Indian Ocean warming – its extent, and impact on the monsoon and marine productivity

Roxy M. K., K. Ritika, A. Modi, P. Terray, R. Murtugudde, K. Ashok, B. N. Goswami, S. Masson, V. Valsala, P. Swapna, S. Prasanna Kumar and M. Ravichandran

Indian Institute of Tropical Meteorology, Pune

WIO

RIO

Indian Ocean warming: o Western Indian Ocean experienced strong,

monotonous warming during the last centuryo Links to asymmetry and skewness in ENSO

forcingo Largest ‘in-phase’ contributor to global SST

warmingPotential Impacts:

o Weakening of the monsoono Reduction in marine phytoplankton

WIO RIO

Indian Ocean warming - backgroundBasin-wide / Warm-pool warming in recent years

Chambers et al. JGR, 1999; Alory et al. GRL, 2007; Rao et al. Climatic Change, 2012; Swapna et al. Climate Dynamics, 2013

Studies note basin-wide warming over Indian Ocean in the last 50 years

Suggested cause: Apart from greenhouse warming,Weakening winds causing warming trendsWarm SST weakening the winds which in turn warm

SST again

Warm-pool enlargement in recent yearsSST trend during last 50

yrs

Indian Ocean during the last centurywestern Indian Ocean warmed up to 1.2degC, in 100 yrs

Roxy et al. J.Climate, 2014

a) Basin-wide warming over the Indian Ocean with enhanced, significant warming over western Indian Ocean.b) The western region has largest interannual variability over Indian Ocean. Warming here might influence monsoon dynamics!c) In early 1900s, the western

Indian Ocean was much cooler than the warm-pool. The monotonous warming over west nullifies zonal SST gradient - may influence monsoon dynamics.

(a) SST trend [°C per year], June-Sept mean

(b) SST standard deviations [°C], June-Sept mean

(c) SST [°C]: WIO vs RIO, June-Sept mean

WIO

RIO

WIO: 50-65°E, 5°S-10°NWIO [CMIP5]RIO: 70-100°E, 20°S-20°N

The historical simulations with CMIP5 models do not reproduce the WIO warming (light pink color), which means that apart from direct radiative forcing due to greenhouse warming, other unaccounted mechanisms might be responsible (eg: ENSO variability)

Mean summer (June-Sept) SST:

SST > 28degC = enhanced convection

Gadgil et al., Nature, 1984; Roxy, Climate Dynamics, 2013

20 21 22 23 24 25 26 27 28 29 30 31

12

10

8

6

4

2

0

SST [°C]

Pre

cip

itatio

n [m

m/d

ay]

Significant increase in precipitation with respect to higher SSTs.

SST-precipitation relationship when the lag is considered

Mean summer (June-Sept) SST:

Asymmetry in ENSO forcingInfluence of El Niño > La Niña

(a) Mean Walker circulation

(b) El Nino Composite - Walker circulation anomalies

(c) La Nina Composite - Walker circulation anomalies

hPa

hPa

hPa

(a) Climatological mean SST [°C]

(b) El Niño Composite - SST anomalies [°C]

(c) La Niña Composite - SST anomalies [°C]

El Niño induce significant easterly anomalies and positive SST anomalies over w.Indian Ocean

but... La Niña events does not result in significant anomalies over the Indian Ocean

Skewness in El Niño forcingIncrease in Frequency and Magnitude of El Niños(a) Correlation: east Pacific SSTa vs Global SSTa, June-Sept mean

(b) SST anomalies [°C]: east Pacific vs WIO, June-Sept mean

1 S.D. = 0.77 °C

r = 0.6

(c) Skewness [east Pacific] and trend [WIO]: SST anomalies, June-Sept mean

east PacificWIO

east PacificWIO

(a) SST Skewness [1901-1950]

(b) SST Skewness [1951-2012]

SST Difference between [1951-2012] and [1901-1950]June-Sept

Detrended anomalies show increase in frequency and strength of El Niños. The warm events over Indian Ocean also has increased. Occasionally, they cross the El Niño criteria (1 S.D. = 0.77 degC).

Indian Ocean warming (above) associated with positive skewness over east Pacific (below)

Indian Ocean warms without greenhouse gas forcingSimulations (with-without) ENSO variability shows IO warming

Nino 3.4 SST anomalies spectra

10-5

10-4

10-3

10-2

10-1

100

Pow

er s

pect

rum

100 10 1Period (year)

0.01 0.10 1.00Frequency(1/year)

HadISST

SST Difference between [ENSOvar] and [noENSOvar]

Model simulations using latest SINTEX coupled model with realistic ENSO variability

Magnitude of warming without greenhouse gas forcing is weak though

Largest contributor to global warming?Indian Ocean warming “in phase” with global warming

Correlation: Annual global mean SSTa vs annual SSTa

Land-sea contrast decrease in the past centuryContradicts model/observations for Northern Hemisphere

Roxy et al. Nature Communications, 2015, Revised

(c) Trend in land-sea temperature difference

(a) mean surf.Temperature °C (b) surf.Temperature trends °C (112 year-1)

Tsurf [°C] Ttrop [°C]

Tsurf

Ttrop

(d) upper troposphere (200 hPa) temperature trends°C (65 year-1)

Though models and observations suggest increase in land-sea contrast over Northern Hemisphere due to global warming, it is different over South Asia/Indian Ocean.

The decrease in land-sea contrast reflects in tropospheric temperature gradients also.Observations suggest an increase in land-sea contrast over Northern Hemisphere during recent decades

Warm Indian Ocean, Weak south Asian MonsoonIndian Ocean warming well correlated with weak Precip.

(a) Trend in IMD Precip. (b) Trend in CRU Precip. mm day-1 (112 year-1)

(c) Correlation: WIO HadISST vs IMD Precip. (d) Correlation: WIO ERSST vs CRU Precip.

New Delhi

Islamabad

Correlation: WIO ERSST vs CRU Precip. [detrended]

New Delhi

Islamabad

(a) & (b) Decreasing trend in precipitation from Pakistan through central India to Bangladesh. Significant over central Indian subcontinent (horse-shoe pattern)

(c) & (d)Trend and correlation with western Indian Ocean warming has similar patterns!

Correlation of detrended anomalies:

Weakened Monsoon precip/winds due to warmingModel simulations with Indian Ocean warming

Precip. and Winds, DIfference between [CFSv2WIO] and [CFSv2CTL]

1 m s-1 (65 year-1)

mm day-1 (65 year-1

Decreasing rainfall over the south Asian subcontinent: horse-shoe pattern in model simulations with increased IO warming

SST difference between [CFSv2WIO] and [CFSv2CTL]

Model simulated warming of WIOModel simulated response to warming

Weakening local Hadley circulation: Convection enhanced over ocean and suppressed over land

Trend in vertical wind velocity, June-Sept (1948-2012) Pa s-1 (65 year-1)

Vertical wind velocity, DIfference between [CFSv2WIO] and [CFSv2CTL] Pa s-1

Observations: trend in vertical velocity

Simulations: change in vertical velocity in response to Indian Ocean warming

Warming – Marine Primary Productionwestern Indian Ocean is a highly productive region...

Reduction in Marine Primary ProductionChlorophyll trends in observations and simulations

Chlorophyll trends

Observations:Merged (SeaWiFS, MODIS, and MERIS)1998 - present

Historical Simulations:Best of CMIP5/MPI-ESM-MR1950-2005

Warming stratifies the ocean - and suppresses the mixing of nutrients from the subsurface, reducing chlorophyll

Chlorophyll

SST

Chlorophyll

SST

Ch

loro

ph

yll (

mg

m-3

)

Ch

loro

ph

yll (

mg

m-3

)

SS

T (°C

)

(b) Chlorophyll and SST anomalies

r = - 0.7r = - 0.9

SS

T (°C

)

(a) Chlorophyll and SST anomalies

Chlorophyll

Wind Speed

Ch

loro

ph

yll (

mg

m-3

) Win

d S

pe

ed

(m s -1)

(d) Chlorophyll and Wind Speed anomalies

r = 0.5

Chlorophyll

Ch

loro

ph

yll (

mg

m-3

)

r = -0.3

(c) Chlorophyll and Wind Speed anomalies

Win

d S

pe

ed

(m s -1)

Wind Speed

(f) SST and Static Stability anomalies

Sta

tic Sta

bility

SS

T (

°C)

SST

Static Stability

r = 0.9

(e) SST and Static Stability anomalies

Static Stability

r = 0.75

Static Stability

Sta

tic Sta

bility

SS

T (

°C)

SST

Chlorophyll

Static Stability

Ch

loro

ph

yll (

mg

m-3

)

(h) Chlorophyll and Static Stability anomalies

r = - 0.62

Sta

tic Sta

bility

Chlorophyll

Static Stability

Ch

loro

ph

yll (

mg

m-3

)

r = - 0.8

(g) Chlorophyll and Static Stability anomalies

Sta

tic Sta

bility

Historical Simulations (1950-2005)Observations (1998-2011)

SST-Chl:significant long-term correlation

Wind-Chl: No significant long-term correlation

Enhanced stratification due to increasing SST

Stratification highly correlated to the reduction in Chl

Reduced plankton might increase the fish stressAlong with the stress from fisheries industries...

Annual catch rates of tuna (N/100 hooks) in the Indian Ocean

tun

a (

yello

wfin

, b

ige

ye a

nd

alb

aco

re)

big

eye

tun

a

bigeye tuna

tuna

Indian Ocean warming – its extent, and impact on the monsoon and marine productivity

Indian Ocean warming: o Western Indian Ocean experienced

strong, monotonous warming during last century

o Links to asymmetry and skewness in ENSO forcing

o Largest ‘in-phase’ contributor to global SST warming

New Delhi

Islamabad

Potential Impacts:o Weakening of the monsoono Reduction of marine

phytoplankton

WIO

RIO