Comparisons between GERB and the Met Office NWP model

Comparisons between GERB and the Met Office NWP model

Richard Allan, Tony Slingo

Environmental Systems Science Centre, University of Reading

Sean Milton, Malcolm Brooks

Met Office, Exeter

Thanks to the GERB International Science Team

Sinergee project: www.nerc-essc.ac.uk/~rpa/GERB/gerb.html

GERB Model

Harries et al. (2005) BAMS; Allan et al. (2005) JGR

Objectives

• Improve experience with satellite datasets including GERB

• Timely Model Evaluation– using geostationary data independent of the

assimilation system

• Understanding of physical processes

Sh

ort

wav

eL

on

gw

ave

Mean model bias: 2006

All-sky Clear-sky

Allan et al. (2007) QJRMS

All-sky Clear-sky

Sh

ort

wav

eL

on

gw

ave

Marine stratocumulus

Convective cloud

Radiative biases in the Met Office global model

Convective outflow

Surface albedo

All-sky Clear-sky

Sh

ort

wav

eL

on

gw

ave

Mineral dust aerosol

Dust impact on longwave radiation

• Large perturbation to Met Office model OLR during summer over west Sahara– Consistent with high mineral dust aerosol optical depth

(see also Haywood et al. 2005, JGR)

– Used in planning of GERBILS

Model minus GERB OLR: July 2006, 12-18 UTC

All-sky Clear-sky dust aerosol

1800

1200

120

0 C

LEA

R

2003 2004 2005 2006 2007

Model-GERB OLR difference time series 15oW-5oE, 15-30oN

Dust impact on longwave radiationModel minus GERB OLR: 12-18 UTC

All-sky Clear-sky dust aerosol

Jun

e 18

-28

2007

Ju

ly 2

006

19 June 2007

180

0

1500

1200

21 June 2007

180

0

1500

1200

Conclusions

• Model evaluation– Satellite data independent of assimilation system– Good feedback for modellers and satellite team– Identified biases relating to cloud and surface

• Mineral dust aerosol over Sahara– Monthly longwave radiative effect up to 50 Wm-2

– Large effect of single events (Slingo et al. 2006, GRL)

– June 18-28 2007: smaller dust events than previous years but detectable OLR bias

Marine Stratocumulus

• Curious banding structure– Transition across model

levels (see Lock et al. 2001, MWR)

• Cloud reflectivity bias– Model low-altitude

stratiform clouds are too reflective

Cloud liquid water path

Bias: model minus GERB; SSM/I; SEVIRI

Albedo Liquid Water Path Cloud

Reduction in model bias from June to July 2006 - relates to cloud liquid water

…but see also Horvath and Davies (2007) JGR

Convective cloud

5th June 2006

Convective Decay Time-scale

• Unrealistically low levels of convective cloud

• On-off; common problem in models

• Simple fix…

Improved shortwave reflectivity

• Increased convective cloud cover

• But is the physics any better?

• Future work: Comparisons with CloudSat