Single column models calculate the time evolution of vertical profiles of temperature and moisture

1
Single column models calculate the time evolution
of vertical profiles of temperature and moisture
Adiabatic tendencies represent advection by the
large-scale flow and are prescribed
represent diabatic contributions to
the temperature and moisture
tendencies from the parameterised physics
PROS
CONS

• simple, inexpensive
• good tool for parameterisation development
  understand, develop and evaluate the physical
  representations through comparison to obs and
  high res simulations
• compare results with other SCMs is the model
  error a community wide problem or model specific

2
cumulative precipitation (mm)
obs
precip 3 hour av (mm/h)
pc2
no pc2
86
14
95
5
30 min average
3
Jon Petch 2D CRM 500
(750m) x 46 (500m)
4
Jon Petch 2D CRM 500
(750m) x 46 (500m)
5
Jon Petch 2D CRM 500
(750m) x 46 (500m)
6
Jon Petch 2D CRM 500
(750m) x 46 (500m)
7
Jon Petch 2D CRM 500
(750m) x 46 (500m)
8
Jon Petch 2D CRM 500
(750m) x 46 (500m)
9
Jon Petch 2D CRM 500
(750m) x 46 (500m)
10
Cloud volume is the prognostic variable but the
radiation scheme needs to know cloud area.
? Uncertainty in whether one should use the
diagnostic convective cloud fraction or
purely the prognostic. ? The prognostic cloud
represents the detrained condensate into the
environment and not the tower cloud. ?
Experience at UKMO is that at times a separate
convective cloud fraction is necessary due to
inability to represent extreme pdf shapes. ?
For the sake of a fair comparison between pc2 and
no pc2 I have included the convective cloud
fraction in the cloud area diagnostic.
11
Jon Petch 2D CRM 500 (750m) x 46 (500m)
1. the combined cloud area
2. CRM cld fraction
3. volume cloud fraction prognostic
4. area fraction not including convective
12
Jon Petch 2D CRM 500 (750m) x 46 (500m)
----- obs ----- CRM
13
Jon Petch 2D CRM 500 (750m) x 46 (500m)
14
Jon Petch 2D CRM 500 (750m) x 46 (500m)
15
(No Transcript)
16
Using combined area cloud fraction
17
Using bulk/volume cloud fraction
18
2 km
5 km
10 km
15 km
19
2 km
5 km
10 km
15 km
20
whole 25 day period
days 19-26 active monsoon
days 26-37 suppressed monsoon
days 37-44 break period
21
timeseries of q increments
obs
conv
microphysics
b. layer pc2
22
whole 25 day period
days 19-26 active monsoon
days 26-37 suppressed monsoon
days 37-44 break period
23
whole 25 day period
days 19-26 active monsoon
days 26-37 suppressed monsoon
days 37-44 break period
24
days 19-26 active
diurnal composite of cloud fraction
days 26-37 suppressed
days 37-44 break
25
Summary

• Temperatures after active period show significant
  cold bias below 11km and warm bias above sub
  period runs may help
• Similarly for moisture, dry bias in lowest 10km
  but quite strong below 2km forcing data derived
  for land only and ocean only may help
• Clouds look generally okay, question about cloud
  area parameterisation which overestimates the
  high cloud area fractions
• PC2 shows good agreement with obs for the pdfs of
  lwp and iwp
• Greatest T bias in the break period and q bias in
  the suppressed phase
• The CRM and PC2 run have deeper cloud for hector
  event Jan 24-26, which also shows up in not
  enough olr but not for no PC2 run
• Models all seem to dissipate cloud too soon after
  the hector event the SCMs more so than the CRM
• During initial suppressed phase there is too much
  downwelling solar and too much olr models have
  thinner anvil, no PC2 is better than PC2 and CRM
• Too much cloud for SCM runs in the break period,
  CRM is better
• PC2 shows more similar behaviour to CRM both
  the good and bad
• Poor simulation of suppressed convection has been
  identified as having a detrimental effect on
  simulations of sub-seasonal variability in
  tropical convection, and in particular the poor
  representation of the MJO by climate models -
  TWP-ICE good case study