© Crown copyright Met Office Met Office SCM and CRM results Adrian Lock, Met Office, UK.

© Crown copyright Met Office

Met Office SCM and CRM results

Adrian Lock, Met Office, UK

© Crown copyright Met Office

Met Office LEM/CRM set-up

• Use 2D “LES” with Dx=50m, Dz=20m in 6.4km domain (top at 8km)

• Started with Dx=100m, Dz=50m in 12.8km domain and higher resolution very similar

• Otherwise LEM set-up as for GCSS-RICO

• Run for 30 days (takes ~1 day to run)

• Run Control and +2K for points 6, 11, 12

© Crown copyright Met Office

Met Office SCM set-up

• Run parallel simulations in the Met Office SCM

• based on runs by Charmaine Franklin at CSIRO

• Run for 100 days

• Confession:

• I’ve used the development version of the climate model, HadGEM3, instead of the proposed AR5 model HadGEM2

• Test sensitivity to resolution

• L38 (~300m at 1km, as used in HadGEM2)

• L63 (~160m at 1km, as used in HadGEM3)

© Crown copyright Met Office

S6 profiles after 30 days 5 day mean in SCM (3 hour mean in CRM)

• SCM gives significantly deeper cumulus than CRM (4km cf 2km) but reasonable cloud cover profile and massflux

• SCM gives drier (warmer) BL

Cloud fractionMassfluxHumidity (g/kg)

Ctl+2K

© Crown copyright Met Office

S6 time seriesUnsmoothed

© Crown copyright Met Office

S6 time series2 day means

• SCM cloud cover matches CRM, especially at L63

• Too much precipitation in SCM (cumulus too deep)

• No sign of any cloud feedback in SCM or CRM

© Crown copyright Met Office

S11 profiles after 30 days 5 day mean in SCM (3 hour mean in CRM)

Cloud fractionMassfluxHumidity (g/kg)

Ctl+2K

• SCM again gives deeper BL than CRM

• SCM gives greater cloud cover underneath the inversion

© Crown copyright Met Office

S11 time seriesUnsmoothed

© Crown copyright Met Office

S11 time series2 day means

• Cloud cover much higher in SCM than CRM; L38 very noisy

• Too much precipitation again in SCM

• SCM has positive feedback, ~independent of resolution

• No feedback in CRM

© Crown copyright Met Office

S12 profiles after 30 days 5 day mean in SCM (3 hour mean in CRM)

Cloud fraction

Ctl+2K

• CRM boundary layer very shallow and cloud free

• SCM similar to s11

Humidity (g/kg)

© Crown copyright Met Office

S12 time series2 day means

• CRM: no cloud, so no feedback

• SCM has positive feedback at L63, negative at L38

© Crown copyright Met Office

Cloud feedback mechanisms for S11See Mark Webb’s talk

• Lower-tropospheric stability (LTS)

• Theta difference between 3km and surface

• Control LTS = 22.75K

• +2K LTS= 23.8K

• So S11 shows increase in LTS, similar to GCMs

• Does LTS implies increase in cloud = negative feedback?

• CTEI parameter, based on inversion jumps:

• Control = (11.0 -15.0) / (-15.0) = - 4.0 /-15.0 = 0.27

• +2K = (11.25-16.6) / (-16.5) = -5.25/-16.5 = 0.32

• So S11 shows increase in , similar to GCMs

• LES (Lock, 2009) suggest increase implies reduction in cloud cover, at least for Sc over Cu, = positive feedback

tp

tp

tp

e

qcL

qcL

qcL

© Crown copyright Met Office

Lock (2009): factors influencing cloud area at the inversion for shallow cumulus

© Crown copyright Met Office

Summary

• Lack of stratocumulus in CRM at s11, s12 is a concern

• I’ll try 3D on our new supercomputer when I get back

• SCM has a tendency to give deeper boundary layers than the CRM

• SCM reasonably resolution independent, except for s12

• Climate change impact on LTS and kappa appears consistent with that seen in GCMs

© Crown copyright Met Office

Summary

SW feedback

(Wm-2)

S6 S11 S12

CRM 0 0 No cloud

SCM L38 0 10 -70

SCM L63 0 10 70

© Crown copyright Met Office

Questions?