Blog Archive

Monday, April 21, 2008

National Snow and Ice Data Center -- Arctic sea ice melt extent graph discrepancies from April 2008


BLOGGER'S update, April 23rd (evening):

Hank Roberts was kind enough to post the explanation (although, looking at this evening's graph, I must admit, after viewing the 4 graphs, that I still have no idea what is going on with them -- only time will tell, it seems):

"The satellite data sources for these products, while generally providing complete coverage, are subject to gaps (shown in dark grey) in coverage because of satellite operations. In the daily extent time series, gaps are replaced with values interpolated from surrounding days, but temporary spurious results may occur. The current satellite source is aging and showing more frequent data gaps. NSIDC is investigating a reliable replacement data source. —Credit: National Snow and Ice Data Center"

Blogger's note -- update of April 23 (morning): This just gets weirder and weirder. April 20 shows one thing, April 21 another, and now April 23 looks more like April 20 again. Anyway, I guess the NSIDC are the only ones who can tell us how they update this graph.



ABOVE: Graph from April 20, 2008.

BELOW: Graph from April 21, 2008.




BELOW: Graph from the morning of April 23rd, 2008.
(Once again, thank you to Leon for pointing it out to me.)



BELOW: Graph from the evening of April 23rd, 2008.


BELOW: Graph from the evening of April 25th, 2008.

Link to updated graph: http://nsidc.org/data/seaice_index/images/daily_images/N_timeseries.png

13 comments:

TimC said...

So my eyes weren't playing tricks on me! Thank you for saving this.

TimC said...

PS

The image was the earlier of the two about 5:30 AM PST -- but by the time I got to work it was the latter (7:30 AM).

Cassandra_Moderna said...

OK, this explanation makes sense to me (with a caveat) -- Phil. Fenton posted this comment on realclimate:

"I noticed the sudden downturn on the graph yesterday and also that it coincided with images that were missing signal in a couple of regions. I anticipated that the graph would shortly be adjusted and as you have shown they were."

Ok, that sounds reasonable, but the portion of the curve affected covered much more than a single day.

Cassandra_Moderna said...

Also, Nick Barnes posted this comment at realclimate:

"Yes, I noticed the NSIDC graph change too. I assume that the earlier graph, which was extreme, was a data processing error. The current graph is still plenty alarming.

These things happen at the cutting edge. We had similarly alarming sea ice area graphs from Cryosphere Today for a short while earlier in the year, before they found and fixed a bug."

leon said...

Thank you a lot for this Tenney!

leon said...

I hope this: http://nsidc.org/data/seaice_index/images/daily_images/N_timeseries.png is also thanks to data processing error!!

Cassandra_Moderna said...

Thanks again, Leon. I am going to update this post with the new graph of April 23rd. That wet noodle sure is going all over the place.

Hank Roberts said...

Contact info for the person (one person, as far as I know) who maintains the site is on the page.

You should ask.

I know the site had a hardware crash a while back and took quite a while before the charting worked reliably again; I nudged in email then about putting up a notice and it was done.

I know the averages for the latter quarters and year 2007 haven't been updated (I asked, that's not done automagically, it's hand work and takes time).

Ask. Be patient. It's not a huge operation, it's another human being over there.

Hank Roberts said...

Oh, I see the question's answered over there already:
____________________________
The satellite data sources for these products, while generally providing complete coverage, are subject to gaps (shown in dark grey) in coverage because of satellite operations. In the daily extent timeseries, gaps are replaced with values interpolated from surrounding days, but temporary spurious results may occur. The current satellite source is aging and showing more frequent data gaps. NSIDC is investigating a reliable replacement data source. —Credit: National Snow and Ice Data Center

Paolo Morelli said...

Also the sea ice extent data pubblished on IARC-JAXA web site
(www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm) is subject to small fluctuations during the day. But they are way smaller the the changes shown here in these graphs from NSIDC.
By the way, IARC-JAXA graph is updated on a dayly basis and it is my personal favourite website to track sea ice extent.

Cassandra_Moderna said...

Thank you so much, Paolo!

I have added that link to the expanded post on this subject:

http://climatechangepsychology.blogspot.com/2008/06/arctic-sea-ice-well-on-its-way-to.html

and I will add it to this post, as well:

http://climatechangepsychology.blogspot.com/2008/06/june-2008-arctic-sea-ice-in-bad-shape.html

Paolo Morelli said...

You're wellcome. I'am very glad you found it usefull.
According to this dataset today the ice extend was below the (same day) 2007 value for the first time since January. And ice loss rate in the first part of June was pretty bad, avaraging almost 90k square km per day.
ciao

Cassandra_Moderna said...

Dear Paolo,

The multi-year ice is already gone, so the melt rate is going to be very bad from here on out.