large-scale orography and monsoon akio kitoh
Post on 12-Feb-2016
40 Views
Preview:
DESCRIPTION
TRANSCRIPT
Large-scale orography and monsoon
Akio KITOH Meteorological Research Institute, Japan Meteorological Agency
1: Introduction
2: Surface temperature change
3: Asian monsoon
4: El Niño/Southern Oscillation (ENSO)
Kutzbach et al. (1993) J.Geology
Effects of mountains on climate
Ruddiman and Kutzbach (1989) JGR
Broccoli and Manabe (1992)
Arid and Semiarid Climatemountain
observed
no mountain
Broccoli and Manabe (1992)
soil moisture precipitation
M
NM
Eurasia is drier in M than in NM
Broccoli and Manabe (1992)
M
NM
transient eddy moisture flux
Larger eddy activity and larger moisture flux over Northern Eurasia in NM
Tibetan Plateau uplift
Ramstein et al. (1997) Nature
Kutzbach et al. (1993) J.Geology
4 types of large-scale forcing or b.c. for the South Asian monsoon
the monsoon is most sensitive to the elevation and radiation (orbital) changes
CCM1+50m mixed-layer
GCM Study on mountain and monsoon
#AGCM perpetual July Hahn and Manabe 1975: Jul GFDL 270km L11 Kutzbach et al. 1989: Jan/Jul CCM R15 L9
#AGCM seasonal cycle Broccoli and Manabe 1992: GFDL R30 L9
NH midlatitude dry climates An et al. 2001: NCAR CCM3
4 stage Himalayan uplift Liu and Yin 2002: COLA AGCM
11cases: 0%, 10%, …, 100%
#AGCM + slab ocean Kutzbach et al. 1993: CCM1 R15 L12 + 50m slab ocean Kitoh 1997: MRI-II 4x5 L15 + 50m slab ocean
#AOGCM Kitoh 2002, Abe et al. 2003: MRI-CGCM1 (4x5) Effect of SST change Kitoh 2004: MRI-CGCM2 (T42) 0% to 140%
・ Exp-M control ・ CGCM coupled GCM・ Exp-NM no mountain ・ SGCM slab-ocean
・ AGCM
Model topography in the control run
Effect of Large-Scale Mountains on Surface Climate
Variance northward of 20N are 3,800 (M), 1,600 (NM) and 2,200 m2 (M-NM). Thus, the land-sea distribution effect (NM) explains about 40%, and the mountain effect (M-NM) explains about 60% of the total variance.
Stationary eddies at 500 hPa in January
200 hPa Winds
January: The Asian subtropical jet M is 15 m s-1 stronger. But zonal mean zonal wind at 30N is the same. July: The subtropical jet in NM stays at 30N.
Surface Winds
Note the difference in trade winds both in Jan and Jul, and different wind direction over the Arabian Sea in July.
Precipitation
An overall precipitation pattern is similar. > land-sea configuration and SST distribution are the main factors. NM summertime Asian precipitation elongates along 10N belt. M has less precipitation over Eurasia. Shape of ITCZ.
Sea-level pressure
January: Shape of the Siberan high. July: strong Pacific subtropical anticyclone in M
non-adjusted
adjusted for 6.5 K/km
Annual mean surface air temperature difference
+ inland area - coastal area / ocean
Large negative temperature change over mountains. < elevation effect
SST also changes.
South Asia and Eastern Asia: precipitation-soil moisture-evaporation, precipitation-cloudiness-insolation Continental interior: precipitable water and moisture flux convergence are less, dry ground, less cloud
Summary (Land surface temperature)Orography induces a warmer continental interior and colder coastal area over land. The land surface temperature drops due to the lapse-rate effect. When this effect is eliminated, the continent interior becomes warmer with a mountain uplift, because clouds become fewer and the surface drier due to a decreased moisture transport. On the other hand, South Asia becomes cooler because the summer monsoon is stronger, and heavier precipitation makes the land surface wetter and increases the clouds.
Summary (SST)
The SST decreases due to orography particularly over the subtropical eastern oceans. This occurs because less solar radiation reaches the surface due to more low-level clouds that are induced by a strong subtropical anticyclone.
Changes of Asian monsoon by uplift
All mountains are varied uniformly between 0% and 140%.Land-sea distribution is the same for all experiments. MRI-CGCM2. No flux adjustment.
0 10 20 30 40 50year
M14 (140%)
M12 (120%)
M10 (control)
M8 (80%)
M6 (60%)
M4 (40%)
M2 (20%)
M0 (no mountain)
Experiments
MRI CGCM2•AGCMAGCM
–MRI/JMA98MRI/JMA98–T42 (2.8x2.8), L30 (top at 0.4 hPa)T42 (2.8x2.8), L30 (top at 0.4 hPa)–Longwave radiation - Shibata and Aoki (1989)Longwave radiation - Shibata and Aoki (1989)–Shortwave radiation - Shibata and Uchiyama (1992)Shortwave radiation - Shibata and Uchiyama (1992)–Cumulus - Prognostic Arakawa-Schubert typeCumulus - Prognostic Arakawa-Schubert type–PBL - Mellor and Yamada level 2 (1974)PBL - Mellor and Yamada level 2 (1974)–Land Surface - L3SiB or MRI/JMA_SiBLand Surface - L3SiB or MRI/JMA_SiB
•OGCMOGCM–Resolution : 2.5x(0.5-2.0), 23layersResolution : 2.5x(0.5-2.0), 23layers–Eddy mixing : Isopycnal mixing, GMEddy mixing : Isopycnal mixing, GM–Seaice : Mellor and Kantha (1989)Seaice : Mellor and Kantha (1989)
•CouplingCoupling–Time interval : 24hoursTime interval : 24hours–Flux adjustment: “without” in this experimentFlux adjustment: “without” in this experiment
120E-140E pentad precipitation obs
M4 0.75
M0 0.71
M8 0.81
M12 0.74
M2 0.74
M10 0.79
M6 0.79
M14 0.66
Numbers indicate spatial cc with obs
50N
10S
100% OBS
0%
20
40
60
80
120
140
Taylor’s diagram
Note the difference in the Pacific warm pool.
Over the Indian Ocean, SST gradient reverses.
What is the merit of using CGCM?
AGCM: only dynamical/thermodynamical effect of mountain
CGCM: air-sea interaction, effect of SST change
Additional AGCM experiments were performed with the same experimental design
A0, A2, A4, A6, A8, A10, A12, A14
Comparison between CGCM and AGCM experiments
Precipitation Precipitable waterCGCM
AGCM
C-A
Rainfall Index IMR: India, land
10N-30N, 60E-100E
SEAM: Southeast Asia
5N-25N, 100E-130E
EAM: East Asia
25N-35N, 120E-140E
CGCM
AGCM
CGCM
AGCM
CGCM
AGCM
Koppen climate: Asia
Koppen climate: India
• “BW” “BS” “Aw” as precip increases
• “BS” in the interior part of southern peninsular India does not appear in the model due to coarse resolution
0% 100%
OBS
Koppen climate: China
• “BW” “BS” dominates in 0% 〜 40% cases; too dry
• “Cw” “Cf” appears from 60% case as precip increases
• “Cs” appears in 80% 〜 120% cases due to larger winter precip
OBS
100%0%
Summary (Monsoon)• Systematic changes in precipitation pattern a
nd circulation fields as well as SST appeared with progressive mountain uplift.
• In the summertime, precipitation area moved inland of Asian continent with mountain uplift, while the Pacific subtropical anticyclone and associated trade winds became stronger.
• The model has reproduced a reasonable Baiu rain band at the 60% case and higher.
• CGCM results were different from AGCM’s: CGCM showed a larger sensitivity to mountain uplift than AGCM.
Changes of ENSO by mountain uplift
Control run: global SST EOF1 and regressions
No-mountain run: global SST EOF1 and regressions
NINO3.4 SST and SOI
→ lower mountain cases have larger amplitude
m0
m14
m12
m10
m8
m6
m4
m2
In M0, the SST pattern is nearly symmetric about the equator.
The spatial pattern (e.g., meridional width) changes with uplift.
In M0, frequency peak is at 7 yr. When mountain becomes higher, it shifts toward high frequency, and explained variance smaller.
Power spectra of each leading mode of SST EOF
33.6%
16.9%
17.5%
18.5%
18.1%
29.5%
25.6%
25.6%
6 4 2 yr
Pacific trade winds become stronger associated with strengthened subtropical high with mountain uplift
Change in Mean Climate: Trade Winds
→ Easterlies in lower mountain cases are strong in the eastern Pacific, but weak in the western Pacific
low mountain
high mountain
Change in Mean Climate:
Upper Ocean Heat Content and its zonal gradient
→ lower mountain cases have larger OHC gradient
low mountain
high mountain
Summary (ENSO)Systematic changes in SST and ENSO as well as precipitation pattern and circulation fields appeared with progressive mountain uplift.
When the mountain height is low, a warm pool is located over the central Pacific; it shifts westward with mountain uplift.
Model El Nino is strong, frequency is long and most periodic in the no mountain run. They become weaker, shorter and less periodic when the mountain height increases.
As mountain height increases, the trade winds intensify and the location of the maximum SST variability shifts westwards.
Smaller amplitude of El Nino with high mountain cases may be related to smaller SST/OHC gradient in the central Pacific.
Short return period of El Nino may be associated with a westward displacement of most variable SST longitude and a decrease in the meridional width.
top related