow, P.: Review article: Earth's ice imbalance, The Cryosphere, 15, 233–246, https://doi.org/10.5194/tc-15-233-2021, 2021.
https://tc.copernicus.org/articles/15/233/2021/tc-15-233-2021.pdf
<snip> RE: the table shown below comparing Gt/year ice loss between periods. I find it odd that the amount of ASI loss in 2000s is higher than for the 2010s. Even multiplying the 2010s number by 1.1 since it only accounts for 2010-2018 (missing 2019) gives a much smaller value of ca. 103 vs. the 384 for the 2000s.
Wipneus' PIOMAS volume graph for Sept. minimum volume shows a change from 2000 to 2009 of ca. 11 to 7 M km3, and ca. 7 to 4 for 2010 to 2019. That is not directly relevant since the study used the winter (October -April) average ASI volume trend not the summer minimum trend for their ice loss measure. But I don't see how the losses in 2000-2009 could so much exceed the 2010-2018 value. And even the 1980s and 1990s rates of winter ASI loss are almost 1.5 and almost 3X higher than the extrapolated Gt/year for 2010s.
Even though it is measuring a different month of the year, I do not see why the Wipneus Sept minimum chart would show a consistent trend in losses while this study apparently finds a much lower rate of ASI loss in the 2010s.
I posted this in the What's New in the Arcitc thread. Kassy suggested moving to the "When will the Arctic Go Ice Free" thread.
Here is the subsequent discussion:
Kassy: Well the winter ice was thicker overall so there was more to melt out then.
Glen: Yes, but there was also a lot more summer ice in the earlier decades, yet the PIOMAS volume (e.g. Wipneus chart) shows a steady rate of decline. So why would winter (Oct-Apr) losses show a 2/3 reduction in loss rate for the most recent decade compared to prior decades, while September minimum losses continued to follow a consistent trend?
Kassy: They are not so much winter losses as snapshots of the overall loss only taken in winter.
Back to me this time: I still don't get it. Why would the trend for Oct-Apr average volume values be so radically different from the trend of Sept. minimum values?
OK, I can almost get that if within a single time period (i.e. one decade in the table) there were strong winter compensatory increases after September minima each winter. That way, even with a trend in Sept. minimums, the Oct-Apr values would not change as much. Which I guess is what's happening, but the strength of that compensatory factor is surprising.
Restated: Why do Oct-Apr losses show a strong rate of decline in 1990s and 2000s, then decline to about 1/3 of those loss rates per year in the 2010s, when the September minimum trend continues going down?
Here is what I see in the Wipneus monthly PIOMAS trend lines for some example months.
Granted the exponential trend overtates the rate of decline and is not accurate for the most recent 3 years, but I think it does so more or less equally for all months and still captures the bigger comparison between decades well enough (values rounded for clarity):
SEP: 1980: 16 1990:14 2000:11 2010:8 2020:3
Nov- 1980:20, 1990:18, 2000:15, 2010:12, 2020:8
Jan- 1980:26 1990:24, 2000:22, 2010:19, 2020:15
Mar- 1980:31 1990:29 2000:27 2010:23 2020:20
Which gives these decadal declines
SEP- 1980-1990:2 1990-2000:3 2000-2010:3 2010-2020:5
Nov- 1980-1990:2 1990-2000:3 2000-2010:3 2010-2020:4
Jan- 1980-1990:2 1990-2000:2 2000-2010:3 2010-2020:4
Mar- 1980-1990:2 1990-2000:2 2000-2010: 4 2010-2020:3
Part of my brain is saying, "Of course there is no difference in the decadal trendline values dummy, each trend line is drawn to smooth out any such differences." Which is true, but the individual years within each decade for each month are roughly 50:50 above and below the trend line through the trendline in that decade, so I don't think that is a major problem.
Partial brain fart: The individual years are NOT 50:50 for the final years of the 2010-2020 trendline. Instead for each of those month/years (and also Oct,Dec,Feb,and Apr), the individual monthly values for the final 1-3 years are above the trend line, thus showing a lower rate of decline than the trend line for the final period 2010-2020, which is the whole point. But that is also true for September, so I think the original argument still stands.
In summary: I don't see a big drop off > 66% in the rate of decline Oct-Apr monthly PIOMAS between 1990-2000 and 2000-2010 vs. 2010-2020. But that is what is indicated by Slater et al. (though their data stop in 2018).