Key land surface coupling processes over semi-arid region analyzed in the DICE case. S. Ait Mesbah, F. Cheruy, J.L. Dufresne,F. Hourdin, JY Grandpeix, M.P. Lefebvre Outline • 2 PBL and LSM schemes developed at LMD/IPSL and used for the 1D DICE case • The DICE simulations • A simple conceptual model to analyze the results • 3D simulations and generalization • conclusion 2 PBL and LSM schemes developed at LMD and used for the 1D DICE cases CASE DICE : Diurnal Land-atmosphere Coupling Experiment ST - Eddy diffusivity Kz, - Counter gradient (1K/km), - Surface PBL : Louis 1979 PBL NP - Yamada (1983) Thermal plume 2 PBL and LSM schemes developed at LMD and used for the 1D DICE cases CASE DICE ST PBL - Eddy diffusivity Kz, - Counter gradient (1K/km), - Surface PBL : Louis 1979 ORCHIDEE2 (OR2) - Moisture : Conceptual model. ( 2m depth. 2 layers, Top layer can vanish, No drainage) - Temperature : heat diffusion eq (7 layers, 5m), thermal properties depend on moisture. SOIL NP - Yamada (1983) Thermal plume ORCHIDEE11 (OR11) - Moisture : Solves Richards diffusion equation. 2m depth. 11 layers, Free drainage) -Temperature heat diffusion eq (7 layers, 5m), thermal properties depend on moisture PLAN • 2 PBL and LSM schemes developed at LMD and used for the 1D DICE cases • The DICE simulations • A simple model to analyze the results • 3D simulations and generalization • conclusion DICE simulation 80% of C3 grassland, LAI = 0.4, 20 % of bare soil, loam, surface roughness=0.01m - KANSAS : 37.65N,263.265E 23 oct 1400 to 26 oct 1999. Dry, no rain, no clouds. DICE simulation Différence OR2 – OR11 • Different moisture profiles ( OR2 ,OR11 )induce different thermal inertia at the surface. LMDZOR - ORC2 et ORC11 IT variable, due to differences in soil moisture parameterization LMDZ-Bucket Beta = 0.03, IT variable PhyA IT 946 PhyB IT 946 DTROR11>DTROR2 ST et NP similar during the day NP amplifies the differences: colder Ts during the night OBS LMDZOR - ORC2 et ORC11 Imposed IT : 1080 (J/(K.m².s-0.5) PhyA IT 2190 PhyB IT 2190 LMDZ-Bucket , IT variable The differences in nocturnal Ts are due to the qualitatively similar to LMDZOR-ORC2 et ORC11. parameterization of the BL DICE simulation • DTR weaker with ORC2 , due to an higher G; A bucket LSM with prescribed b coupled with LMDZ reproduces the behavior • DTR is less affected than expected ; is less than expected (G divided 2, same DF> DTR divided 2) DF (atmospheric processes) are impacted • Strong day/night dissymmetry in the DTR sensitivity to G • The atmospheric physics modulates the temperature differences (lower with ST / NP) DICE simulations: Questions 1- How to explain the greater sensitivity to the thermal inertia during the night ? 2- What is the role of the boundary layer scheme in the asymmetry? PLAN • 2 PBL and LSM schemes developed at LMD and used for the 1D DICE cases • The DICE simulations • A simple model to analyze the results • 3D simulations and generalization • conclusion SIMPLE MODEL, ENERGY BALANCE SWdn LWup LWup LWdn LE H LE H SWup G Swdown – Swup + Lwdown – Lwup = H +LE + G LWdn G Swdown – Swup + Lwdown – Lwup = H +LE + G SIMPLE MODEL, ENERGY BALANCE ATMOSPHERIC FORCING LWup SWdn LWdn SWup Swdown – Swup + Lwdown – Lwup = H +LE + G Swdown – Swup + Lwdown – Lwup = H +LE + G SIMPLE MODEL, ENERGY BALANCE Response of the surface to atmospheric forcing LWup LWup LE H LE H G Swdown – Swup + Lwdown – Lwup = H +LE + G G Swdown – Swup + Lwdown – Lwup = H +LE + G Simple model Equilibrium a Duration of the day fs Surface balance Sensitivity of X to the thermal inertia : From 3days DICE SP NP Bucket , b constant G2 = 2140 J.m-2.K-1.s -.5 , G1= 940 J.m-2.K-1.s -.5 Simple model SP NP • If a = .5 , sensitivity of atmospheric flux to I identical changes without impacting Tmean (DICE, a =0.36) = 1 and DTR • day/night differences arises from H’ and LE’ Both are significantly higher during the day • DICE dry LE’ << H’ : Dissymmetry arises from the sensible heat flux H’day and H’night function of stability of the PBL Day= PBL unstable, small variation of Ts translates into turbulence Night = stable PBL, decoupled surface, small turbulence, The energy stored in the soil released, only radiation responds to change of Ts , a small change of G is changed in a change of Ts Sensitivity to the PBL scheme • The larger the contrast of PBL stability between day and night, the higher the dissymmetry of the sensitivity of Ts to the thermal inertia. • SP produces more mixing than NP during the night, the day/night stability contrast is lower Ts(SP) is less sensitive to the thermal inertia than Ts(NP) PLAN • 2 PBL and LSM schemes developed at LMD and used for the 1D DICE cases • The DICE simulations • A simple model to analyze the results • 3D simulations and generalization • conclusion From DICE to arid and semi-arid regions Increase G NP Low thermal inertia High thermal inertia Conclusions • The thermal inertia (parameterization of soil moisture and thermal properties of the soils) impact DTR and Tmean • It impacts the atmospheric fluxes as well • The diurnal variation of the stability of the atmosphere is essential • The sensitivity to the thermal inertia is higher with PBL schemes which produces less mixing during the night • Its parameterization in GCM’s can lead to biases in the surface temperature