comparison of gerb l2 fluxes (v002 and v999) with ceres and the met office global forecast model
DESCRIPTION
Comparison of GERB L2 fluxes (V002 and V999) with CERES and the Met Office global forecast model. Richard Allan Environmental Systems Science Centre, University of Reading. Recent comparison: 12 th December 2005 1200 UTC. November 2005 animation. GERB (BARG)Model. Spot the difference…. - PowerPoint PPT PresentationTRANSCRIPT
Comparison of GERB L2 fluxes (V002 and V999) with CERES
and the Met Office global forecast model
Richard Allan
Environmental Systems Science Centre, University of Reading
Recent comparison:
12th December 2005 1200 UTC
November 2005 animation.
GERB (BARG) Model
Spot the difference…
Comparison of V002 and V999 L2 GERB fluxes
- Diurnal cycle- Monthly mean comparison with CERES- Comparison with Met Office global model- Time-series including latest data
V999 V002
Diurnal animations of V999 and V002 OLR, July 2004
V999 Improvements
• Dawn/Dusk “wobble” and artifacts• Line of anomalously high OLR ~ equator• Hot spots removed (N. Sahara)
• ?problem with sun-glint
• Also note: Gulf of Guinea, Brazil convection, mid-latitude “noise” (see Ruth’s talk)
Comparison with CERES monthly-mean ERBE-like data
Use combined Aqua-Terra FM1+FM4 ES4 product
July 2004 monthly-diurnal mean
GERB (BARG, V999) CERES FM1+FM4
OLR
Albedo
July 2004 monthly-diurnal mean differences GERB (BARG, V999) minus CERES FM1+FM4
∆OLR~ -6 Wm-2
∆RSW~ -8 Wm-2
∆Albedo~ 0.02
∆ISW ~ .2 Wm-2
July 2004 monthly-diurnal mean differences GERB (BARG, V002) minus CERES FM1+FM4
∆OLR~ -5 Wm-2
∆RSW~ -2 Wm-2
∆ISW~ 0 Wm-2
∆Albedo~ -0.01
December 2004 monthly-diurnal mean diffs GERB (BARG, V999) minus CERES FM1+FM4
∆Albedo~0.03
∆OLR~ -6 Wm-2
∆RSW~ 10 Wm-2
∆ISW~ 0 Wm-2
CERES comparison: summary
• LW: GERB is 2.5% < CERES (~ 6 Wm-2)– V999 has removed view-angle dependence
• SW: GERB(V999) ~10% > CERES– ~10 Wm-2 RSW or 0.02-0.03 albedo – V002: smaller differences
• Differences apply for July and December 2004
Comparison with Met Office global model simulations:
SINERGEE
∆OLR ∆AlbedoJuly 2004 1200 UTC
Model minus GERB
V002
V999
OLR
OLRc
AOD
0.55um
Daytime clear-sky OLR: model minus GERB (Jun/Jul/Dec) V002 V999
Daytime clear-sky άSW: model minus GERB (Jun/Jul/Dec)
V002 V999
0000 0600 1200 1800
Model minus GERB, December 2004. V002(top); V999 (bottom)
December 2004
1200 UTC
Model minus GERB
∆OLR ∆Albedo
V002
V999
November 2005
1200 UTC
Model minus GERB
Clear-sky
V002
∆OLR ∆Albedo
Summary: V002-V999
• Improved diurnal cycle of OLR (sunglint?)• Monthly-diurnal mean vs CERES
– GERB LW ~2.5% < CERES (V002 similar)– GERB SW ~10% > CERES (V002 less)
• Sahara: increase in model-GERB diffs in LW and SW– but regions of known error (mineral dust, surface albedo)
• Oceans: reduction in model-GERB SW difference– Model low-clouds are too reflective
• Clear-sky fluxes over ocean: agreement with model remains– some improvement although GERB is now more reflective over
Mediterranean and Canaries
“Global” Ocean, 1200 UTC data
Differences (model minus GERB)
“Global” Ocean, 1200 UTC data
CLEAR-SKY differences (model minus GERB)
“Global” Ocean, 1200 UTC data
Clear-sky Mediterranean 1200 UTC
Stray Light: evident in 0000 UTC comparisons
Continuing model evaluationComposites
Model-GERB OLR (0, 6, 12, 18 UTC)
Model-GERB RSW (6, 12, 18 UTC)
Clear-sky composites 2003-2005
MODEL
GERB
Clear-sky LW clear-sky SW
Analysis of model marine strato-
cumulus
Cloud radiative effect and fraction
composites:
Model (left) GERB(right)
Diurnal & daily variation in stratocumulus (July 2004):
Model
MPEF
RMIB
PLANS
• SINERGEE– Continued near-real time comparisons– July 2004 model experiments– SEVIRI simulations?
• DABEX
• RADAGAST
Model simulations…
Proving problematic!
DABEX/RADAGAST