HELP! Are Ob Gulf and Yenisey Gulf frozen in September?

[xuhua2005]

According to the ice charts from NSIDC, it seems that the sea ice in Ob Gulf and Yenisey Gulf still remains even in September. But this area is quite south!
Is this true?
 

[ivica]

According to the ice charts from NSIDC, it seems that the sea ice in Ob Gulf and Yenisey Gulf still remains even in September. But this area is quite south!
Is this true?
 

Seems very unlikely to me. If I recall it correctly, that area had on occasions temp well above 20 C this summer.

Take a peek at satellite photos, using for example 2013-08-09 and NASA worldview
http://earthdata.nasa.gov/labs/worldview/
or http://www.arctic.io/observations/

[werther]

This is presented on their map avery year. As I posted last year, there's no ice in Baidaratskaya Gulf, Khatanga Bay, Ob Estuary a.s.o. On the Canadian side, even Coronation Gulf filled with ice seems overdone. Check it on MODIS, like Disko Bay etc.

[BenB]

I asked the NSIDC about this, and they put it down to land-to-ocean spillover. Personally I find it strange that the ocean mask for August didn't eliminate this error, but I'm sure they have their reasons. Apparently "the Sea Ice Index data set and the source data for it do tend to exhibit these errors even after corrections have been made", which suggests that this ice is included when the Sea Ice Index is calculated.

Ben

ps. Thanks for the great website and forum Neven!

[Wipneus]

About the land spillover.

The NSIDC maps calculates ice concentration using microwave radiation intensities at 19 and 37GHz. These are measured by the SSMIS instruments on DMSP satellites. Footprints are:

Channel    FOV
19.3 GHz    70x45 km
37.0 GHz    38x30 km
85.5 GHz    16x14 km
91.7 GHz    16x13 km

http://nsidc.org/data/docs/daac/nsidc0001_ssmi_tbs.gd.html

Narrow channels are a big problem, and realistically these measurements methods are unsuitable to measure ice concentration there.

The AMSR-E and AMSR2 instruments are superior in this regard. Algorithms that use the ~90 GHz channels are advantageous in this respect as well.

[xuhua2005]

Thank Wipneus.
But I find in high-resolution charts Ob Gulf and Yenisey Gulf are also frozen in September.
This is the website providing charts from the sensor of SSMIS.

http://www.iup.uni-bremen.de:8084/ssmis/

[Wipneus]

Thank Wipneus.
But I find in high-resolution charts Ob Gulf and Yenisey Gulf are also frozen in September.
This is the website providing charts from the sensor of SSMIS.

http://www.iup.uni-bremen.de:8084/ssmis/

That website is unavailable at the moment, I use Uni Hamburg's instead.

Attached are images of
NSIDC based on SSMIS 19/37 GHz
Hamburg based on SSMIS 85GHz
Hamburg Based on AMSR2 89GHz

Pick the one that has "high resolution".

[xuhua2005]

Thank Wipneus.
But I find in high-resolution charts Ob Gulf and Yenisey Gulf are also frozen in September.
This is the website providing charts from the sensor of SSMIS.

http://www.iup.uni-bremen.de:8084/ssmis/

That website is unavailable at the moment, I use Uni Hamburg's instead.

Attached are images of
NSIDC based on SSMIS 19/37 GHz
Hamburg based on SSMIS 85GHz
Hamburg Based on AMSR2 89GHz

Pick the one that has "high resolution".

Thank you.
Resolution is the cause of the problem.
That's OK.

[nukefix]

How about ASCAT, 25km experimental resolution:

http://www.esa.int/esapub/bulletin/bullet102/Gelsthorpe102.pdf

[Wipneus]

This is the website providing charts from the sensor of SSMIS.

http://www.iup.uni-bremen.de:8084/ssmis/

The web site is available now. Attached the same region by their SSMIS +85 GHz map.

As you see, most of the discussed area is grayed out, my interpretation: not available/computable.

It is not 100% enough: even outside the grey bands some spurious low concentration ice is visible. That is antennas: outside the specified field of view some land contamination is possible: side lobes.

It is a neater solution than NSIDC's IMO.

[xuhua2005]

This is the website providing charts from the sensor of SSMIS.

http://www.iup.uni-bremen.de:8084/ssmis/

The web site is available now. Attached the same region by their SSMIS +85 GHz map.

As you see, most of the discussed area is grayed out, my interpretation: not available/computable.

It is not 100% enough: even outside the grey bands some spurious low concentration ice is visible. That is antennas: outside the specified field of view some land contamination is possible: side lobes.

It is a neater solution than NSIDC's IMO.

Thanks. I got it.
It is difficult to treat the edge of landmass.

[xuhua2005]

Thank Wipneus.
But I find in high-resolution charts Ob Gulf and Yenisey Gulf are also frozen in September.
This is the website providing charts from the sensor of SSMIS.

http://www.iup.uni-bremen.de:8084/ssmis/

That website is unavailable at the moment, I use Uni Hamburg's instead.

Attached are images of
NSIDC based on SSMIS 19/37 GHz
Hamburg based on SSMIS 85GHz
Hamburg Based on AMSR2 89GHz

Pick the one that has "high resolution".

Excuse me, where did you download the last chart (Hamburg Based on AMSR2 89GHz)? Could you give me the link? It's quite useful!
Thanks.

[Wipneus]

Attached are images of
NSIDC based on SSMIS 19/37 GHz
Hamburg based on SSMIS 85GHz
Hamburg Based on AMSR2 89GHz

Pick the one that has "high resolution".

Excuse me, where did you download the last chart (Hamburg Based on AMSR2 89GHz)? Could you give me the link? It's quite useful!
Thanks.

xuhua,

I created these three images from gridded ice concentration data, linear from 0%->black and 100%->white.
The last month of these (Hamburg amsr2)  images, at least during the melting season, is available on https://sites.google.com/site/apamsr2/home/pngcby32

Similar images (and the gridded data files that I use) are on Uni Hamburgs server:

ftp://ftp-projects.zmaw.de/seaice/AMSR2/

Oh and let me repeat the caveat from the Uni Hamburg scientists:

These data are not officially released and are still work-in-progress.

[xuhua2005]

Many thanks!