PosterPDF Available

Interdecadal Variability of the Indian Monsoon in an Atmospheric General Circulation Model

Authors:

Abstract

Interdecadal Variability of the Indian Monsoon in an Atmospheric General Circulation Model
Role of Cloud Relaxation Parameter in Inter-annual and Interdecadal
Variability of the Indian Monsoon in an Atmospheric General
Circulation Model
Deepeshkumar Jain, Arindam Chakraborty, Ravi S. Nanjundiah
Deepeshkumar Jain, Arindam Chakraborty, Ravi S. Nanjundiah
Centre for Atmospheric and Oceanic Sciences,
Centre for Atmospheric and Oceanic Sciences,
Indian Institute of Science
Indian Institute of Science
Inter-annual and Interdecadal variability of Indian summer monsoon rainfall (ISMR) is studied in with the National Centers for Environmental Prediction (NCEP)
Inter-annual and Interdecadal variability of Indian summer monsoon rainfall (ISMR) is studied in with the National Centers for Environmental Prediction (NCEP)
seasonal forecast model (SFM) in Atmospheric Model Intercomparision Project (AMIP) style simulations. One of the major processes that control the strength of
seasonal forecast model (SFM) in Atmospheric Model Intercomparision Project (AMIP) style simulations. One of the major processes that control the strength of
monsoons in an AGCM simulation is the parameterization of deep clouds. A factor that governs the effect of deep convection in the model is the cloud relaxation time
monsoons in an AGCM simulation is the parameterization of deep clouds. A factor that governs the effect of deep convection in the model is the cloud relaxation time
scale. Our previous studies have shown that with larger as well as cloud type dependent relaxation time scales, the simulation of mean monsoon rainfall improves. In
scale. Our previous studies have shown that with larger as well as cloud type dependent relaxation time scales, the simulation of mean monsoon rainfall improves. In
the present study we try to understand the role of deep convection on the simulation of interannual and interdecadal variability of the monsoons by varying the cloud
the present study we try to understand the role of deep convection on the simulation of interannual and interdecadal variability of the monsoons by varying the cloud
relaxation time scales in cumulus parameterization. Interconnections between Nino 3.4 SST anomaly, IOD index and ISMR is also studied
relaxation time scales in cumulus parameterization. Interconnections between Nino 3.4 SST anomaly, IOD index and ISMR is also studied
Figure 1. The model simulated
precipitation is overestimated over most
parts of the tropics in the control
experiment. Alpha=0.10 simulates
precipitation more realistically.
Renno and Ingersoll’s (1996) definition of Cloud
adjustment time scale
Interconnectio
ns
Figure 7. Inter-decadal variability of JJAS mean
ISMR (mm/day) over Indian land (70-90 E, 8-28 N)
in IMD data and two model simulations. The two
simulations show opposite sign of anomaly for all the
three decades. Sign of anomaly for 2 decades (1982 to
1990 and 2000 to 2009) is simulated correctly with
alpha=0.10. Control simulation (alpha=0.30) gives the
sign of anomaly correctly only for one decade (1991
to 2000) .
Figure 3. Monthly mean Rainfall over Indian land (8-28
N, 70-90 E) from June to September. Alpha=0.10 gives
correct phase of peak precipitation as compared to
control simulation (alpha=0.30)
Figure 4. Inter-annual variability of JJAS mean ISMR in observation
(IMD data) and two model simulations with cloud relaxation parameter
as 0.30 and 0.10. Also shown is inter annual variability of Nino 3.4
SST anomaly. A strong positive Nino 3.4 (bounded by 120W-170W and
5S-5N) SST anomaly is associated with drought over India.
Figure 5. Interconnections between precipitation over
Equatorial Indian ocean (EIO) and ISMR. The number shown
is the correlation between precipitation over these
regions and ISMR over Indian land. It can be seen that in
observations, the correlation is not very high.
However, alpha=0.10 simulation produces strong coupling
between EIO rainfall and ISMR
Figure 6. IOD index is defined as difference between SST anomaly over east EIO (10S-10N, 50 – 70E, the right box in Fig 5) and that over
west EIO (10 – 0S, 90 – 110 E, the left box in Fig 5). In observations, a clear straight line can be drawn separating drought years from excess
years. This is true with alpha=0.10 simulation as well. We do not see a clear separation between excess years and drought years in control
simulation
alpha=0.3 alpha=0.1
Hit 16 18
Miss 12 10
Figure 8. Skill Scores of two simulations for
28 years.
Hit => When simulated precipitation gets the sign
of precipitation anomaly correct.
Miss => When simulated precipitation gets the
sign of precipitation anomaly wrong.
Conclusions
1. A slower cloud relaxation parameter produces better monthly mean rainfall
during June-September over Indian land.
2. In the 28 year simulation,many years having droughts associated with negative
Nino 3.4 SST anomaly (El-Nino) (such as 2002 and 2009) are simulated well by the
model when alpha=0.10. The year of 1994 and 1997 are simulated with wrong sign
of anomaly in both the cases (alpha=0.10 and alpha=0.30).
3. The correlation between precipitation anomalies over EIO and ISMR is stronger
in simulation when a slower relaxation parameter is used.
4. The observed IOD, Nino 3.4 SST anomaly, and ISMR relationship is strong and
this is simulated well by the model when alpha=0.10. The correlation is not as
strong in control experiment.
5. In terms of inter-decadal simulations also, alpha=0.10 simulates the sign of
anomaly correct for the decade of 1980's and 2000's. The control simulation,
however, simulates the sign of anomaly correct only for the decade of 1990's.
6. The skill scores for 28 years of simulations is also better for alpha=0.10
simulation.
Figure 2. Cloud adjustment time can be imagined as the time
taken by the parcel in the cloud to travel frokm bottom to top of
it. Relaxation parameter in control simulation was 0.30
Model Used -
The Seasonal Forecast Model (SFM) from NCEP used for the present study was
run at T62L28 resolution. The model has 28 unequal vertical sigma levels and a
horizontal resolution of 1.875 degrees. For uniform resolution throughout the globe,
the model uses reduced grid. Chou (1992) short-wave radiation parameterization is
used in the experiment while the long-wave parameterization is from Chou and
Suarez (1994). Planetary boundary layer as parameterized by Hong and Pan (1996) is
used. Cloud fraction is based on Slingo (1987). Mountain induced gravity wave drag
parameterization is by Alpert et al (1988). Land process parameterization by Pan and
Mahrt (1987) is used in the model. Smoothed mean orography is used in the study
and ozone is prescribed using climatology. The Relaxed Arakawa Schubert cumulus
parameterization scheme is based on Moorthi and Suarez (1992). Semi-implicit time
integration is used for model dynamics. Kanamitsu et al (2002) provides detailed
description of the model.
Experimental Details -
Experiment was started by giving an Atmospheric Model Inter-comparison Project
(AMIP) style run starting on 1st January, 1982. The model was integrated for 28
years till 31st December, 2009. Monthly mean sea surface temperature’s (SST’s) are
from Reynolds and Smith (1994) interpolated linearly to the model time step. Initial
conditions are taken from NCEP reanalysis. Diagnostic variables are output as daily
averages (once every day). The focus of the present study is to study the sensitivity
of annual and inter annual variability of the simulated precipitation over India during
the Indian summer monsoon (ISM) season to the choice of relaxation parameter. A
thirty minute model time step was used for the integrations and the output was
saved once a day. The present study include sensitivity studies between two
simulations, one with α=0.10 and the other with α=0.30. We call the case in which
α=0.30 as the control case or the default value of α. α=0.10 refers to cloud
adjustment time of 300 minutes while in the control case cloud adjustment time is
100 minutes.
References - .
1. Kanamitsu M, Kumar A, Juang H, Schemm J, Wang W, Yang F, Hong S, Peng
P, Chen W, Moorthi S, et al (2002) NCEP dynamical seasonal forecast system
2000. Bulletin of the American Meteorological Society 83(7):1019–1037
2. Jain D, Chakraborty A, Nanjundiah R (2012) On the role of cloud adjustment
time scale in simulating precipitation with relaxed arakawa–schubert convection
scheme. Meteorology and Atmospheric Physics pp 1–13
3. Renno, Ingersoll,(1996), Natural convection as a heat engine: A theory for
CAPE. Journal of the Atmospheric Science. Volume 53. Page 572.
ResearchGate has not been able to resolve any citations for this publication.
Poster
Full-text available
We find the role of cumulus parameterization closure assumption on mean Indian monsoon precipitation prediction
Article
Full-text available
routinely since April 2000, and the products are available online at NWS's ftp site.
Article
Full-text available
In this study we look at planetary convection from a large-scale perspective, that is, in statistical equilibrium. We focus on the physical process by which convection adjusts unstable atmospheres. Our hypothesis is that atmospheric convection is a natural heat engine. Our objective is to present a framework usefull for the basic conceptual understanding of the equilibrium state of convecting atmospheres.