Much importance has been ascribed to tracking the change in Arctic sea ice, but what about the global trend? That doesn’t seem to get much press. However there is some important information that needs to be presented related to the global trend of sea ice as measured by satellite since 1979. The results are surprising. – Anthony
A guest post by Jeff Id, from The Air Vent
2nd Update 12/24/08 It turns out that an error in documentation at NSIDC is the cause, see this new post for a full explanation.
Update and correction:
To my readers, Anthony Watts received a comment from our friend Tamino on the ice data I used for the area analysis. Unfortunately for me he is right this time. It appears that a correction to the data is required prior to 1987 which will create an approximate negative trend of 0.88 million sq kilometers per 30 years. It is a fairly small trend in the scheme of a 20million sq kilometer signal, but understand this mistake is entirely mine and is unrelated to Anthony Watts excellent blog.
Unfortunately the change makes the Area signal difficult to determine prior to 1988 because the percent fill is unknown. Anthony cannot check every detail of a post which took me days of research and he simply requested if he could copy it here.
The link to my corrections is:
I calculated a true global sea ice anomaly in this post using the National Snow Ice Data Center data. What would you say if I told you that over the last 30 years the sea ice area has stayed flat or even trended — Up!!!???
This isn’t a small deal. We have been told, well screamed at really, that CO2 is causing unprecedented rise of temperature on a global scale. We hear constantly that the ice is melting and the result will be dramatic flooding of the earth; movies have even been made. Those of us who pay attention to the scientists have heard that the most significant warming will be at the poles (according to the computer models). We also hear that the Antarctic has added ice during the same time the Arctic lost sea ice. This is explained in that the Antarctic ice increase is a local situation and the Arctic ice loss is a result of global warming. A unique form of cherry picking but should be treated with an open mind.
If you’ve been paying attention, you have heard that the net ice level is going down. The Antarctic gain cannot compensate for the Arctic loss. Well, I set out to see how bad the situation is.
First, anthropogenic global warming scientists use two measures, extent and area.
Extent looks at all the square Kilometers (Km^2) with more than 15% ice in them and adds them up.
Area looks at all the square Kilometers (Km^2) with more than 15% ice in them and adds them up but multiplies the Km^2 by the amount of ice in the square kilometer. i.e -(if you have 1 Km^2 of sea filled 15%, ice- extent counts it as 1Km^2 while area counts it as 15% of 1Km^2 or 0.15Km^2)
This post deals with the amount of sea ice so I used Area. In the future Iwill do it with extent. The NSIDC uses two algorithms for calculation of sea ice, nasateam and bootstrap. We will look at both here.
Without modification the NSIDC data for bootstrap runs from 1978-Dec. 2006 and the nasateam runs from 1978-Dec 2007, these near 30 year trends comply close enough with current science which states (conveniently) that climate requires a 30 year trend to see the result.
This is a graph of the global sea ice area from the nasateam algorithm.
The red line is the slope of the global sea ice data from nasateam in its raw format. The slope is negative by only 6803 Km^2/year and the mean is 18,290,000 km^2.
We should look at sea ice anomaly to be the most accurate for trend. To calculate sea ice anomaly I took the average shape of the annual signal and subtracted it from the curve above.
The average ice variation globally looks like this on an annual basis.
I subtracted this curve above to get the sea ice anomaly.
The downward slope of this graph is more extreme but the scale is highly magnified. The net downslope in 30 years of global warming is – 10173Km^2/year. Over 29 years of data this means that we have lost 302025 Km^2 of ice. This is a 1.65 percent drop in global ice level in 30 years. Remember though that this data ended on an extraordinary high melt year of 2007, the ice level can be seen recovering in dec 07 leading into 2008. This shows as a slight change in slope of the very tip of the first graph (a subtle, difficult to see effect).
Well NSIDC recommends using the Bootstrap algorithm for research instead of Nasateam because of certain errors which have been corrected for.
The bootstrap algorithm plot for global data looks like this.
The red line is slope again, and this time it is positive, indicating an increase in ice level from 1978-Dec 2006. The slope of the red line is plus 6341 km^2 per year indicating that the earth in 28 years has added 177,000 sq kilometers of ice with a mean ice level of 20.42 million Km^2.
The anomaly is better for calculating trends because it cleans up the end points making the slope insensitive to the start and stop point of the annual cycle.
The up trend for the anomaly in sea ice from 1978 to end 2006 is 804Km^2 per year. Which in our timeframe the preferred bootstrap algorithm says the earth ADDED 22,000 Km^2 of ice area!!
Here are the anomalies rescaled to actual by adding the mean of the original data back in.
Obviously people cannot make the claim that sea ice is being lost. It isn’t. The data shows that our trend is basically flat during this time of unprecedented temperatures. It’s clear that there has been no significant change in sea ice area.
This is almost enough to make me turn in my Skeptic union card, but increased CO2 warming the earth makes some sense to me, the magnitude is in question. The fact that polar sea ice not melting is not an insignificant point. It is also important to realize that the changes are too small to fit with IPCC statements about the trend. Unlike trees, ice does make a good thermometer. I can’t say this strongly enough— This is a strong indication of substantial errors in the computer models and temperature data which needs to be addressed before we throw what’s left of our global economy to the wind. How would Earth’s total sea ice ignore such substantial warming? It’s a good question which deserves an answer.
I will update this when new data becomes available and will also attempt to demonstrate that the net slopes we see are within the margin of error for the measurement in a future post. In the meantime, lets let the world know the truth. We aren’t going to drown any time soon!
I had a request for description of the difference between the bootstrap and nasateam algorithms. It is a bit complex but it seems well documented on the NSIDC here are a few links and descriptions from that site. From FAQ section.
2. What is the difference between the NASA Team algorithm and the Bootstrap algorithm?
For general analyses or creation of simple images, either algorithm will suffice. The Bootstrap sea ice concentration data set is believed to be more useful for modeling and process studies in the polar regions because it is generally free of residual errors that could not be removed by conventional techniques. A temporally more consistent time series of sea ice concentrations is provided, offering improved accuracy over the ice concentration maps created from the original Bootstrap algorithm.
More interesting to me was the table provided which shows the strenghts and weaknesses of each process. The original table is at the link above.
HERE is a link to the R code to make the above graphs.