Sea Ice Products
An automatic analysis system, which derives sea ice products by combining satellite
data from different sensors, has been developed and implemented for operational use.
Four pre-operational products are defined:
- Ice concentration (%) (Global)
- Ice edge: closed ice/open ice/open water (Global)
- Ice type: multi-year/first-year (Global)
- Low Resolution Sea Ice Drift (Global)
These are defined as pre-operational because they switched to SSMIS on 19-01-2013 and
are still under final validation.
One re-processed dataset is available:
Finally, some demonstrationl products are available. Those products require
more scientific and/or technical validation before they can be declared operational.
- Medium Resolution Sea Ice Drift (NH)
Operational Sea Ice Products
|
| PRE-OPERATIONAL |
Global Sea Ice Concentration |
|
|
Title | OSI-401: SSMIS Sea Ice Concentration Maps on 10 km Polar Stereographic Grid |
| Description | The sea ice concentration product from the EUMETSAT OSI SAF. Ice concentration is computed
from atmospherically corrected SSMIS brightness temperatures, using a combination of state-of-the-art
algorithms. It is operational since 2005. |
| Data used | SSMIS (DMSP F17), ECMWF forecast for atmospheric correction (SSM/I before 19-01-2013) |
| Formats | HDF5, NetCDF, GRIB |
| NRT access | FTP (last 31 days in NetCDF3, HDF5 and GRIB) and EUMETCast (GRIB). |
| Archive | FTP (since 2005) and
EUMETSAT Data Center |
| Documentation | Product User's Manual |
| Links | Quicklooks NH, Quicklooks SH,
Validation & monitoring |
| Remarks | N/A |
|
|
| PRE-OPERATIONAL |
Global Sea Ice Edges |
|
Title | OSI-402: Sea Ice Edge Maps on 10 km Polar Stereographic Grid |
| Description | The sea ice edge product from the EUMETSAT OSI SAF. Ice classes are assigned
from atmospherically corrected SSMIS brightness temperatures and ASCAT backscatter values, using a Bayesian approach.
It is operational since 2005. |
| Data used | SSMIS (DMSP F17), ASCAT (Metop-A), ECMWF forecast for atmospheric correction (SSM/I before 19-01-2013) |
| Formats | HDF5, NetCDF, GRIB |
| NRT access | FTP (last 31 days in NetCDF3, HDF5 and GRIB) and EUMETCast (GRIB). |
| Archive | FTP (since 2005) and
EUMETSAT Data Center |
| Documentation | Product User's Manual |
| Links | Quicklooks NH, Quicklooks SH,
Validation & monitoring |
| Remarks | N/A |
|
|
| PRE-OPERATIONAL |
Global Sea Ice Types |
|
Title | OSI-403: Sea Ice Type Maps on 10 km Polar Stereographic Grid |
| Description | The sea ice type product from the EUMETSAT OSI SAF. Ice classes are assigned
from atmospherically corrected SSMIS brightness temperatures and ASCAT backscatter values, using a Bayesian approach.
It is operational since 2005. |
| Data used | SSMIS (DMSP F17), ASCAT (Metop-A), ECMWF forecast for atmospheric correction (SSM/I before 19-01-2013) |
| Formats | HDF5, NetCDF, GRIB |
| NRT access | FTP (last 31 days in NetCDF3, HDF5 and GRIB) and EUMETCast (GRIB). |
| Archive | FTP (since 2005) and
EUMETSAT Data Center |
| Documentation | Product User's Manual |
| Links | Quicklooks NH, Quicklooks SH,
Validation & monitoring |
| Remarks | Note that during Arctic summer season (May-September) the ice type product is
dubious because melting of the ice surface obscures the ice type signals. |
|
|
| PREOPERATIONAL |
Low Resolution Sea Ice Drift |
 |
Zoom from map above |
|
Title | OSI-405: Sea Ice Motion Maps with 48 hours span, on 62.5 km Polar Stereographic Grid |
| Description | The low resolution sea ice drift product from the EUMETSAT OSI SAF. Ice motion vectors
with a time span of 48 hours are estimated by an advanced cross-correlation method
(the Continuous MCC, CMCC) on pairs of satellite images. Several single-sensor products are available, along with a
merged (multi-sensor) dataset.
|
| Data used | SSMIS (85 GHz, DMSP F17), ASCAT (Metop-A) |
| Formats | NetCDF3 |
| NRT access | FTP (last 31 days in NetCDF) |
| Archive | FTP (since December 2009),
HTTP and OPENDAP (from October 2006 to December 2009) |
| Documentation | Product User's Manual,
Validation Report |
| Links | Quicklooks NH, Validation and Monitoring |
| Remarks |
- Due to high atmospheric Liquid Water Content and to ice surface melting, it is not possible to track ice motion
during Arctic summer, from the channels we are processing. Accordingly, the NH product files distributed between
May, 1st and September 30th do not contain any valid vector.
- The product would have to be extended to Southern Hemisphere before it can be declared as operational.
|
|
Other Sea Ice Products
|
| REPROC DATASET |
Global Sea Ice Concentration reprocessing dataset (1978-2009) |
|
|
Title | OSI-409: Global Sea Ice Concentration reprocessing dataset (1978-2009) |
| Description | The reprocessed sea ice concentration dataset of the EUMETSAT OSI SAF,
covering the period from October 1978 to October 2009 (SMMR and SSM/I).
Ice concentration is computed from atmospherically corrected SSM/I brightness
temperatures, using a combination of state-of-the-art
algorithms and dynamic tie-points. The dataset includes error-bars for each
grid cell (uncertainties).
Version 1 of the dataset was released early 2010. Version 1.1
of the dataset, with longer temporal coverage
was released late 2011.
|
| Data used | SMMR and SSM/I, ECMWF re-analysis and forecast for atmospheric correction |
| Formats | NetCDF |
| Access Data | FTP (whole dataset) |
| Documentation | Product User's Manual,
Validation report |
| Links | Quicklooks (single day),
Quicklooks (compare 2 days) |
| Remarks |
WARNING: There is a bug in the uncertainty estimates in v1 of the data set.
This bug has no effect on the maps of sea ice concentration. A new improved v1.1 dataset
was made available in November 2011, with more accurate uncertainty estimates.
Use of this dataset should be acknowledged by referring to:
EUMETSAT Ocean and Sea Ice Satelitte Application Facility. Global sea ice concentration reprocessing dataset 1978-2009
(v1.1, 2011), [Online]. Norwegian and Danish Meteorological Institutes. Available from http://osisaf.met.no. |
|
|
| IN DEVELOPMENT |
Medium Resolution Sea Ice Drift |
 |
Zoom from map above |
|
Title | OSI-407: Sea Ice Motion Maps with 24 hours span, on 20.0 km Polar Stereographic Grid |
| Description | The Medium resolution sea ice drift product from the EUMETSAT OSI SAF. Ice motion vectors
with a time span of 24 hours are estimated by a maximum cross-correlation method (MCC) on pairs of satellite images.
Input data are ~1km Infra Red or Visible data from the AVHRR instrument on board the METOP satellite.
|
| Data used | AVHRR channel 2 and channel 4 (Metop-A VIS and TIR respectively) |
| Formats | NetCDF |
| NRT access | FTP (last 31 days in NetCDF, coming soon) |
| Archive | FTP (coming soon) |
| Documentation | Product User's Manual,
Validation Report |
| Links | Quicklooks NH (coming soon) |
| Remarks |
- Due to pronounced cloud cover during summer, and occasional during winter periods, the daily data coverage can be very sparse.
- The mr ice drift production is based on TIR data during winter and VIS data during summer, with an overlapping period in spring and autumn.
|
|
The concentration product is currently based on passive microwave
data from SSMIS. A sea ice concentration algorithm has been established
as a smooth combination of two algorithms, Bristol algorithm and
the Bootstrap frequency mode algorithm. To ensure an optimum performance
over both marginal and consolidated ice, and to retain the virtues of
each algorithm, the Bristol algorithm is given very little weight at low
concentrations, while the opposite is the case over high ice concentrations.
The algorithm is run on brightness temperatures corrected using the Numerical
Weather Prediction (NWP) model based correction scheme and using the SAF monthly
tiepoint set.
Atmospheric corrections
Contamination arising from atmospheric water vapour
content and wind roughening of the open water surface is a problem. Using a
radiative transfer model and input data from the ECMWF (European Center for
Medium Range Weather Forecast) NWP-model weather an atmospheric correction is
computed and applied to the SSMIS brightness temperatures. The corrected
brightness temperatures are used in the ice concentration calculations as well
as in the multi sensor ice analysis.
A tool for combining various data sources containing
uncertain information is given by the Bayesian approach. By this, several
measured parameters can be combined to yield an optimal estimate taking
into account the uncertainties involved in the various data sources. We
obtain an estimate of the most probable ice class, and also of the
uncertainty of this estimate.
A multi sensor analysis system using several sensors to analyse sea ice
classes has been built. This system has been developed on SSM/I, scatterometer and
AVHRR data. At present, SSMIS and ASCAT data is used operationally. The analysis system is
implemented for ice edge analysis defining three classes: open
water, open ice and closed ice. The basic step in building the analysis
system was to derive the probability distributions for each ice class and
given the measured quantity. To obtain this statistical knowledge of the
average and the scatter of the expected measurement values, large sets
have been collocated with satellite data from areas with know ice properties.
In a similar way an algorithm for sea ice type analysis,
distinguishing between first year and multi year ice has been developed
and implemented. The operational product uses SSMIS and ASCAT data.
Contact
Contact person for OSI SAF Sea Ice products is Steinar Eastwood, met.no,
osisaf-manager[at]met.no.
|