290 likes | 373 Views
GCSS-GPCI / BLCL-RICO 18-21 sept 2006. Impact of new ARPEGE physics on RICO case. P. Marquet, CNRM . Toulouse . Météo-France. (19-sept-2006). + ARPEGE CLIMATE & N.W.P. team + Meso-Scale CNRM team + ENM (School of Met.) + CERFACS + IPSL-LMD …. Introduction.
E N D
GCSS-GPCI / BLCL-RICO 18-21 sept 2006 Impact of new ARPEGE physics on RICO case P. Marquet,CNRM. Toulouse. Météo-France. (19-sept-2006) + ARPEGE CLIMATE & N.W.P. team + Meso-Scale CNRM team + ENM (School of Met.) + CERFACS + IPSL-LMD …
Introduction First aim of the study : • Run the French Météo-France (CNRM) ARPEGE-CLIMAT SCM for the new RICO_composite (L80 and L31) … • Send the results for the inter-comparison : done / Pier … + 4 Impact studies : • Test STANDARD versus NEW physics (Turb, Conv,-Phys) ? • Test of ( L80 / t = 5 mn ) versus ( L31 / t = 15 mn ) ? • Impact of the explicit Top PBL entrainment ? • Impact of the (Dry)-thermals ?
The Climate ARPEGE physics : STANDARD / NEW1 / NEW2 STANDARD / V3-V4 NEW1 / Mixed NEW2 / IPCC / full-Diagnostic Diag+Prog full-Prognostic TURB DIAG. Mellor-Yamada DIAG. M&Y PROGN. / C.B.R. e / t = 0 e / t = 0 e / t = P + Dif – Dis moist PDF / Bougeault moist PDF / moist PDF / Bougeault Bougeault / Bechtold Micro-Phys DIAG. PROGN. PROGN. Smith/Kessler Smith/Kessler Bulk - Lopez q_liq / q_i ce q_liq / q_ice q_cloud / q_rain Shallow part in TURB , but why ? Mass-Flux Mass-Flux Convection + Mass-Flux Bougeault ? CAPE / Gueremy CAPE / Gueremy Deep Mass-Flux / Bougeault Mass-Flux Mass-Flux Convection Convergence of HU CAPE / Gueremy CAPE / Gueremy Top-PBL NO YES YES Entrainment Grenier & Breth. Grenier & Breth.
Validation of N_low : GCM (T63-L31) (DJF+JJA) N_low - ISCCP Strato Cu <- STD Lopez + CV_GY + TKE-CBR + Ent_PBL Strato Cu <- NEW
The NEW2 turbulent scheme : TKE-CBR TKE-C.B.R. (2000) + B.L. (1989) for Mixing Length + F2 & 3 / Bougeault (1982) & Bechtold (1995) ; EUROCS : GCM <-> EPCI+GPCI+ACI <-> SCM
Variance of q_cloud : PDF TKE-C.B.R. (2000) + B.L. (1989) + F2 & 3 / Bougeault (1982) & Bechtold (1995) Micro-physics : pdf Cu Sc
An Explicit Top-PBL Entrainment Vertical Diffusion of the Betts variables : _l and q_t Grenier (ARPEGE) ( A1 = 0.16 ) ( A2 = 0.0 )
SCM / EUROCS_ARM_Cu (Lenderink) LES-KNMI ARP-STD / L19 q_cloud ARP-NEW2 / L19 + Top PBL ent. ARP-NEW2 / L19
SCM / EUROCS_ARM_Cu (Lenderink) LES-KNMI ARP-STD THETA(L19) ARP-NEW2 + Top PBL ent. ARP-NEW2
Ayotte 24SC / L96 Hourdin, Couvreux, Menut, J.A.S., 2002.
And for RICO ? Impact studies : • Test STANDARD versus NEW physics : STD versus NEW2 • L80 / t = 5 mn ; L31 / t = 15 mn • Impact of theexplicit Top PBL entrainment : NEW2_noent • Impact of the (Dry)-thermals : NEW2_noDTh
STD NEW2 NEW2_noent NEW2_noDTh RICO L80 / THETA_l STD instable ; NEW2 better, But with a drift z>2km ? a –0.5 K bias in the PBL ? a very small impact of « entr. » a large impact of Dry Thermals
STD NEW2 NEW2_noent NEW2_noDTh RICO L31 / THETA_l STD noisy ; NEW2 better, with a smaller drift z>2km ! ? a –0.5 K bias in the PBL ? a greater impact of « entr. » and larger impact of Dry Thermals
STD NEW2 NEW2_noent NEW2_noDTh RICO L80 / Zonal Wind STD noisy ; NEW2 better, But why these oscillations ? A small impact of « ent » A greater impact of Dry Thermals
NEW2_noent NEW2 STD NEW2_noDTh RICO L80 / Cloud Cover STD very noisy !! NEW2 better, Cloud Base and Maxi OK, Drift of Cloud top <-> theta ? A small impact of « ent » A large impact of Dry Thermals What about the values of C.C. ??
RICO / Cloud Cover Average 24h - 30h L80 LES KNMI L31 STD very noisy up to 3.5 km !! NEW2 better L80 : from 0.5 to 2.5 km : OK ; the good shape L31 : too deep cloud 0.2 to 4 km !! L31 : Large impact of Top PBL entr. ! What about the values of C.C. ??
RICO / Cloud Content Average 24h - 30h L80 LES KNMI L31 SAME as for CLOUD-COVER… Except that values compare to LES… STD very noisy up to 3.5 km !! NEW2 better L80 : from 0.5 to 2.5 km : OK ; the good shape L31 : too deep cloud 0.2 to 4 km !! L31 : Large impact of Top PBL entr. !
RICO / <w’’l> Average 24h - 30h NEW2 L80 realistic… except close to the surface !! L80 K*(m/s) LES KNMI W/m2 L31 W/m2
L80 LWP L80 Cloud Cover NEW2 NEW2 L80 Precip. mm/day L80 LH (W/m2) NEW2 NEW2
RICO / TKE Average 24h - 30h L80 LES BOMEX W/m2 m2/s2 L31 W/m2
Conclusions Mainly, questions !… • Apart from the (internal) validation of the ARPEGE physics… • Why this drift above 2.5 km ? smaller with L31 /t =15 mn : w ? • A deep impact of top-PBL/TURB shallow convection • Small for EUROCS- Cu / large for RICO… Precip ? • Are the oscillations for (u,v) Wind observed by others ? • How to compare Precip & C.C. to LES or Obs. ? => Radiation ! • A real interest for this RICO case ! “Composite” -> “Long_Run” ? RICO GPCI & AMMA-CI (next talk)… Continue EUROCS’ method…
STD NEW2 NEW2_noent NEW2_noDTh RICO L31 / Cloud Cover STD very noisy up to 4 km !! NEW2 better, But worse than L31 (clud base ?) A large impact of « ent » And larger impact of Dry Thermals
RICO / <w’q’t> Average 24h - 30h L80 LES KNMI W/m2 (g/kg)*(m/s) L31 W/m2
RICO / <w’’vl> Average 24h - 30h L80 LES KNMI W/m2 K*(m/s) L31 W/m2
STD NEW2 L31 L31 RICO SH (W/m2) L80 L80 STD NEW2
L31 L31 NEW2 STD RICO LH (W/m2) L80 L80 STD NEW2
NEW2 L31 L31 STD RICO Cloud Cover L80 NEW2 L80 STD
L31 L31 NEW2 STD RICO Precip. L80 L80 NEW2 STD
NEW2 L31 L31 STD RICO LWP L80 NEW2 L80 STD