Validation of the IPSL models used for Mediterranean climate studies

Laurent Li

Laboratoire de Météorologie Dynamique (LMD)

Institut Pierre-Simon Laplace (IPSL)

CNRS/UPMC, Paris, France

Mediterranean climate modelling workshop, Trieste, 13-15 Oct 2008

IPSL: a federation of 5 laboratories (700 persons)And several institutions

Who we are IPSL/LMD

IPSL Earth system model

Continents Atmosphère Océan

Surface continentaleSol et végétation

Circulation atmosphérique

Physique

Circulation OcéaniqueGlace de mer

LMDzORCHIDEE

PISCES

INCA

STOMATE

ORCALIM

C02 CarboneCarboneCarbone

Biogéochimie continentale

Biogéochimie etBiologie marine

ChimieGaz&

Aérosols

DMS

NutritifsChimie

CH4, COVAérosols

Continents Atmosphère Océan

Surface continentaleSol et végétation

Circulation atmosphérique

Physique

Circulation OcéaniqueGlace de mer

Surface continentaleSol et végétation

Circulation atmosphérique

PhysiqueGlace de mer

LMDzORCHIDEE

PISCES

INCA

STOMATE

ORCALIMLMDzORCHIDEE

PISCES

INCA

STOMATE

ORCALIM

C02 CarboneCarboneCarbone

Biogéochimie continentale

Biogéochimie etBiologie marine

C02 CarboneCarboneCarbone

Biogéochimie continentale

Biogéochimie etBiologie marine

ChimieGaz&

DMS

NutritifsChimie

CH4, COVAérosols

ChimieGaz&

Aérosols

DMS

NutritifsChimie

CH4, COVAérosols

Circulation Océanique

• Atmosphere: LMDZ-4• Land: ORCHIDEE• Ocean: ORCA• Sea-ice: LIM• Coupler: OASIS-3 Medium resolution:

LMDZ 96x71x19 - ORCALIM : 2°

The IPSL-CM4 climate model

• LMDZ-Med is a global atmospheric GCM with variable grid and a zoom over the Mediterranean basin. Local resolution: 30 km.

• It is run as a regional climate model, with nudging conditions (every 6 hours) from a global model (LMDZ-g, ERA40, IPCC, etc.) at low resolution outside the zoom. The model is free to have its own behaviours inside the zoom.

LMDZ-Mediterranean model

IPCC-A2 scenarios regionally-enhanced with LMDZ-Med

IPSL CNRM GFDL

Middle of the 21st century

End ofthe 21st century

Middle of the 21st century

End ofthe 21st century

T

P

Changes (2070/2099 minus present) of seasonal extreme precipitations(30-year return values, from GEV distribution). Units: mm/day

Two-way (self-) nesting between LMDZ-regional and LMDZ-global

Globa

lR

egional

Globa

lR

egional

Med Sea oceanic model (MED8)

Global oceanic model (ORCA2)

Schematic of the quartriple coupling in IPSL

SST from coupled model(MED8-LMDZr)

Jan Apr

Jul Oct

SST from coupled model(MED8-LMDZr-LMDZg)

Jan Apr

Jul Oct

Climatological SST (AMIP)

Jan Apr

Jul Oct

Simulations planned in CIRCE

• Regional O-A coupled model (MED8 + LMDZr) used for the scenario A1B from 1951 to 2050 (either IPSL or MPI-M global inputs).

• Quartriple coupled model (MED8 + LMDZr + LMDZg + ORCA2) used for the scenario A1B from 1951 to 2050.

Two configurations:

•LMDZ-regional forced by prescribed lateral boundary conditions;

•Two-way nesting system between LMDZ-regional and LMDZ-global.

Two experiments (10 years for each simulation):

•ObsVeg: Observed Vegetation of present day,

•NatVeg: Natural vegetation in the Mediterranean basin – idealized situation without any anthropogenic land use (statistical model).

An example of the two-way nesting atmospheric system

Difference (NatVeg – ObsVeg)

2-way

1-way

Precipitation T2m

Precipitation T2m

Difference (NatVeg – ObsVeg)

Regional

Global

Precipitation T2m

Precipitation T2m

An example of LMDZr and LMDZg two-way nesting system for climate change scenario

(Yangtze River Basin)

Annual-mean rainfall (mm/d) (top), and its future variation (bottom: 2050-2000)

LMDZ-global LMDZ-regional LMDZ-sn

An example of LMDZr coupled to a 50-m constant-depth slab model of the Med Sea

• This study is designed to investigate how large-scale synoptic processes exert their influences on the Mediterranean regional climate.

• Some synoptic perturbations are artificially filtred out from the lateral boundary conditions of the LMDZ-mediterranean model.

Experimental design schematicCTRL NoSynExtra

NoSynTropiNoSyn

6-hr varying BC

climatological BC

6-hr varying BC

6-hr varying BC

climatological BC

climatological BC

January 2-m temperature (K)CTRL NoSynExtra - CTRL

NoSynTropi - CTRLNoSyn - CTRL

January precipitation (mm/day)CTRL NoSynExtra - CTRL

NoSynTropi - CTRLNoSyn - CTRL

Geopotential (NoSynExtra – CTRL)

500 mb300 mb

850 mb 1000 mb

• Suppression of extratropical synoptic perturbations makes a negative-phase NAO situation, i.e. more precipitation and warmer temperature in the south, but reverse in the north.

• Suppression of tropical synoptic perturbations makes a more uniform cooling and drying (but weak) in the domain.

Perspectives • more resolution at regional scale• more interactivity in the system

L.F. Richardson’s view of numerical weather prediction (1920s) (painting: François Schuiten, 2000)