Titel corps 28 News Gothic

1
On land surface temperature trends in calm and
windy nights Gert-Jan Steeneveld and Bert
Holtslag
2
Year-Glob
DJF-NH
Windy conditions (upper tercile) Calm conditions
(lower tercile)
Conclusion by Parker (Nature, 2004) No
influence of wind speed on long-term surface
air temperature trends!
JJA-NH
3
However
Pielke and Matsui (GRL, 2005) argue that
long-term climate records of surface air
temperature are not expected to have the same
trends for light and stronger winds at night
So there seems to be a contradiction with
Parkers study Here we use state of the art
atmospheric boundary layer model coupled to the
land surface to investigate to what extend
the surface temperature trend at night is
dependent on wind speed
4

• Three different SBL-regimes
• Fully turbulent
• Intermittent turbulent
• Radiation dominated

Cloudy Clear
Radiative
Fully turb.
Int. turb
(see Van de Wiel et al, 2002)
Mechanical forcing
Can we model these different regimes with a
state of the art single column model?
5

• 1D Model description
• Local Turbulence Scaling (Duynkerke, 1991)
• Radiation (Garratt Brost, 1981)
• Soil Vegetation layer
• High resolution in soil and atmosphere,
• in particular near the surface

See for details Steeneveld et al, 2006,
J.Atmos.Sci., 63, 920-935
6
Modeling 3 diurnal cycles with contrasting nights
in CASES99
Int. turb
Rad.
Fully turb
October 23-26
Steeneveld et al., JAS, 2006
7
Int. turb
Fully turb
Non- turb
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11

• Now use the validated model
• Run model again using set-up for 23/24 Oct in
• CASES-99
• Vary the Geowind in range of 0.5 to 18 m/s
• Use 100 layers 10 km atmospheric column
• Uniform CO2 concentration across atmosphere
  starting with
• 330 pm
• To calculate climate sensitivity we study a
  uniform CO2 increase
• (in steps of 10 up to 40) impacting on downward
  longwave
• radiation

12
Model results for low and high wind cases at end
of night
G9 m/s
G2 m/s
Note lines from left to right indicate increased
CO2 in steps of 10
13
Sensitivity near surface rather constant for
Geowinds lt 10 m/s!
14
Conclusions
State of the art atmospheric boundary layer model
coupled to the land surface shows good
agreement with observations for CASES99 (and
elsewhere) given sufficient vertical
resolution Model calculations for near surface
air temperature indicate that increase due to
enhanced CO2 is rather independent from wind
speed for geowinds lt 10 m/s Results differ from
study by Pielke and Matsui (2005) who found a
much stronger variation of air temperature change
with wind speed
15
Results by Pielke and Matsui (GRL, 2005)
16
Concepts of Stull as used by Pielke and Matsui
Here the heat flux FH is prescribed independent
from wind speed U
17
Observations CASES-99
Red dot Value used by Pielke and Matsui for all
wind speeds in their sensitivity study!
18
Ta
Fast responding vegetation layer
H
Tveg
Vegetation
Tsoil
G
soil
Td
19
(No Transcript)
20
(No Transcript)