TU1.L10 - Arctic Sea Ice dynamics for Global Climate Models: Results from the GlobICE Project.

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 1

Arctic Sea Ice Dynamics for Global Climate Models:Results from the GlobICE Project

Steven Baker1, Seymour Laxon2, Christian Haas3, Alan Muir1, Jeff Ridley4, Andy Ridout2, Daniel Feltham2, Daniela Flocco2,

Ron Kwok5.

1UCL-MSSL, 2UCL-CPOM, 3Uni.Alberta, 4UK-MetO, 5JPL.

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 2

Talk Overview

• What is the GlobICE project?

• What are the Products?

• How are they made?

• How are they validated?

• Examples of results

• How to get the data

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 3

GlobIce Origins

• User requirements for Sea Ice Dynamics Observations

• Response to specific user needs

• Scientific users- climate & process modellers etc...

• Global/Regional coverage for motion, deformation, mass fluxes,...

• Exploitation of Historical EO Data Archives

• Long-term provision of data to users

ESARequirements

DUE ITT

“SeaIce Community”

RegionalModellers

ProcessStudies

DataAssimilation

ClimateModellers

SeaIce Dynamics ProductsMotion, Deformation, Flux

ERSSAR

EnvisatASAR

Data Archives

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 4

CliC Project Areas Supported by GlobIce

• CPA3: The marine cryosphere and its interactions with high latitude oceans and atmosphere– What will be the nature of changes in sea-ice distribution and mass

balance in both polar regions in response to climate change and variability?– Outputs to include: re-analyses of oceanographic and sea-ice

quantities; Reduced uncertainties regarding past variability and trends in sea-ice coverage...

• CPA4: Links between the cryosphere and global climate– Impact on atmospheric and oceanic circulation of changes in cryosphere?

Likelihood of abrupt climate changes resulting from changes in the cryosphere?

– Outputs to include: Improved models of high latitude climate; Understanding of the cryospheric processes w.r.t climate; Clearer view of role of cryosphere in global water & energy cycles...

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 5

Main Objectives

Parameter Spatial Resolution Temporal resolution Accuracy Ice Deformation

5km 3 days < 1%

Ice Motion 5km Daily 0.1 km/day Open Water Fraction

25km Daily 0.1%

Thickness 100km 1 month 0.1m Mass Flux n/a 1 month 25km3

Product requirements - specified by CliC (ideal)

Produce a historical Arctic-wide data set of sea ice dynamics to scientific user community, primarily in climate research.

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 6

GlobICE Level - 1

• Ice Motion from cross-correlation of pairs of SAR images

• Eulerian motion: motion at fixed grid points every few days

• Lagrangian motion: follow a point starting on a fixed grid for as long as possible over a season– add new points as tracking of old ones end

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 7

GlobICE Level - 2

• Eulerian Motion + external ice concentration --> ice area flux estimates- Ice concentration from Nimbus-7 SMMR and DMSP SSM/I Passive

Microwave Data, Cavalieri et al.

• Lagrangian motion and cell connectivity database allows derivation of deformation metrics– deformation rate – Vorticity rate – Shear rate– Open Water Fraction (OWF)

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 8

GlobICE Level - 3

Ice Deformation - Lagrangian Motion

• Motion of cell vertices -->Computation of deformation metrics

• Ice thickness estimation – Converging -> rafting– Divergence -> new ice growth

• Open Water Fraction estimation

Animation example from RGPS (Ron Kwok)

(for Mac)

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 9

GlobICE Level - 3

Ice Deformation - Lagrangian Motion

• Motion of cell vertices -->Computation of deformation metrics

• Ice thickness estimation – Converging -> rafting– Divergence -> new ice growth

• Open Water Fraction estimation

Animation example from RGPS (Ron Kwok)

(for PC ?)

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 10

Gateway ID and Name Location

0 - The Fram Strait 17W to 10E at 80N (100 km)

1 - Svalbard – Franz Josef Land 28E to 45E at 80N (100 km)

2 - Franz Josef Land – Severnaya Zemlya 66E to 91E at 81N (100 km)

3 - Franz Josef Land – Novaya Zemlya 76N to 80N at 60E (100 km)

4 - The Nares Strait 78W to 71W at 78N (200 km)

5 - The Bering Strait 172W to 166W at 65.7N (150 km)

6 - The Davis Strait 67W to 55W at 70N (100 km)

GlobICE Level - 4• Ice thickness is derived from

Altimetry data at UCL (Laxon et al 2003)

• combined with concentration & ice motion -> volume flux through Gateways(Kwok et al 2006)

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 11

User Access To Products - via Live Access Server

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 12

GlobICE Full Product SetProduct Grid

ResolutionProduct Frequency

Product TimeSpan

ProcessingChain

Ice Motion – Image Pair Displacements

5km Daily 2-9 days Eulerian

Average Velocity 5km Daily 3 days Eulerian

50km Monthly Monthly Eulerian

Lagrangian Motion 5km Monthly Monthly Lagrangian

Deformation, (Thickness,) Open Water Fraction

5km Daily 3 days Eulerian

20km Monthly Monthly Eulerian

100km Monthly Monthly Eulerian

Open Water Fraction 20km Monthly Monthly Eulerian

Gateway Mass Flux per gate Monthly Monthly Eulerian

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 13

Product Formats

• All Gridded products are on a polar stereographic projection - same as RGPS and SSMI– Standard parallel 70deg N

• Net cdf 3 (classic) – self-describing file format

• Product documentation available

• Measurement confidence data included

• Available via LAS & ftp

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 14

5km

t=3 days

Eulerian Image Pair Products

• each product contains displacements from 1 pair of images at 5km grid resolution

• each product spans between 2 and 9 days.

Atypical (sparse) example used for clarity!

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 15

Ice Motion from image pair and Sedna GPS Buoy drift (in white)

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 16

Eulerian 3-day Average Velocity Products

• 5 km Grid

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 17

Eulerian Average Monthly Velocity

• 50 km Grid

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 18

IAPB drift buoys compared with Monthly Average Velocity

Black dot at end of period, 8 Oct-17Nov 08

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 19

Lagrangian Motion compared to SEDNA GPS Drift Buoys

SEDNA GPS Buoy 74353 compared with GlobICE Lagrangian tracking data between 12/10/07 and 15/11/07. Motion is from top to bottom as time increases

Sedna: 12 GPS Buoys - data courtesy Jenny Hutchins

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 20

Lagrangian MotionTrack Length - Jan 2009

Percentage Track Lengths<= 3 days : 7.3%4 to 7 days : 62.7%8 to 14 days: 20.0%15-21 days: 6.5%21-31 days : 3.5%

ASAR Coverage – Jan 2009

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 21

Deformation, Thickness, OWF Products

• Lagrangian derived DTO product (Deformation, Thickness, OWF)replaced by a Eulerian derived D(T)O product with identical format.

• Due to ASAR coverage, Lagrangian derived DTO product would give very sparse coverage due to rapid cell decay.

• Eulerian derived DO product provides good spatial coverage.• Core Users consulted & recommended this change.• Lagrangian Motion Product still produced at Monthly intervals.• Lagrangian derived DTO product could be added in the future in the

event of improved coverage from ASAR or other instrument.

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 22

Open Water Fraction

Open Water Fraction from GlobICE (left panel) and SSM/I (right) for November 2008. The scale is % of open water

In areas of new ice passive microwave is known to over estimate OWF

e.g. Kwok has shown that SSMI gives a higher OWF than RGPS

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 23

Month GlobIce Fram Strait

(km3/month)

GCM Fram Strait

(km3/month)Nov 154 145±44Dec 135 171±35Jan - 182±45Feb 173 194±54Mar 142 232±83

• Ice volume fluxes in km3 / month • 7 gateways from the Arctic• Helps to understand fresh water export budget

Gateway Volume Fluxes

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 24

Current Status

• GlobICE System is fully developed and tested

• Prototype Products Validation by core user group in progress

• First User workshop - 30 May at IGS 2010 in Tromsø

• Final testing and Delivery was completed at UCL in early July

Next:

• Phase 3 - distribution of products to users

• Evaluation by wider community

• Integration of data within models

• 2nd User Workshop - Feedback!

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 25

GlobICE Full Product Set (recap)Product Grid

ResolutionProduct Frequency

Product TimeSpan

ProcessingChain

Ice Motion – Image Pair Displacements

5km Daily 2-9 days Eulerian

Average Velocity 5km Daily 3 days Eulerian

50km Monthly Monthly Eulerian

Lagrangian Motion 5km Monthly Monthly Lagrangian

Deformation, (Thickness,) Open Water Fraction

5km Daily 3 days Eulerian

20km Monthly Monthly Eulerian

100km Monthly Monthly Eulerian

Open Water Fraction 20km Monthly Monthly Eulerian

Gateway Mass Flux per gate Monthly Monthly Eulerian

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 26

Access to the GlobICE Products?

If you want to have access to these data products to evaluate them for yourselves, please contact one of the following:-

• Steve Baker / Seymour Laxon / Jeff Ridley

sgb@mssl.ucl.ac.uk

swl@cpom.ucl.ac.uk

jeff.ridley@metoffice.gov.uk

• ESA Technical Officer for GlobICE : Pascal Lecomte

http://globice.mssl.ucl.ac.uk/

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 27

Thanks For Listening

Ron Kwok

Steve BakerSeymour Laxon

Tove TennvassåsOle Søslett

Jeff Ridley

Fanny Ardhuin(Robert Ezraty)

Rüdiger Gerdes

Phil Eales

Dan FelthamAndy Ridout

http://globice.mssl.ucl.ac.uk/

Alan Muir

Christian Haas

GlobICE Project

IGARSS : 25-30 July 2010 : Honoluluhttp://dup.esrin.esa.int 28

OWF In Dense Pack-ice

HadGEM1

HadISST

GCM diagnostics: ice fraction

• Relationship between OWF in GlobIce and models