WW

RP

Sub-seasonal to seasonal prediction

David Burridge (THORPEX IPO)and a cast of …….

An element of the WWRPAn element of the WWRP

WW

RP

Sub-seasonal to Seasonal Prediction – “….the Commission requested the JSCs of the WWRP and the WCRP and also the THORPEX ICSC to set up an appropriate collaborative structure to carry out an international research initiative on sub-seasonal to seasonal forecasting. It recommended that this is closely coordinated with the present existing CBS infrastructure for long-range forecasting (with centres producing long-range forecasts and regional climate centres) and with the future developments in WMO climate service delivery and the Global Framework for Climate Services called for in the High-Level Declaration of WCC-3.”

WMO CAS XV Resolution at

Incheon, South Korea, November 2009

WW

RP

Collaboration between the Weather and Climate Communities to Advance

Sub-Seasonal to Seasonal Prediction

To achieve progress in long-range prediction, the coordination of research is needed in: multi-model ensemble prediction system, tropical convection, and its two-way interaction with the global circulation, data assimilation and its socioeconomic applications.

Gilbert Brunet, Melvyn Shapiro, Brian Hoskins, Mitch Moncrieff, Randal Dole, George N. Kiladis, Ben Kirtman, Andrew Lorenc, Brian Mills, Rebecca Morss, Saroja Polavarapu, David Rogers, John Schaake and Jagadish Shukla.

WW

RP

First StepWorkshop

“Sub-seasonal to Seasonal Prediction” Met Office, Exeter –1 to 3 December 2010

The main goals of this Workshop were to establish current capabilities in sub-seasonal to seasonal prediction, to identify high-priority research topics and demonstration projects and to develop recommendations

for the establishment of an international research project.

An element of the WWRPAn element of the WWRP

WW

RP

Operational Prediction Systems

Medium-range weather predictions (~10-15 days)Medium-range weather predictions (~10-15 days)

Monthly or extended-range predictions (~30-45 days)Monthly or extended-range predictions (~30-45 days)

Seasonal predictions (~12 months)Seasonal predictions (~12 months)

WW

RP

WMO PTC Meeting, 7-9 February 2011, Geneva.

MJO connection to Canadian surface air temperature: high-impact weather?

Lagged winter (DJF) SAT anomaly in Canada for 1979-2004

Significant warm anomaly in central and eastern Canada 1-2 pentads after MJO phase 3

Acknowledgment to Hai Lin

WW

RP

WW

RP

Social and Economic Utilization of Sub-Seasonal and Seasonal Predictions

A need for closer ties between weather and climate research:Understanding how information at the

weather/climate interface, including uncertainty, connects with decision-making

There is also a great need for much easier access to forecast data by the user community. These need to be available in special user-oriented products. How could we achieve this service?

The post-processing techniques that are needed by many users may require an archive of past forecasts (e.g. for water cycle applications). Some user applications require an archive of re-forecasts from fixed models for periods as long as 20 years or more.

WW

RPSouth American Demonstration Project – Celeste Saulo (CIMA)

Improve predictability by understanding summertime precipitation

La Plata Basin –Forecast Demonstration ProjectVariability induced by SACZ / MCSs / blocking

events /ENSO / SSTs / Soil Moisture2003 Salado Flood – initial flooding event ~24 days /

secondary event ~2.5 monthsOutputs – probabilistic predictions of precipitation /

documentation of scientific findings / support from numerical modelling centres in Brazil & Argentina for forecasts & leveraging existing systems & collaborations Ongoing research

WW

RP

The predictability of MCSs in continental regions is a global issue

Biomass burning as a consideration within the models - would the inclusion of other environmental parameters be beneficial to improve forecasts?

Inclusion of hydrological modelling to generate flood forecasts

Landslide risk modelling

South American Demonstration Project – considerations

WW

RP

South Asian Demonstration Project – Violeta Toma (GIT)

Implemented 15 day ensemble TC forecasting system Intensity of TCs (Nargis) not fully captured by models

Flood Forecasting in Bangladesh – 3 tiered systemExperimental real-time 1-10 day flood forecasts Model lead time & accuracy not the only issue with regard to preventing

losses

Importance of understanding the user community for forecast dissemination

Communicating uncertainty – what the warnings mean to the user community

Experimental look at Pakistan floods!

WW

RP

Floods in Pakistan

10°N10°N

20°N 20°N

30°N30°N

40°N 40°N

40°E

40°E 60°E

60°E 80°E

80°E

10°N10°N

20°N 20°N

30°N30°N

40°N 40°N

40°E

40°E 60°E

60°E 80°E

80°E

Contours at 1% levelShaded areas significant at 10% levelensemble size = 51 ,climate size = 90

Verification period: 26-07-2010/TO/01-08-2010Precipitation anomaly

Analysis and ECMWF VarEPS-Monthly Forecasting System

10°N10°N

20°N 20°N

30°N30°N

40°N 40°N

40°E

40°E 60°E

60°E 80°E

80°E

10°N10°N

20°N 20°N

30°N30°N

40°N 40°N

40°E

40°E 60°E

60°E 80°E

80°E

10°N10°N

20°N 20°N

30°N30°N

40°N 40°N

40°E

40°E 60°E

60°E 80°E

80°E

ANALYSIS

FORECAST 22-07-2010: DAY 5-11 FORECAST 15-07-2010: DAY 12-18

FORECAST 08-07-2010: DAY 19-25 FORECAST 01-07-2010: DAY 26-32

<-90mm

-90..-60

-60..-30

-30..-10

-10.. 0

0.. 10

10.. 30

30.. 60

60.. 90

> 90mm

WW

RP

Hydrological and atmospheric coupling – addition of a storm surge component

Importance of communication with users to understand requirements, best methods of dissemination and to encourage understanding and use of probabilistic forecasts for decision making

How do we verify over these timescalesNeed to be able to forecast parameters that can directly inform

disaster mitigation decision making – that is Tropical Cyclone intensity intensity and landfall probability at a range of timescales

South Asian Demonstration Project – considerations

WW

RP

Societal and Economic - overall considerations

Important to look at areas with skill & that have willing involvement from community & regional centres

Project Areas: Agriculture Rainy season onset Africa Polar Eritrea/Somalia/Kenya Southern Mexico/Central

America South East Asia

enhancing model predictability through better dynamics and assessing model error/limitations

enhancing predictability through user focussed metrics or indices that relate to the decision making process

Will involve Variable verification

techniques Decision on how much

intricacy is required for the end user to make actionable decisions

Uncertainty is more than the meteorology

WW

RP

WMO Lead-centres for verification and archiving

Two Lead Centres, facilitating user access to GPC products

Lead Centre for Long-range Forecast Multi-model Ensembles (LC-LRFMME)

Lead Centre for the Standard Verification System for Long-range Forecasts (LC-SVSLRF)

Aim: improve access and usability of global LRF products to aid production of regional/national climate services

WMO CBS has designated Global Producing Centres (GPCs) - currently 12 GPCs satisfying designation criteria

WW

RP

GPC Centre System Configuration (ensemble size of forecast)

Resolution (atmosphere)

Hindcast period used

Beijing Beijing Climate Centre Coupled (48) T63/L16 1983-2004

CPTEC Centre for Weather Forecasts and Climate Studies

2-tier (15) T62/L28 1979-2001

ECMWF European Centre for Medium Range Weather Forecasts

Coupled (41) T159/L62 1981-2005

Exeter Met Office Hadley Centre

Coupled (42) 1.25°x1.85°/L38 1989-2002

Melbourne Australian Bureau of Meteorology

Coupled (30) T47/L17 1980-2006

Montreal Meteorological Service of Canada

2-tier (40) T32/T63/T95/2.0°x2.0° (4- model combination)

1969-2004

Seoul Korean Meteorological Agency

2-tier (20) T106/L21 1979-2007

Tokyo Japan Meteorological Agency

Coupled (51) T95/L40 1979-2008

Toulouse Météo-France Coupled (41) T63/L91 1979-2007

Washington National Centres for Environmental Prediction

Coupled (40) T62/L64 1981-2004

Moscow Hydromet Centre of Russia

2-tier (10) 1.1°x1.4°/L28 1979-2003

Pretoria South African Weather Service

2-tier (6) T42/L19 1983-2001

The 12 WMO-designated GPCs

WMO Lead-centres for verification and

archiving

WW

RP

JJA forecasts (precipitation)Initialized in May

Beijing ECMWF Exeter Melbourne

Montreal Moscow Seoul Tokyo

Toulouse Washington

From WMO Lead Centre

http://www.wmolc.org/Downloaded on 1st July

TAC 42 Verification 2010

WW

RP

18

Forecast and verification products are available to RCCs /NMHSs / RCOFs Lead Centres will play a key role in WMO Global Seasonal Climate Updates

(monitoring and outlook) – part of Global Framework for Climate Services (GFCS) vision.

Key plans for the WMO Lead Centres (guided by the Expert Team for Extended and Long-range Forecasts) include:

development of probability forecast products extend prediction range (to ~6 months) verification of multi-model products possible centralisation of the verification process – currently self-verify

Issue of length of hindcast – but also of consistency of choice of the hindcast period across the GPCs

Working well for seasonal, but not really set up for sub-seasonal timescales Climate updates for WMO provides pull for 6-month forecasts Common verification needed for current predictions, not just for retrospective projects Suggestion to follow the “CMIP model” in making all data available (including hindcast data) rather than focusing on a few measures – many ways of looking at the data

WMO Lead-centres for verification and archiving

WW

RPEUROSIP multi-model ensemble

Three European models so far:ECMWFMet OfficeMeteo-FranceGermany planning to contribute

NCEP has just become an associate partnerNot yet integrated into system

An evolving systemReal-time since mid-2005Common operational schedule (products released at 12Z

on 15th)Monthly mean data in ECMWF operational archive (daily

from some partners)

WW

RP

Seasonal forecast – Nino SST, annual range

EUROSIP forecasts of SST anomalies over the NINO 3.4 region of the tropical Pacific from July 2009, December 2009 and May 2010. Showing the individual

ensemble members (red); and the subsequent verification (blue)

WW

RP

Data bases 2

CHFP (Climate Historical Forecast Project)

Database of hindcasts from many models Using best-available models and data for initialization

Being built now – monthly averages, eventually add daily output Initial states 4 times a year, from ECMWF or NCEP, 7 month runs, 1979-date Output (high frequency) at 24h intervals Hosted in Argentina and UK (exact copies, like TIGGE) Open to anyone, like TIGGE, for research Focus on specific research issues through subprojects such as GLACE, SHFP, SeaIce

HFP Suggest important to figure out user needs and how to get data to them first!

YOTC

Includes both model and observed data, and field campaign data Integrated observational data + high resolution global analysis from ECMWF, NCEP and

GMAO/NASA For research into MJO, Monsoons, Easterly waves and cyclones, tropical-extra tropical

interaction; diurnal cycle. Second copy of data at NCAR soon ECMWF data portal (like TIGGE), 380 registered users

WW

RP

Stratospheric Influences on the Troposphere

Monthly VariabilitySudden Stratospheric Warming – Rossby Wave

BreakingNAO/AO Response to Stratospheric Forcing

Apparent Downward PropagationCan we Predict the Warm?

Yes, 8-14 days in advance

S-I VariabilityStratosphere has long memory – QBOENSO Tele-connections in the StratosphereModeling the Stratosphere Improves the Surface

Tele-connectionsStratospheric Historical Forecast Project

(SHFP)

WW

RP

23

Forecast System StandardizationSeasonal predictions

More standardization across different operational centres is required Seasonal forecast systems are run on a daily, weekly, or a monthly

basis Some data exchange (mostly real-time forecast anomalies) are in

place via the efforts of the WMO LC-LRFMME Some coordination via LC-LRFMME, CHFP etc. Need common verification data sets (e.g., some use GPCP for

precipitation, some Xie-Arkin Optimum period for hindcasts?

WW

RPCollaboration between the Weather and Climate

Communities to Advance Sub-seasonal-to-Seasonal Prediction:

Research Issues

Seamless weather/climate prediction with Multi-model Ensemble Prediction Systems

Data assimilation for coupled models as a prediction and validation tool for weather and climate research

Utilization of sub-seasonal predictions for social and economic benefits

Significant progress: The multi-scale organisation of tropical convection and its two-way interaction with the global circulation

Sub-seasonal variability modulates the frequency of high-impact weather (e.g. AO, PNA, Atlantic blockings)

WW

RP

25

Monthly / sub-seasonal prediction

In principle, all GPCs that run seasonal predictions, In principle, all GPCs that run seasonal predictions, also have monthly predictions – but monthly mean also have monthly predictions – but monthly mean products are not sufficient for monthly to sub-products are not sufficient for monthly to sub-seasonal time-scalesseasonal time-scales

Probably not adequate level of standardization Probably not adequate level of standardization across different operational centersacross different operational centers

Monthly forecasts run on a daily or weekly basisMonthly forecasts run on a daily or weekly basis Mix of coupled, or atmospheric alone, prediction Mix of coupled, or atmospheric alone, prediction

systemssystems No data exchange efforts (?)No data exchange efforts (?)

WW

RP

26

Operational monthly to sub-seasonal forecasting

The JMA issues operationally, monthly forecasts every week The JMA issues operationally, monthly forecasts every week The Met Office will start testing a monthly forecast system in The Met Office will start testing a monthly forecast system in spring 2011.spring 2011.The ECMWF carries out operational monthly prediction with plans The ECMWF carries out operational monthly prediction with plans to extended the range to 45 daysto extended the range to 45 daysStarting in January 2011, National Centers for Environmental Starting in January 2011, National Centers for Environmental Prediction (NCEP) will also be running a monthly prediction system.Prediction (NCEP) will also be running a monthly prediction system.Canadian plans for the monthly forecast system are to base the Canadian plans for the monthly forecast system are to base the monthly forecast system on the Canadian global EPS.monthly forecast system on the Canadian global EPS.In Australia, the Bureau is currently experimenting with a monthly In Australia, the Bureau is currently experimenting with a monthly forecast system.forecast system.

There was a strong consensus that these operational There was a strong consensus that these operational approaches should be coordinated and the feasibility of data approaches should be coordinated and the feasibility of data exchange should be investigated exchange should be investigated

WW

RP

JMA: Yuhei Takaya

AGCM(JMA-GSM0803C)

TL159L60 (~110km)

JMA Global Data Assimilation System

Atmos. perturbationsTrop.&Ext.-trop. bred vec

Atmos I. C.

27

JMA Land Surface Analysis

25 members start from Wed25 members start from Thurs

50 members in total

coupler(w/ flux adjustment)

Climate DASJRA-25/JCDAS

JMA/MRI-CGCM

JMA GSM（ TL95L40: ~180k

m ）

Ocean model (MRI.COM )

1.0°×(0.3°-1.0°) ×L51 ODAS

MOVE/MRI.COM-G

Atmos perturbsTrop.&Ext. trop BV

Atmos I. C.

Oceanic I. C.Atmospheric BGM + Lagged Averaging

Forecast

Seasonal Forecast SystemMonthly Forecast System

9 member ensemble every 5 days

WW

RP

Capabilities in sub-seasonal prediction

Considerable progress in improving medium-range weather forecasts and developing operational seasonal prediction

Forecasting in the intermediate range between medium range and seasonal is difficult as the importance of the initial conditions wanes, and the importance of slower boundary conditions such as sea surface temperature increases but has only a modest influence on the weather and climate, especially away from the tropical regions.

Progress with representing the ENSO but not solved (including its associated tele-connections)

The MJO is an important source of predictability for the extended range and improving it’s representation improves the prediction of ENSO, theIndian Ocean Dipole, the Northern and Southern Annular Modes and the NAO ----

WW

RP

Capabilities/improvements - continued

Parametrization of physical processesRepresenting tropical convection and the MJORepresentation of blockingLand-surface conditionsThe influence of the stratosphereTreatment of the upper oceanCoupled data assimilation

Data bases to support researchTIGGECHFPSHFPGLACEYOTC

WW

RPMAIN RECOMMENDATIONS OF THE

WORKSHOP

Establishment of an International Research Project:

Based on the outcome of this workshop, it is recommended that a Panel/Project for Sub-seasonal and seasonal prediction research should be established. Panel members should include representatives from WWRP-THORPEX, WCRP, CAS, CBS, CHY and CCl and their relevant programme bodies. The first task for the Panel will be the preparation of an Implementation Plan which is consistent with the outcome of this Workshop - the main emphasis being on sub-seasonal prediction.

WW

RP

MAIN RECOMMENDATIONS continuedThe Implementation Plan should give high priority to: Sponsorship of a few international research activities ……. The establishment of collaboration and co-ordination

between operational centres undertaking sub-seasonal prediction to ensure, where possible, consistency between operational approaches to enable the production of data bases of operational sub-seasonal predictions to support research

Utilization of data collected for the CHFP, TIGGE GLACE and YOTC for research

The establishment of a series of regular Workshops on sub-seasonal prediction

In a separate plan, or as part of the Implementation Plan, the WWRP/SERA Working Group and WCRP should outline plans for a number of regional projects