P1254325706sIQCZ

1
Response of the western European climate to a THC
collapse
A. Laurian S. Drijfhout W. Hazeleger B. van den
Hurk
Koninklijk Nederlands Meteorologisch Instituut,
The Netherlands
2
The Thermohaline Circulation

• Transports heat and salt poleward- A change
  in its strength can lead to global and regional
  climate changes
• (Vellinga and Wood 2002 Dong and Sutton 2002
  Chang et al. 2008)

3
The Thermohaline Circulation

• Transports heat and salt poleward- A change
  in its strength can lead to global and regional
  climate changes
• (Vellinga and Wood 2002 Dong and Sutton 2002
  Chang et al. 2008)
• ? What is the response of Western European
  climate
• to a THC collapse?

4
Outline

• Introduction
• Model and experiment
• Response of the 2-m air temperature
• Mechanisms leading to a different response over
  the ocean and above land
• - Response of the net atmospheric fluxes
• - Change in clouds and associated feedbacks
• - Change in moisture transport
• Conclusions

5
Model and experiment

• Coupled model ECHAM5/MPI-OM
• - Good skills at reproducing the Atlantic MOC
• - Realistic atmospheric circulation over Europe

6
Model and experiment

• Coupled model ECHAM5/MPI-OM
• - Good skills at reproducing the Atlantic MOC
• - Realistic atmospheric circulation over Europe
• ECHAM5 horizontal resolution T63, 31 vertical
  levels
• MPI-OM horizontal resolution 1.4 x 1.4,
  near the Equator 1.4x0.5, near the poles
  O(20-40 km), 40 vertical levels

7
Model and experiment

• Coupled model ECHAM5/MPI-OM
• - Good skills at reproducing the Atlantic MOC
• - Realistic atmospheric circulation over Europe
• ECHAM5 horizontal resolution T63, 31 vertical
  levels
• MPI-OM horizontal resolution 1.4 x 1.4,
  near the Equator 1.4x0.5, near the poles
  O(20-40 km), 40 vertical levels
• ESSENCE project (Sterl et al. 2008)
• 17-member ensemble of runs (1950-2100)
• - From 1950 to 2000 runs forced by the
  observed concentrations of GHG and tropospheric
  sulfate aerosols
• - From 2001 to 2100 runs followed the
  SRES A1b scenario

8
Model and experiment

• We investigate 2 five-member ensemble runs.
• In one ensemble, a freshwater anomaly of 1 Sv was
  uniformly applied between 50 and 70N over the
  northern North Atlantic Ocean from 2001 onwards,
  starting from the five initial states of the
  other ensemble.

9
Model and experiment

• We investigate 2 five-member ensemble runs.
• In one ensemble, a freshwater anomaly of 1 Sv was
  uniformly applied between 50 and 70N over the
  northern North Atlantic Ocean from 2001 onwards,
  starting from the five initial states of the
  other ensemble.
• The ensemble mean of the perturbed runs (HOSING)
  is compared to the ensemble mean of the
  associated control runs (ENSMALL) over the period
  2091-2100 in western Europe (30W-20E,
  30N-60N).
• ? How does global warming evolve when a THC
  collapse occurs during the process?

10
Model and experiment
ENSMALL HOSING

• In ENSMALL the THC gradually weakens by 20
  after 100 years due to global warming.
• The freshwater supply leads to a collapse of THC
  within 50 years.

11
Model and experiment
ENSMALL HOSING

• In ENSMALL, warming of 4C in 150 years.
• THC collapse ? Cooling of 0.7C during 25 years
  which tempers global warming

12
Response of the 2-m air temperature

• Cooling everywhere, 6C in the North Atlantic
  ocean, 2.5C in
• Western Europe
• ? Strong land/sea contrast what maintains this
  pattern?

13
Convergence of zonal MSE transport
(MSE gz CpT Lvq)
ENSMALL (m .s ), column-integrated

• MSE in the whole column is transported eastward
  and advection from the ocean is a source of MSE
  between 36 and 52N
• ? One may expect that the prevailing westerlies
  reduce land/sea contrast

14
Outline

• Introduction
• Model and experiment
• Response of the 2-m air temperature
• Mechanisms leading to a different response over
  the ocean and above land
• - Response of the net atmospheric fluxes
• - Change in clouds and associated feedbacks
• - Change in moisture transport
• Conclusions

15
Response of the net atmospheric fluxes
(SW-LW-SH-LH) surface
85 W/m² in the North Atlantic Ocean ?
Latent heat release over the ocean

• Positive values above sea indicate that the ocean
  cools the atmosphere and the atmosphere warms the
  ocean, to compensate for the absent ocean heat
  transport from the South.
• The ocean is driving the net atmospheric flux
  response

16
Response of the net atmospheric fluxes
SW surface
SW at the surface clearly supports the
suggestion of a cloud response. SW by about
6.5 W/m² over sea ? Suggesting an of clouds
there blocking the penetration of SW SW by
about 4.5 W/m² over land ? Suggesting a of
clouds there allowing more SW to penetrate to the
surface.
17
Radiative effect of clouds
(SWLW)toa, clouds - (SWLW)toa, clear sky in
ENSMALL
? Net effect of clouds in the mid latitudes
Cooling of the atmosphere (60 W/m² over the
ocean, 25 W/m² over land)
18
Change in radiative effect of clouds
(SWLW)toa, clouds - (SWLW)toa, clear sky in
HOSING-ENSMALL

• Over the ocean (8 W/m²) consistent with an
  of clouds

• Over land (4 W/m²) consistent with a of
  cloud cover

19

• Change in liquid water
• (column integrated)

liquid water over land (0.1g/m²) liquid
water over the ocean (0.04g/m²)
THC collapse ?
20

• Seasonality of the liquid water response

DJF JJA

• over land in winter
• over the ocean in summer

21

• Vertical distribution of the liquid water
  response (48N)

DJF JJA

• Over sea low clouds in all seasons (below
  900 hPa)
• Over land mid-level clouds in winter
  (between 900 and 600 hPa)

22

• Why is the liquid water increasing over the ocean?

DJF JJA
The cooler air temperature over the ocean ?
stability of the atmospheric BL, inhibiting
convection and maintaining a wetter marine
BL. ? shallow clouds over the ocean
23

• Why is the liquid water decreasing over land?

DJF JJA
? Other mechanism to explain the decrease of
cloud cover over land
24

• Change in specific humidity (q) at the surface

• q over the ocean by about 2 g/kg
• q over land by about 1.2 g/kg
• (Evaporation and the colder air is more
  easily saturated)
• ? Similar zonal contrasts as the 2-m temperature
  response

25

• Change in relative humidity (RH) at the surface

• RH above the ocean by about 5
• RH above land by about 1
• This strong land/sea contrast in the RH response
  is consistent with the clouds over land and
  the clouds over sea

26

• Change in relative humidity (RH) at the surface

• Over the ocean, in RH dominated by in T
• Over land, in RH dominated by in q

27
Change in moisture transport
Réponse du budget dhumidité

• The change in moisture sink (P-E) is balanced by
  the change in moisture divergence by advection.
• Over the ocean the transport of moist air
  towards land by about 1 mm/day.
• The transport of moisture towards land is
  consistent with the strong in E over the ocean
  related to the strong in LH flux.

28
Change in moisture transport
Réponse du budget dhumidité
A decomposition of the anomalous advection of
moisture into a dynamical term and a
thermodynamical term shows that the reduction of
moisture transport towards land is mainly driven
by the in q
29
Change in moisture transport
Réponse du budget dhumidité

• This results in cloud cover above land which
  acts as a
• secondary response maintaining the land/sea
  contrast of the
• temperature response.

30
Summary
31
Conclusions
The response to a THC collapse features a strong
zonal gradient of surface air temperature between
the ocean and the continent. Above the ocean,
more low clouds are formed due to an increase in
relative humidity and a more stable marine BL.
These clouds have a cooling effect on the ocean
which enhances the cooling due to the primary
ocean-driven response. A decrease in
convergence of moist static energy above land is
responsible for a decrease of clouds above land.
The effect of this is to warm the atmosphere over
land, thereby weakening the cooling over land
that results from advection of colder and dryer
air from the sea. The secondary cloud response
therefore acts to maintain the strong land/sea
contrast in surface air temperature response.
32
Discussion

• 1. The temperature rise of the last decades is
  larger over land than over sea and a similar
  cloud feedback modulates the temperature response
  to a global warming (Joshi et al. 2008).
• 1. How would Western European climate respond to
  a realistic THC collapse?
• ? Sensitivity experiments with different
  amplitude for the freshwater pulse
• 2. How sensitive is the response of the climate
  to a THC collapse?
• Sensitivity experiments with different climate
  scenarios
• 3. If the THC variability is predictable, can we
  predict the impacts on European climate?
• ? Useful for society

33
Thank you!
34
Temperature response due to a THC collapse