Imposed versus Dynamically Modeled Sea Ice A ROMS

  • Slides: 32
Download presentation
Imposed versus Dynamically Modeled Sea Ice: A ROMS study of the effects on polynyas

Imposed versus Dynamically Modeled Sea Ice: A ROMS study of the effects on polynyas and waters masses in the Ross Sea John M. Klinck, Y. Sinan Hüsrevoglu and Michael S. Dinniman Center for Coastal Physical Oceanography Old Dominion University

Introduction Sea ice is important for ocean surface fluxes and water transformations at high

Introduction Sea ice is important for ocean surface fluxes and water transformations at high latitude. Imposed sea ice is easier and has some accuracy. Atmosphere, ocean and ice can be inconsistent. Can not represent dynamics of polynyas or variation from observed conditions.

Questions What is the sensitivity of dynamic sea ice to forcing and OBC? How

Questions What is the sensitivity of dynamic sea ice to forcing and OBC? How realistic are sea ice results for the Ross Sea compared to observations? What is the influence of Terra Nova Bay Polynya on High Salinity Shelf Water?

Model Overview • ROMS v 2. 1 primitive equation model • 5 km grid

Model Overview • ROMS v 2. 1 primitive equation model • 5 km grid spacing • Includes silicate and nitrate with simplified uptake • Includes water under Ross Ice Shelf • Daily wind stress for years 2000 and 2001 • Surface fluxes from bulk formulae and surface atmospheric conditions • Imposed and dynamic (CICE) sea ice • Additional information by Mike Dinniman.

Sea Ice Model • CICE V 3. 1 (Hunke and Dukowicz, 1997: 2002) •

Sea Ice Model • CICE V 3. 1 (Hunke and Dukowicz, 1997: 2002) • 5 ice categories with 4 layers in each category (cat 4, 5 are rare). • One snow layer • Same grid as ocean model ( point is common) • Changed tiling and data structure to that of ROMS.

Coupling • WRF I/O API MCT (V 2. 0. 1) Coupling Implementation (Weather Research

Coupling • WRF I/O API MCT (V 2. 0. 1) Coupling Implementation (Weather Research and Forecasting) (Model Coupling Toolkit) • Loose Coupling (1 hr step) • Spherical Coordinate Remapping and Interpolation Package (SCRIP) from LANL to map between C-grid (ROMS) and B-grid (CICE) • Freezing/Melting potential calculated by ROMS (algorithm of CCSM, not Steele et al 1989) • CICE/CSIM data structure converted to ROMS tiling.

Domain • Thin lines are isobaths • Shaded area is under Ross Ice Shelf

Domain • Thin lines are isobaths • Shaded area is under Ross Ice Shelf • Black line is section for later figures

Forcing • Atmospheric forcing from ECMWF. • Daily values of wind, air temp, pressure,

Forcing • Atmospheric forcing from ECMWF. • Daily values of wind, air temp, pressure, humidity and clouds for 2000 and 2001. • Daily precipitation is from NCEP/NCAR reanalysis. • Added winds from Terra Nova Bay AWS at two points in ECMWF grid (called AWS case).

Annual Average wind stress AWS ECMWF 2000 2001

Annual Average wind stress AWS ECMWF 2000 2001

Open Boundaries • T and S specified by WOA 01 monthly climatology (radiation+adaptive nudging)

Open Boundaries • T and S specified by WOA 01 monthly climatology (radiation+adaptive nudging) • Integrated transport from OCCAM (radiation+adaptive nudging) • Radiation for baroclinic flow • Monthly SSM/I climatology for ice concentration on eastern boundary (imposed+nudging area) • Ice thickness is 30 cm for ECMWF and 50 cm for AWS

Forcing and Coupling Ocean only

Forcing and Coupling Ocean only

Flow diagram MPI barrier at symbols

Flow diagram MPI barrier at symbols

Results Shown Zonal salinity section in February Monthly average ice concentration Interannual differences in

Results Shown Zonal salinity section in February Monthly average ice concentration Interannual differences in ice concentration and thickness Relative importance of dynamics and thermodynamics Terra Nova Bay effect on salinity Bottom salinity

Salinity year 4 Higher salinity with dynamic sea ice Too much salinity Convection over

Salinity year 4 Higher salinity with dynamic sea ice Too much salinity Convection over bathymetry Mixed layer thickness?

Ice Concentration Nov 2000 AWS Climatology ECMWF SSM/I

Ice Concentration Nov 2000 AWS Climatology ECMWF SSM/I

Ice Concentration Dec 2000 OBC

Ice Concentration Dec 2000 OBC

Ice Concentration Jan 2001

Ice Concentration Jan 2001

Annual Average Ice 2000 AWS forcing 2001

Annual Average Ice 2000 AWS forcing 2001

Annual Average Ice processes for 2001 Thermodynamic growth, dynamic loss shows polynyas. Dynamic growth

Annual Average Ice processes for 2001 Thermodynamic growth, dynamic loss shows polynyas. Dynamic growth at north shows OBC problem

Bottom Salinity Initial Year 3

Bottom Salinity Initial Year 3

Bottom Salinity Year 3 Climatology

Bottom Salinity Year 3 Climatology

AWS wind effect on Terra Nova Bay (year 3) HSSW Effect may be too

AWS wind effect on Terra Nova Bay (year 3) HSSW Effect may be too much due to expanded area

Conclusions • Small scale winds important for polynyas • Ice open boundary conditions need

Conclusions • Small scale winds important for polynyas • Ice open boundary conditions need good information • Dynamic sea ice produces more salt than imposed sea ice. • HSSW is created and drains to shelf break as expected.

The End The BEDMAP data were provided courtesy of the BEDMAP consortium. Computer facilities

The End The BEDMAP data were provided courtesy of the BEDMAP consortium. Computer facilities and support were provided by the Commonwealth Center for Coastal Physical Oceanography. This work was supported by the U. S. National Science Foundation Grants OCE-99 -11731 and OPP-03 -37247.

Ice Concentration Jan 2001

Ice Concentration Jan 2001

Ice Concentration Dec 2001

Ice Concentration Dec 2001

Ice Concentration, Nov 2001 AWS climatology ECMWF SSM/I obs

Ice Concentration, Nov 2001 AWS climatology ECMWF SSM/I obs

Salinity year 2

Salinity year 2

Salinity year 3

Salinity year 3

Tnb year 2

Tnb year 2

Tnb year 1

Tnb year 1

Winds near Terra Nova Bay Black arrows are NCEP/QSCAT Red arrow is AWS station

Winds near Terra Nova Bay Black arrows are NCEP/QSCAT Red arrow is AWS station