Arctic Sea Ice Uncertainties

The largest ice cap in the Eurasian Arctic – Austfonna in Svalbard – is 150 miles long with a thousand waterfalls in the summer.

About NOAA and MASIE Ice Extent Statistics

As we approach the serious Arctic melting season toward the September minimum, it is important to have a context to interpret various upcoming media reports.

Two factors are paramount: 1) The Sea Ice Prediction Network (SIPN) uses the September Monthly Average as reported by NOAA; and 2) This year NOAA adjusted its measurement system, resulting in a difference in extent statistics.

NOAA says this:

March 2015
The Sea Ice Index processing was updated to use the smaller SSMIS pole hole instead of the SMM/I pole hole, and the erroneous use of the SMMR pole hole in SSM/I and SSMIS data was also corrected. In addition, a new residual weather climatology mask was applied to the Northern Hemisphere that better represents where ice will and will not be, and the extent values in the daily extent data files have been rounded to three decimal places instead of six because that is the precision of the data. The entire time series was reprocessed and now reflects these changes.

As many are aware, NOAA numbers come entirely from passive microwave sensors on satellites, while MASIE ice charts are prepared by the National Ice Center based on multiple sources, including the microwave results, but also satellite imagery and field reports. More of the difference in methodologies and historical results is described here:

Something Different This Year

At the moment we are seeing that NOAA is now reporting ice extent figures that are much closer to MASIE than previously. The following table shows the comparison.

Monthly 2015 2015 2015 2014 2014 2014
February 15.032 14.498 0.534
March 15.170 14.758 0.413
April 13.650 13.954 -0.304 14.318 14.088 0.230
May 12.646 12.485 0.161 12.916 12.701 0.215
June 10.841 10.889 -0.049 11.324 11.033 0.292
July 9.573 9.473 0.100 8.482 8.108 0.374
August 6.353 6.078 0.275
September 5.364 5.220 0.144
October 7.697 7.232 0.464

All figures are in M km2. MASIE results stopped last year after October and did not resume until April 2015. The July 2015 average includes only the first 12 days, so it can not be compared to July 2014 30-day average.

Note that last year MASIE showed higher extents in all months, ranging higher by 200-500k km2, except for the September minimum. However, in 2015 NOAA changes show results much closer to MASIE, at times even larger extents. June was almost the same, something that didn’t happen in the past.


In some charts showing Arctic daily ice extent from several years, NOAA 2015 results exceed 2014 partly because of an adjusted system. The newer numbers are more in synch with MASIE results.

So far 2015 monthly averages are running slightly below last year when comparing MASIE to itself, or NOAA to itself. And SIPN median prediction is for a slightly lower minimum.

However, in the last 2 weeks 2015 is showing higher extent than the same period last year, presently an increase of ~ 500k km2.  Will that trend continue?

What will NOAA show in September? In addition to natural uncertainty, some differences may arise from system changes. At least this time, the adjustments are not in an alarming direction.

NOAA data is here:

MASIE Update July 13, 2015

2015 retains 2% lead over 2014 in BCE Region

Some Arctic ice watchers are focused on the BCE region: Beaufort, Chukchi and East Siberian Seas. It seems that when multi-year ice collects in this region, the Arctic Sea ice margin is protected, and the melting is reduced, resulting in a higher September minimum. Thus an early melting in BCE region can signal a lower summer minimum for NH ice extent, and vice-versa.

To monitor this, I have added a BCE index, being the total 2015 ice extent in BCE as a % of total 2014 extent in the same region. All figures from MASIE.

Note that the BCE maximum ice extent is comparable in size to Arctic Sea max. Historically BCE melts much more than the Arctic Sea; for example, in 2014 BCE lost 58% of its max compared to only 10% for Arctic Sea.

BCE Index recent results:

Day BCE 2015 % of 2014
187 2597170 100.2%
188 2594289 99.8%
189 2593287 99.2%
190 2538316 99.1%
191 2540197 100.6%
192 2534781 102.8%
193 2529403 102.6%

Part of the interest in BCE this year comes from the warm water blob in the N. Pacific, that may add melting to this region located on the Asian side. The two years were virtually identical with little melting prior to day 130. Daily losses since then have been similar and the 2 years were tied on day 146. For 3 days 2015 took some losses while 2014 held on to gains. Since day 150 the gap has been ~3-4%, until recently.

The Blob may have melted out Bering Sea early, and that may now be causing Chukchi to have lower extent than last year.  Yet the BCE region had more ice than 2014 for 13 days until slipping behind for 3 days, then recovering to again lead by 2%.

For more on the Blob:

July 13, 2015

Day 193, July 12 results from MASIE. Arctic ice extent lead over 2014 dips to 471k km2: A day when 2014 regains ice while 2015 has a small loss. .

2014 gained 58k on this day while 2015 lost 21k, reaching a new seasonal minimum of 9.13M km2. The loss is now 37.2% from NH max on day 93.

2014 extent now trails 2015 by 5.4%, which is about 471k km2 difference.

2015 losses were spread, the largest being 10k in Chukchi.

The seas that have lost ice are: (% lost from each sea’s max)

Baltic 100%
Bering 100%
Okhotsk 99.2%
Barents 87.8%
Baffin Bay 73.7%
Kara 71.1%
Hudson Bay 49.6%
Chukchi 39.2%
Greenland 27.8%
Laptev 19.2%
Beaufort 13.3%
Can Archipelago 12.4%
East Siberian 6.7%

The other seas have lost less than 5% from their maximums.

The seas contributing most to the total NH ice extent loss:

(5) Kara_Sea 12.0%
(6) Barents_Sea 9.6%
(8) Baffin_Bay_Gulf_of_St._L 23.4%
(10) Hudson_Bay 11.3%
(12) Bering_Sea 12.2%
(14) Sea_of_Okhotsk 11.4%

2015 melt still trails 2014 by 5 days.

masie day 193


At this point, the median outlook for NH ice extent average for September 2015 is about 5M km2, slightly below last year. That seems reasonable to me, given the lower March max, but also considering the higher ice thickness. Of course, there is no predicting what weather events will affect the ice melting and compacting between now and October.

What’s at stake this year? If September average is higher than last year, then it supports the recovery narrative. Slightly lower than 2014 (the consensus prediction) and the generally declining trend is supported. A major fall off in ice extent would be followed by mass media alarm bells.

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