Coupled Ocean-Atmosphere model for Regional Climate studies in the

Mediterranean Sea

William CabosDepartamento de FísicaUniversidad de Alcalá

European ROMS Workshop 6-8 November

The aim of our projectRegionalization of Climate change projections of IPCC 4th AR models for the

Iberian Peninsula

ToolRegional coupled Ocean–Atmosphere Model

Atmospheric componentREMO (Regional Model)

Ocean component

ROMS (Regional Ocean Model System)

Coupler OASIS3 coupler

Each of these components has been successfully used for a wide range of climate studies.

Coupling of REMO to ROMS

Regional simulations of high resolution consistent with global circulation when resources are limited

avoid inconsistences between global and regional models

reduce the effects of grid size changes from global to region of interest

Advantages of nested climate models

Global model

Intermediate nested model

Region of interest

Model Setup

Model domain: Atlantic Ocean, Mediterranean and Black Sea

Mediterranean circulation is poorly simulated in global models

Model Setup

Conditions to take into account in the model

It should simulate the Mediterranean present time climate and climate change projections

The Atlantic ocean is necessary for the global climate signal

The ocean and the atmosphere should have the same grid:

no interpolation

1 hour exchange frecuency

The masks should be as like as possible

The Black Sea should be included

COUPLED MODEL

AtmosphereREMO

Coupler OASIS

OceanROMS

Atmospheric General Circulation Climate Model.

-developed in the Max Planck institute of Metereology

-Hidrostatic

- maximal grid resolution: 1/10 degree.

This grid resolution enables the model to resolve the most important topographic features

-Sofisticated vegetation scheme

-State of the art soil, hydrographic and glaciar models

-Widely used for climatic and impact studies in Europe

-Echam4 physics

-Rotated coordinates

REMO (REgional MOdel)

REMO Validation : Temperature

REMO and CRU temperature: South Spain

REMO and CRU temperature differences

Different regions

REMO Validation : Precipitation

North Spain Central Spain

All Spain

REMO coupled model domain

REMO coupled model setup

OASIS

Flexible, modular, well supported

Sinchronization of component models

Coupling fields exchange: arbitrary number, frecuency

I/O actions

Interpolation

PSMILe communication: MPI, direct communication between models with the same grid and partition.

ROMS

Roms-Agrif version

0.22 degrees resolution

30 leves, s=6.0b=0.

Open boundaries: restoring to global coupled model data

Rotated coordinates:

Topography and ocean forcing fields are rotated with Eulerian angles, f calculated in geogrphical coordinates

GDEM3 initial conditions

Spinup with REMO 1958-2001 climatology : winds, heat and fresh water fluxes

ROMS

ROMS: spinup forcing

REMO wind stress:

February

REMO wind stress:

July

ROMS: spinup forcing

REMO Heat Flux:

February

REMO Heat Flux:

July

ROMS without nudging

ROMS Black Sea with nudging (.11 degrees)

ROMS .22 with nudging

ROMS .22 with nudging

-Finish the coupling interface

-Fix the bosphorous problem

-Use 6-hour REMO data for spinup

-Include tides

-Validate ROMS

-experiments with the coupled model

Future work

REMO

REMO

A hierarchy of nested models with growing resolution allows:

Regional simulations of high resolution consistent with global circulation when resources are limited

To avoid inconsistences between the global model and the regional model in the region of interest

To reduce the impact of grid size difference

Nested models

Global model

Intermediate nested model

Region of interest

OASIS

Well structured routines, allows model exchange

REMO

REMO

REMO

REMO

REMO

ROMS: downscaling of SODA data

REMO

Activities of our group related to ROMS

Regional Studies: Climate change and present climate validation:

Present climate: 1958 - 2001

Climate change: different scenarios:2000 - 2100

Global data:

IPCC 4th AR ocean-atmosphere models

Regional models:

Ocean – ROMS; forced with global and regional atmosphere

Atmosphere – REMO

Coupling necessary for local effects

REMO

ROMS: downscaling of SODA data

SODA

Ocean reanalysis: conbines observations with ocean data

from 1958 to present

.5 degree resolution

Does not show the most important features of Mediterranean circulation : resolution ?

ROMS

.12 degrees simulation with SODA boundary data

Can be used for higher resolution local climate studies

Weak nudging to SODA data?