skill of coupled model for indian monsoon: seamless …mjo-induced anomalies of ppt. and winds are...

PROWESS, 17-19 Feb 2014, NCMRWF, India

Skill of Coupled Model for Indian Monsoon: Seamless Assessment

Ashis K. Mitra

Imran M. M., D.K.Mahapatra, A.Gera, S. Prakash, P.P.Saheed, E.N.Rajagopal

NCMRWF, ESSO/MoES, India e-mail: [email protected]

Collaborators: UKMO

Alberto Arribas, Sean F. Milton, Gill Martin and Matthew Martin


Outline

Background: Need for coupled Modelling system at NCMRWF

(Days to Season - Seamless Model Development/Verification)

Coupled Model configuration at NCMRWF: NCUM-C

Mean Monsoon & Biases: UKMO and CFS coupled Models

(Atmosphere & Ocean Components)

CFS: Pentad Rainfall Biases , few weeks from CFS hindcast data

Monsoon 2013: Extended range fcst , few weeks

Summary and way ahead

Moving towards a Unified Modelling Framework

Artificial boundaries between meso-scale short-range prediction,

synoptic scale medium range prediction, and monthly/seasonal prediction

have no scientific basis

Due to practical considerations of computing and of model complexity different

prediction systems for different time scales were being practiced

Simulation and prediction of meso-scale systems, synoptic scale disturbances,

intra-seasonal, seasonal and inter-annual variations are linked

(Error comes in during first few days and gets modulated)

World Modelling Summit (2008), adopted an Unified Modeling Framework

(Seamless Prediction) as one of the important consensus theme

In annual European Geophysical Union's meet of 2009 ECMWF organised a

special session on 'Seamless Prediction' approach

Dec 2010: Joint WCRP, WWRP, THORPEX -> days to season seamless prediction

During active spell of MJO, pptn. and low-level circulation fcst. skill in tropics

found to be systematically reduced in MRF, as compared to quiescent times.

MJO-induced anomalies of ppt. and winds are systematically forecast

weaken and even retrograde.

By day-7 the anomalies were not seen in forecast

MERSEA Project of ECMWF

Medium Range to Seasonal Forecast from ECMWF Coupled Model

Up to 10 days beyond 10 days to a season

T 799L91 T399L62 ( 15 members Ens)

25 km 50 km

0.25 deg Ocean L50 0.25 deg Ocean L50

Findings:

Tropical Systems show much improved simulations

------------------------------------------------------------------------------------------------------------------

Operational VarEPS + Monthly: up to 32 days (Merged MRF and Monthly System)

EPS resolution T639L62 (32 km) for the first 10 days, then T319L62 (65 km) coupled

(Ocean resolution is 1 deg in zonal, at eq. 1/3 deg, L 29)

High Resolution Coupled Prediction System, include ocean meso-scale

Coupled atmosphere, ocean state, ocean waves

BMRC Australia - Project

ECMWF Project Latest studies at ECMWF: an interactive upper-ocean (high

resolution) coupled to AGCM significantly enhance the

prediction skill of weather in medium range and extended

range ( up to a month/season) for Monsoon

Klingaman, UReading

Summary

Due to scale interactions, study of monsoon variability and predictability from

hours to a season using a realistic model is of vital importance.

Under the MoES National Monsoon Mission, NCMRWF is focussing on the

week-2 forecast of monsoon rainfall using a coupled model and its

initialisation.

Improving the skill of models within the time-scale of first two weeks period is

vital because of its seamless connection to the monsoon intra-seasonal and

seasonal variability and prediction.

Model development for monsoon rainfall has to be dealt in a holistic way

including the scales from Days to a Season (Short-term Climate).

Kamal Puri, 2012: Aust. Met. & Ocean Journal , ‘ ACCESS: Scientific Justification and options for

System Development ‘

Mitra et al., 2013, Current Science, 104(10), 1369-1379.

S2S: Sub-Seasonal to Seasonal Prediction Project

(WMO Coordinated)

http://www.google.co.in/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&cad=rja&docid=fqlRfUNStYPYBM&tbnid=G4QCw08304iE-M:&ved=0CAUQjRw&url=http%3A%2F%2Feh.lenin.ru%2Fenglish%2F7or%2Fun%2Funo-sp.htm&ei=zREDU6zKNonYrQeo7oG4Aw&bvm=bv.61535280,d.bmk&psig=AFQjCNERMDYFyYW9XybqHis8AQBkCFNq5w&ust=1392796425892348

A fully atmosphere-ocean coupled model is being implemented at NCMRWF

for advanced research on impact of coupling on medium, 2-weeks, 1-month,

1-season time scales [ Focus in week-2 in Monsoon Mission ]

MoES/NCMRWF has an MoU with UKMO

NCMRWF/MoES and UKMO is working together to further improve the

coupled model for Seamless prediction framework particularly for monsoon

This has serious positive implications on improving the monsoon simulation

for scales from days to a season including week-2 and monthly forecast.

UKMO HadGEM3AO based Global Coupled Model On IBMP6 machine at NCMRWF

Atmosphere : UM 1.875 x 1.25 horizontal

L85 (going up to 85 km)

50 levels are below 18 km

Ocean : NEMO 1 x 1 (Near equator 0.33 )

L75 (1 m near surface;

35 layers in upper 300 mts)

Sea-Ice : CICE Model of Los Almos Lab, USA

OASIS Coupler (CERFACS and collaborators)

Hindcasts: May initial conditions (from MO)

Initialised atmos., ocean, sea-ice

HadGEM3AO: Range of specific model configurations incorporating

different levels of complexity but with a common physical framework

Well resolved stratosphere in UM atmos (around 85 km , hi-top system)

Better stratosphere-troposphere interaction

Sea-ice and monsoon variability

Climate forcings (aerosols, methane, CO2 concentrations, etc) are set to

observed values for the period 1960-2000 and follow the emissions

scenario A1B afterwards.

Ozone is fixed to observed climatological values and includes a

seasonal cycle.


UM Atmos

Monsoon Seasonal Skill

UKMO: [N96L85O1L75]

CFSV2: T126L64OL40

Hind-cast for Monsoon

Observed data: CMAP rainfall , ERA Interim for other parameters

Valid period is JJA (monsoon): May start date


Monsoon ISV in Coupled Model


Northward Movement of Rainfall Isochrones onset/Progress Phase of Monsoon

T.N.Krishnamurti, D.R.Sikka and others , 2012, JAS

Surface Position of ITCZ

Northward Movement of Rainfall Isochrones onset/Progress Phase of Monsoon

T.N.Krishnamurti, D.R.Sikka and others , 2012, JAS

Movement of Isochrones of Rainfall In Coupled Model During Onset/Progress of Monsoon

Indian Land Region Western India

Central Indian Region Bay of Bengal Region


Simulation of Mean Indian Ocean Features in Coupled Model

WOCE 2009

1 x 1

1955-2006

OSCAR

Scatterometer

Wind

SSH altimeter

1 x 1

1993-2003

Drifter Ind Ocn.

Buoy

Sat. Tracked

2 x 2; 15m

1976-1998

NIO, J. Mar. Sc

OSCAR

Scatterometer

Wind

SSH altimeter

1 x 1

1993-2003

OSCAR

Scatteromet

er

Wind

SSH

altimeter

1 x 1

1993-2003

Geo. Curr

Drifter Ind Ocn.

Buoy

Sat. Tracked

2 x 2; 15m

1976-1998

NIO, J. Mar. Sc

Compare with INCOIS ARGO data

2002-2012 mean

Coupled NWP & Skill in Extended Range: Few Weeks

CFSV2 Hindcast: 1998 to 2009

GloSea5 : 2013 onset/progress case


-1

-0.75

-0.5

-0.25

0

0.25

0.5

0.75

1

JUN JUL AUG

Months

Anomaly CorrelationCentral India [73-90o E; 22-28o N]

PA1

PA2

PA3

PA4

-1

-0.75

-0.5

-0.25

0

0.25

0.5

0.75

1

JUN JUL AUG

Months

Anomaly CorrelationAll India [67-100o E; 7-37o N]

PA1

PA2

PA3

PA4

-1

-0.75

-0.5

-0.25

0

0.25

0.5

0.75

1

JUN JUL AUG

Months

Anomaly CorrelationPeninsular India [74-85o E; 7-12o N]

PA1

PA2

PA3

PA4

-1

-0.75

-0.5

-0.25

0

0.25

0.5

0.75

1

JUN JUL AUG

Months

Anomaly CorrelationWest Coast of India [70-78o E; 10-20o N]

PA1

PA2

PA3

PA4

(a) (b)

(c) (d)


Start date:

Dry or Subdued

Start date:

Wet

26May IC: Week – 1 : 27 May to 2 June , [ mm/day ]

Example

from a high

resolution

Coupled

Model

HadGEM3AO

N216

26May2013 IC: Week – 2 : 3 – 9 June, [ mm/day ]

26 May 2013 IC: Week – 3 & 4 : 10 – 23 June, [ mm/day ]

June 2013 Rainfall Forecast : Rainfall Anom. [ mm/day ]

26 May 2013 IC

Ocean Data in Coupled Model (Assimilation and Verification)

( Through INCOIS of MoES ) in NEMOVar

ARGO Floats: T and S profiles up to 2000 mts

Moored Buoys: Surface Met. Ocean Parameters

Drifters: Surface Met. Ocean Parameters

XBT Lines: T Profile in shipping Routes

Current Meter Array: Ocean currents at various depths

Tide Gauges : Sea level

Research Vessels : AWS, Upper Ocean parameters , XBT, ADCP

Satellite Data: Sea Surface heights, Sec Ice elevation (Indian Satellite SARAL)

SST, Ocean surface wind Vectors (Indian Oceansat II)

Sea Ice Cover (MW), Surface wind speed, Fluxes


Conclusions & Way Ahead

NCMRWF to undertake advanced research in tropical/monsoon

simulation/prediction across a range of time-scales from hours to a season has

made one incremental step by implementing a state-of-the-art coupled model

A course resolution of the coupled model hindcast data were analysed;

The model has reasonable mean monsoon rainfall and its variability;

Needs improvement

NEMO ocean model is able to capture the mean Indian Ocean features

realistically in the coupled set-up. Equator and BoB region biases.

JULES LS model and LDA are being implemented

High resolution coupled model show some skill in monsoon pptn. Prediction in

extended range up to few weeks; needs more work.

To remain at the forefront of monsoon research/prediction, a higher resolution

version of the coupled model and its initialisation component will be

implemented on a more powerful super computer.

The focus will be to improve simulation of coupled monsoon variability in time-

scale of hours to a season with an aim to improve forecast skill in week-2 time

period over Indian region.


Thank You

skill of coupled model for indian monsoon: seamless …mjo-induced anomalies of ppt. and winds are...

Documents