Hi AWP interested people,
The Albedo Warming Potential was
never intended to be used as a
melting forecast indicator. It's purpose was first to rank years differently from the September minimum. Only later it turned out be be a good indifcator for the
re-freeze season. Maybe I should include this on the website. It's been four years now since the original ungridded version and new people stumbling on it don't find my old posts.
First...thanks for the wonderful contribution. Love your site.
Thank you
I'm wondering if you would considering some kind of tool to validate your model in terms of its ability to predict sea ice volumes.
No. I have a sea ice forecast model which uses the same major albedo physics, but includes a few other "melt mechanisims". You could describe them as curve fitting variables. The albedo model on it's own is not good to predict actual sea ice loss.
Current extra factors in my 2020 modelHeat loss to space: otherwise it would never re-freeze
bottom melting: for stronger late august to early september melt than possible from albedo
fake cooling/heating layer: to replicate icedrift. Mostly Fram export, northern Beafort Sea and into Canadian Achipelago
CO2 levels: to correct heat loss to space over time. If I use a fixed value my model overpredicts iceloss in the 1980s and underpredicts ice loss in the 2010s. (This correction is 2-3 times as strong as the percentage increase in CO2 levels.)
https://www.arcus.org/sipn/sea-ice-outlook/2020/juneI investigated an Ice-Melt AWP model, which doesn't consider ice-free areas. It certainly gives interesting results and maps which look completly different, but I'm not sure it's any good as a melting indicator. I can share it in 1-2 days.
However we are currently in a situation where the CAB is made up of gray ice full of extensive melt ponds and hardly any snow, with an albedo which is probably 50%-60% and certainly not 85%. So in a typical July day it is probably receiving double your number at about 200 W/m2.
Are these albedo values just guessed? If you actually measure gray ice in paint you get 200/255 RGB values or equivalent to 78% albedo. I have a program to calculate the albedo of an image and even the melt pond areas are in the 70% range. I feel in general people tend to overestimate albedo drop from snowfree areas and meltponds. See my attched images for measurements.
he is assuming an albedo of 80% for ice and snow, and 85% for ice and snow in the High Arctic. For open water he is assuming 0% albedo
Close but not quite. My energy values already include water albedo. This is one reason why my high arctic energy values during summer solitice are below lower latitude ones. My early anamoly only model had an 80% value, but since I calculate absolute numbers (1-2 years ago) and not only anomaly values it changed.
0% SIC results in 100% absorption of my water albedo corrected energy values.
100% SIC results in 15% absorption of my water albedo corrected energy values.
Formula:AWPdaily = ((1-SIC) * MJ) + 0.15 * MJ * SIC
Example:MJ = 20
SIC = 75%
AWP = (1-0.75)*20 + 0.15 * 20* 0.75
AWP = 5 + 2.25
AWP = 7.75