Arctic ice: what is “normal”?

Although I respect the expertise of Wadhams, a puzzling thing was the insistence that the downward trend since the measurements is necessarily “abnormal” and caused by our emissions. I have no problem accepting that he saw drastic changes in the Arctic concerning the height / volume of the polar cap during his long career and that this was the reason why he thought that the Arctic would be ice free soon.

For the record, I have no problem whatsoever believing that the Arctic underwent changes in the last 40-50 years. But logically, we only start measuring the ice of the poles since the end of the 1970s. If we only started to measure AFTER the supposed cause started, then logically what is the “normal” extent and volume of the Arctic?

The measurements of the pole ice and the observations made by Wadhams all started at the end of what was then considered a cold period, in which it was reported that temperature of the Northern Hemisphere dropped by a hefty 0.6 °C since the 1940s and the sentiment was that we were heading for an imminent ice age. We could for example as well be in the downwards slope of a cycle. How could one then differentiate between an abnormal melt and a multidecadal cycle?

When I first looked at the PIOMAS data (which originally was used to predict the ice free Arctic), then I had the impression that the trend could be interpreted in more than one way. This is the graph as it shows on the website of the Polar Science Center:

I found the already added trendline rather distracting. To better compare the different views, I removed all trend information from the image and also added extra color to the line to make it stand out more. Without any trend information, this is how this graph now looks:

PIOMAS ice data without trend information

PIOMAS ice data without trend information

On the Polar Science Center website it is presented with a linear trendline. Adding that back to the previous graph is rather simple knowing the start point and the end point:

PIOMAS ice volume with linear trend

PIOMAS ice data with linear trend

The fit is not that bad, but it doesn’t really fit in the beginning, in the middle and the end. Wadhams also didn’t see it that way. He saw a exponential trend downwards. When I add this to the image this is the result (for illustrative purposes only):

PIOMAS ice data with exponential trend

PIOMAS ice data with exponential trend

This seems a better fit in the beginning and in the middle, but at the end it gave the problem that the ice volume didn’t follow the supposed trajectory after 2012. If that would be the case, the ice would have been gone by 2015, which obviously didn’t happen. Taking the last part into account gives something that resembles a cycle (for illustrative purposes only):

PIOMAS ice data with cycle trend

PIOMAS ice data with cycle trend

Of course, the future will tell what the exact trend is. Maybe it continues to go up, confirming a cycle. Or go back down in a linear way, confirming the PSC trendline. Or stay stable for a while. One thing is for sure, it is not going down exponential.

The puzzling thing to me however is why this trend is presented as if it is really clear that the decrease is abnormal and caused by our emissions. There is nothing in the data that shows this. If we want to prove that anthropogenic emissions are to blame, then we need to know what the natural base line is, which is not the case because the measurements started at a time the supposed effect already was having an influence. Therefor I have no problem believing that the Arctic changed a lot during the last 40 years, but am wary about how it is being explained without it following from the data.

2 thoughts on “Arctic ice: what is “normal”?

  1. manicbeancounter

    Thank you for taking time to look at the data. You illustrate how difficult it is to discern whether the decline in Arctic Sea Ice is
    a) an exponential trend downwards.
    b) a linear trend downwards.
    c) part of a cycle that extends beyond the data.

    This is part of the larger problem of determining between three hypotheses
    i) that ALL of the recent surface temperature warming and Arctic sea ice decline is human-caused.
    ii) that SOME of the recent surface temperature warming and Arctic sea ice decline is human-caused.
    iii) that NONE of the recent surface temperature warming and Arctic sea ice decline is human-caused.

    The first three categories do not cover the same scope as the second, but in both cases there is a huge middle ground between the extremes. The problem is that the satellite data on the extent of the sea ice has an extremely limited time period. If there is a cycle, then it extends beyond the 35-40 years of satellite data.
    But we have surface temperature data for land that goes back many decades before 1979. The issue here is two-fold. First, there were no thermometers in the Arctic Ocean, just a few weather stations on the perimeter. Second is that the recent warming in the Arctic appears to be a number of times the 0.6 °C you quote since the 1940s for the Northern Hemisphere. An example is the Berkeley Earth estimated temperature anomaly for Greenland. This shows a 2 °C rise in average temperatures in the 1920s and a similar rise after 1990. Other areas, such as Svarlbard, show a bigger rise in both periods and maybe a bigger rise in the later period than the earlier.
    Despite the temperature data is only being a loose proxy for sea ice extent, it would not seem to support the notion that all the recent drop fall in sea ice extent is predominantly human caused. It is far more suggestive that the change is largely or entirely due to some natural variation.


    1. trustyetverify Post author

      The post indeed focused mainly on the shape of the curve and not whether all, some or none of the warming is caused by human factors. The focus on this shape was because of the insistence that such a downward trend is “abnormal” and attributed to human emissions. Which is, according to me, not necessarily the case.

      The downwards motion of the graph is also consistent with a cycle, much more than the exponential downward trend that Wadhams counted on (and lost, his prediction failed). As you said, if there is a cycle it is longer than the 35-40 years of the satellite measurements. But then the measurements only started at a time the supposed effect already was having an influence and we don’t have clue how a natural cycle looks like. Therefor I think it is not possible to conclude from the observations of the Arctic in the last 40-45 years that this trend is “abnormal”. Probably Wadhams relied on other data to conclude that the changes in the Arctic were caused by our emissions.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s