Weather, Climate and the OCEANS

B. N. Goswami

Indian Institute of Tropical Meteorology, Pune

3rd OSICON, 26-28 Nov, 2013, IITM, Pune

The OCEANS and Indian Monsoon

Climate Variability

The Sun-Earth System

Climate :

Incoming solar heat – radiated loss of heat from earth-atmos.

Broadly three factors influence

Net heat flux at the top of the atmosphere is Positive

over Tropics and Negative over the polar region

-100100 Wm-2

Top-of-atmosphere net (solar minus Earth longwave)

radiative flux

Atmospheric general circulation: moving energy from

near the equator toward the poles…

Circulation is setup due to pressure gradients under the Coriolis force

Transports heat from tropics to the polar region

Large north-south temperature gradient also gives rise to some unstable planetary waves. They also transport heat pole ward. From an initial ONE cell, a THREE cell meridional structure emerges

Dynamic equilibrium is maintained.

How is equilibrium of temperature maintained?

Cold polar region-High

pressure

Hot tropics-Low pressure

Cold polar region-High

pressure

Required Heat Transport

2/

1

cos2

dFrTTTAOA

Atmospheric transport

Oceanic transport

The net heat balance at the TOA also indicates that, for the earth’s climate to be in equilibrium, there must be mechanisms in place that continously transports heat from equatorial regions to the polar regions.

FTA

Transport of energy required by observed heat balance

The thermohaline circulation of the world ocean

External forcing, namely solar forcing has long term variations,

In time scales of solar forcing has oscillations in

21 thousand years

43 thousand years (Milankovich cycles)

100 thousand years

But no significant short tern variations!

However, there are significant short-term climate variability

Ex. ENSO – approx 4 years : TBO -- aprox 2 years

PDO -- 15-25 years & 50-60 years

AMO and Indian monsoon – aprox 60 years

What is responsible for these short term climate variability?

Ocean and Atmosphere interact to produce Climate Variability on a variety of time scales!

ENSO

IODM

To Understand these interactions, we must understand;

How Ocean forces Atmospheric motion

How Atmosphere forces Changes in SST distribution in Ocean

How does Ocean forces Atmospheric Motion ?

Atmosphere feels the

Ocean through SST

SST modulates

moisture supply

through evaporation

Moisture supply

modulates Atmos.

Heating Distr.

Atmosphere Heating

forces surface winds

How does Atmosphere forces Changes in SST ?

Atmospheric surface

winds forces Ocean

currents

Upper ocean currents

redistribute water and

influence SST distr.

Atmospheric

convection and winds

influence Qnet at

surface

How does atmosphere and ocean interact in the tropics?

Changes in SST Ts

Changes in evaporation Es

Changes in atmospheric heating Q

Changes in atmospheric circulation C (surface stress)

Surface stress drives ocean currents

Where is this Air-sea Interactions most Effective? TROPICS

For this,

Small Change in SST large change in moisture

availability High mean SST in tropics makes it possible. Low SST in middle lat makes it less effective

Available moisture should result in latent heat

release Conditionally unstable thermal structure makes it possible in the tropics

Ocean-Atmosphere Interaction also produce

one Multi-decadal Mode of Indian Monsoon

Variability!

Another Example…

JJAS All India Rainfall (AIR)

Interannual Variability

Mean : 86 cm; S.D. : 8.5 cm

No long term increasing trend

A 50-80 year multi-decadal

variability ?

Decreasing trend in the last 5

decades!

Climatological mean

JJAS Precipitation

Decreasing trend of AIR

between 1941-2012

A Big Question!

What is responsible for the recent decreasing trend ISMR?

Is this trend forced by anthropogenic forces (GHG, aerosols etc) or is part of a natural multi-decadal oscillation?

Bollasina, Ming and Ramaswamy, 2011, Science, 334, 502

Claims that this trend is forced by anthropogenic aerosols based on

A series of experiments with a coupled model with active aerosols!

However, the model produces strong decreasing trend in south

China where observed trend is increasing! Hence, the model

trend over Indian region could not be trusted!

Atlantic mutlidecadal variability (AMO) and Indian monsoon

SST

Goswami et al. 2006, GRL, vol.33, L02706

+ (-) AMO phases of NA SST

Persistent increase (decrease) Meridional gradient of TT over monsoon region.

Strong (weak) Indian summer monsoon

Higher frequency of Strong + (-) NAO events

Changes in the Jet stream and storm tracks

Persistent + (-) TT anomaly over N. India and S. Eurasia

How does AMO modulates South Asian Monsoon?

11-year running mean of AIR and an AMO index

EQIO SST is always out of phase with AIR, strongly so in recent

decades.

Interdecadal mode of AIR and Eq. IO SST extracted using

singular spectrum analysis (SSA)

Trend in JJAS SST

Onset

Withdraw

LRS

Scatter plot of LRS AIR and TISM and resultant Correlation.

TISM : Integral of positive gradient of TT

Xavier, Marzin and Goswami , 2007, QJRMS

EQIO SST can also influence AIR through TT gradient…

Trend of TT in

the Northern box

Trend of TT in

the South box

(black) and trend

of EQIO SSTA

Trend of TT in

the Northern box

(solid) compared

to that in the

south box

(dashed)

Increasing Trend of

EQIO SST

Strong Increasing

Trend of EQIO TT

Decreasing Trend of

∆TT, as TT in the

northern Indian box has

weaker increasing trend

Decreasing Trend of

TISMAIR

Increasing trend NA SST in recent decades

increasing TT in the north box

Increasing trend of Eastern Equatorial Pacific SST

(ENSO) decreasing trend of TT in the north box

AGCM experiments:

CTL AGCM forced by global

SST with increasing trend, (1980-

2011)

IO trend removed AGCM forced

AGCM

experiments forced

by SST trends

globally and trends

removed from IO

SST shows that,

The trend in the

NB is weak while

the increasing

trend in the SB is

much stronger, as

in the

observations.

Thus, the large

scale air-sea

interactions make

the EQIO SST

Swapna, Krishnan and Wallace, 2013, Clim.

Dyn.

What is responsible for increasing trend of IO SSTA?

IOSST increase weakens ISM further increases IOSSTweak ISM

Thus, the recent increasing trend of IOSST is result of air-sea interaction!

Based on these analyses, we propose that the recent decreasing trend of AIR is due to an air-sea interaction involving IO SST, Pacific SST as well as NA SST.

Could the recent decreasing trend of AIR be part a natural mode of variability?

Is there a preferred periodicity of the Multi-decadal Variability of the Indian Monsoon?

India Tree-ring

Thailand Tree-ringSpeleothem δ 18O

India Tree-ring : (A.D. 1481-2003); Palaeo-3, Borgaonkar et al. 2010

Thailand Tree-ring (A. D. 1558-2005); Clim.Dyn, Buckley et al. 2007

Speleothem δ 18O: (A.D. 652-2007); GRL, Sinha et al, 2011 (Dandak and

Jhumar caves)

Asian Monsoon Proxies

A - RWI-India (1481-2003)

B - δ18O-CI (625-2007)

C - RWI-Thailand (1558-2005)

Red line : 21-year moving average

With Borgaonkar,

Kriplani and Preethi,

Unpublished!

Multi-decadal Periodicities of Asian Monsoon

RWI-India

δ18O-CI

RWI-Thailand

Can this Oscillation be considered a Mode of variability?

Empirical Mode Decomposition (EMD) Decomposes the data time series into finite number of

Intrinsic Mode Functions (IMFs) each associated with a unique frequency

An IMF has the following properties:

1. In the whole data set, the number of extrema and the number of zero-crossings must be

either equal or differ at most by one;

2. At any time point, the mean value of the “upper envelope” (defined by the local

Maxima) and the “lower envelope” (defined by the local minima) must be zero.

3. The EMD method avoids spurious harmonics and the components of the EMD are usually

physically meaningful.

The EMD is implemented through following steps

(sifting process)

1. Identify all extrema of x(t)

2. Interpolate the local maxima to form an upper

envelope u(x)

3. Interpolate the local minima to form an lower

envelope l(x)

4. Calculate the mean envelope: m(t) =

(u(x)+l(x))/2

5. Extract the mean from the signal: h(t)=x(t)-m(t)

6. Check whether h(t) satisfies the IMF stoppage

criteria,

YES: h(t) is an IMF, stop sifting

NO: let x(t)=h(t), keep sifting

Ref : Huang et al. (1998)

Intrinsic Mode

Functions

(IMFs)

of

Asian Monsoon

Proxies

δ18O-CIRWI-India RWI-Thailand

Periodicities

of

IMFs of Asian Monsoon

Proxies

Blue line : 95% Confidence level

Green line : 5% Confidence level

RWI-Thailandδ18O-CIRWI-India

Staistical Significance of IMFs

δ18O-CI

RWI-India

RWI-Thailand

Blue line : 95% Confidence level

Green line : 5% Confidence level

Fig. The spread function. The groups of the dots

from upper left to the lower right are the energy

density as a function of the spectrum weighted

mean period of IMFs 1-9 for 1024 samples with

1024 data points. The black solid line is the

theoretical line. Black dots correspond for the pairs

of averaged mean energy density and averaged

extrema-counting mean period. Dashed lines

represent first and 95th percentiles.

Wu and Huang (2005).

Intrinsic Mode Functions (IMFs)

Global SST Proxies of AMO, ENSO and PDO

AMO ENSO PDO

Mann et al. 2009,

Science

Periodicities ofof

IMFs of Global SST Proxies

Staistical Significance of IMFs

Blue line : 95% Confidence level

Green line : 5% Confidence level

AMO ENSO PDO

AMO

ENSO

PDO

Coherence between

Multi-decadal Mode

of

Asian Monsoon and Global

SSTs

IMF 5 of RWI-India and IMF 2 of AMO

IMF 5 of RWI-India and IMF 3 of ENSO

IMF 5 of RWI-India and IMF 2 of PDO

HadGEM2-AO

MPI-ESM

GFDL-CM3

Simulation of

Indian monsoon

rainfall (JJAS) by

three coupled

climate model

participating in

CMIP5 under two

RCP scenarios

each.

HadGEM2-AO MPI-ESM GFDL-CM3

Wavelet spectra detrended Indian monsoon rainfall index for the

three coupled climate models for (top) natural + RCP4.5 and

(bottom) natural + RCP8.5

HadGEM2-AO MPI-ESM GFDL-CM3

Wavelet spectra detrended AMO index for the three coupled climate

models for (top) natural + RCP4.5 and (bottom) natural + RCP8.5

The Good News… The weakening trend of the Indian monsoon during

the past five decades is driven by the warming trend

of the equatorial Indian Ocean (EQIO) SST.

The discovery that Indian summer monsoon has a 50-

80 year multi-decadal Mode of variability as an

integral part of a global coupled ocean-atmosphere

50-80 year multi-decadal Mode of variability, indicates

that the current decreasing trend of AIR is part of this

natural multi-decadal variability .

The increasing trend of IO SST may be part of a multi-

decadal global coupled ocean-atmosphere mode.

This means that the decreasing trend may recover

within the next couple of decades!

The Challenge…

The existence of the 50-80 year mode of variability indicates that there will some decadal predictability of AIR. However, due to the broad band nature of the mode, the predictability will be limited.

The current climate models to provide reliable projection of monsoon climate requires,They need to simulate the 50-80 year multi-decadal mode correctly.

Even on seasonal time scale some predictability comes from extra-tropical SST. The climate models need to explore this earnestly.

Thank you