Declining Arctic Sea Ice

You've got a conspiracy theory to explain everything.

Since that's growing boring, there doesn't seem to be any reason left to talk to you.

 

Who is paying Mann's massive legal bills from all his unsuccessful lawsuits against people who told the truth about him?

 

Here's more:

 

I went to the source... danish meteorological institute
Could not find this
But in looking found the following
Which says exactly the opposite

 

The point of my post is that you can find data to show anything you want on the Web. Why did they start at 1980 ??? Why September ???

 

Probably you can show anything you want on the web
That is why I rely upon the opinions of the consensus of scientists from around the world y

because it is a graph of minimum ice.... and that happens after the heat of summer... which happens in September.... if you are looking for the impact of warmer temps on ice... this is when you would look

and I do not know why 1980.... why does this matter?

 

High temperatures do not melt arctic ice. It is ocean currents.

Why aren’t other months used as a comparison ??

What is the trend starting in 1900 ??

 

Yeah... Ocean currents that are what? Warmer??? Why???

 

yes, but warmer oceans cause that melting..., were it not so... the oceans would have melted the ice long ago. And btw, glaciers are melting too

because they are studying the impact of warming

And btw, warming will cause the atmosphere to hold more water vapor and hence result in heavier winter snow fall

What is your point. Recent atypical warming started in about 1980

Also there was not instrumentation to do these measurements in 1900

 

They follow cycles.

 

The main accumulation of solar irradiance on the earth is in the oceans. There are multiple oceanic circulation cycles (which follow the NAO, PDO, SOI, AMO) which impact the temperatures of the water circulated into the arctic ocean. Glaciers are not receding due to melting. They recede because of reductions in the snows which feed the glaciers at their starting points. You can see this in Glacier Bay Alaska where the Grand Pacific Glacier is receding and the Margerie Glacier is flowing spectacularly into the bay within miles of each other. I witnessed this 2 years ago.

Arctic sea ice levels of today are not typical of Holocene minimums.

421 Polyak, L. et al. (2010) Quatern. Sci. Rev. 29, 1757--‐1778.

422 Opel, T. et al. (2013) Clim. Past 9, 2379--‐2389.

423 Fauria, M.M . et al. (2010) Clim. Dyn. 34, 781--‐795 (see also p. NNN)

424 The latter from Polyakov et al. (2003) J. Climate 16,2007


https://www.academia.edu/35571845/D...h_the_most_extensive_peer_reviewed_references

 

That's the beginning of the satellite era. September is the month in which the lowest monthly mean occurs. "ice-free" is defined as < 1 million sq km of sea ice. This will most likely occur in the month of September first. That's why this month is tracked closely.

FWIW...the IPCC still believes that the first occurrence of an ice-free Arctic won't occur until well after 2050 and that's down from their predictions in the 1990's of 2100. It'll be a miracle if we make it to 2050 given current trends though.

 

It's actually pretty complicated. It's a combination of air temperature, humidity, SSTs, water salinity, and winds.

The reason why 2012 still holds the minimum extent record is because of the winds. There was the Arctic Dipole Anomaly that was forcing ice through the Fram Strait where high SSTs were melting it. 2019 experience the Arctic Dipole Anomaly as well but not during the peak runup to the minimum. 2019's melt was primarily caused by incursions of high SST waters into the Arctic circle itself and the record breaking warmth well above the surface. Remember, the latent heat of fusion is a bottleneck on temperatures. That's why temperatures just above sea ice rarely exceed 32F. But the 850mb temperatures mix down and provide a lot of energy required for the phase transition without raising the temperature of the ice itself.

 

While we're on the topic the IPCC's worst prediction so far is in regards to Arctic sea ice. In AR3 from 2001 they predicted the first occurrence of annual mean < 10.5e6 sq km wouldn't occur until after 2040. It actually first occurred a mere 6 years later and 33 years ahead of forecasts in 2007.

Down in the SH they predicted that sea ice would increase until at least 2030 and possibly 2040 before declining. Record lows began occurring in 2016.

The IPCC has a long checkered history of badly underestimating sea ice declines.

 

What's the trend over the last 15 years?

 

-0.47e6 km^2/decade per NSIDC.

 

I

 

But the graph self-evidently says no such thing. The trend since 2008 -- 11 YEARS -- is INCREASING.

 

I think you meant 2007. The NSIDC trend from 2007 to 2019 for the September monthly mean is -0.01e6 km^2/decade...essentially flat or even slightly DECREASING. The math doesn't lie. That's how it is whether you perceive it correctly or not.

There are two problems going on here.

1. You have a cognitive bias that does not allow you to visually estimate linear or exponential regression trends. This is typically the result of the anchoring or framing bias and has implications on how individuals perceive noisy data.

2. You are cherry picking a specific window of time to maximize the match to your preconceived conclusion. If I were to cherry-pick an equivalent time period (actually 13 data points...not 11) then I would pick the period ending in 2012 in which the trend is an astonishing -2.0e6 km^2/decade. For the record, I'm highly opposed to cherry-picking. The trend from 1979, which the limit of the satellite era dataset, is -0.8e6 km^2/decade.

If the -0.8e6 km^2/decade trend holds firm then the Arctic will be consistently ice free in September by 2060. And given the current energy imbalance on the planet, continued emission of anthroprogenic GHGs, and continued reduction in anthroprogenic aerosols it is not unreasonable to think this tend will continue.

 

i'm buying a kayak first thing tomorrow, this could get ugly!

 

No, the trend since 2008 is increasing.

No. YOU have a cognitive bias that does not allow you even to consider that a cyclical variation could be present in data that are too restricted in time to show even one full cycle.

Right: you can't see a cyclical variation because of your anchoring and framing bias.

Nope. I picked the MOST RECENT window of time long enough to show a significant trend, but short enough to start AFTER the cyclical low, and thus show a net increase.

See? Your cognitive bias makes it impossible for you to understand the concept of a "cycle."

That trend is already decisively busted.

It not only won't continue, it already HASN'T continued.

 

Nope. I was trying to help you out. The trend from 2008 is -0.20e6 km^2/decade which is even less (by a lot) than the trend from 2007 which is only -0.01e6 km^2/decade.

See for yourself.

ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/seaice_analysis/Sea_Ice_Index_Monthly_Data_by_Year_G02135_v3.0.xlsx

Enter "=LINEST(J32:J43, A32:A43) * 10" to have Excel compute the linear regression trendline slope for you.

I do consider that there are cyclical variations, but that's not what we're discussing here. We're discussing the trend over a period of time that YOU cherry-picked. And don't think the irony of you lecturing me about restricted times when it was YOU that picked 2008 to present is lost on me.

Remember, I recommend using ALL data points available in this particular dataset and plugging them into an objective mathematical function for optimally computing the trend. That is the exact opposite of bias. You eyeballed a narrow range and did it incorrectly. That is the epitome of bias.

Ok, fine. Let's stick with 2008 to present and the -0.20e6 km^2/decade trend. Is that really what you want? Your call.

Yeah right...just like other 13 times in the dataset when the short term trend reversed to positive? Just understand that despite 13 distinct periods of an increase the trend is STILL decisively down.

And it's likely that this trend will go even more negative as the melting sea ice activates the albedo feedback leading to even more sea ice melting.

 

Those data don't match the data in the graph ARDY posted, which is what I was responding to. So you are just trying some kind of sneaky and dishonest bait-and-switch scam.

I am.

From ARDY's graph. Not your dataset. You are just trying a simple bait-and-switch scam. It's not going to work.

Look at ARDY's graph and try to tell me the trend since 2008 is not up.

I didn't eyeball it incorrectly. I identified the trend in the graph. You are trying a bait-and-switch. Disgraceful.

Look at the graph ARDY posted.

I'll take all your money on that one. Even in your dataset the decline has slowed, despite continued exponential rise in CO2 and your claimed positive feedback effect.

 

First, understand that the 2008 data point is higher than 2007. Look at the graph again. There are two anomalously low points that stand out: 2012 and 2007. I think you're getting confused because the September data points are closer to the next year on the graph than the year in which they occurred due to September being the 9th month. Do this...start from the last point in 2019 and work backwards. That should convince you. Second, a trend is not computed from the first and last data points. It is computed from ALL data points in the time range. If you pick 2007 as your start (not 2008...again, just trying to help you out) and 2019 as your end and draw a straight line then the slope of that line is indeed up. But, that slope does NOT represent the trend. The trend is computed via a linear regression of ALL data points.

I did look at the graph. I even downloaded the data (see here for OSISAF and here for NSIDC). The trend is most definitely NOT up from 2008 to 2019 nor is it up from 2007 to 2019 for either OSISAF or NSIDC. In fact, the trends are nearly identical which isn't a surprise because it is the same data. Again, I'm using a linear regression of ALL data points. I'm intentionally not cherry-picking the first and last points.0.

I stand by statements. The trend from 2008 to 2019 is DECISIVELY down (-0.20). The trend from 2007 to 2019 is flat (-0.01). Don't take my word for it. Compute it yourself.

 

Sorry, you're right: the data points that are closest to the year lines are actually for the previous Septembers. So I did mean 2007. My mistake. Thank you for your patience.

You're right, and I retract my claim.

Right.

OK, I did, I was wrong and you're right. The trend for the last 13 years is flat. So, why is 13 years not long enough to establish that a 30-year down-trend has been broken?