IRI Experience in Forecastingand Applications for the

Mercosur Countries

Mason & Goddard, 2001, Bull.Amer.Meteor.Soc.

PurelyEmpirical(observational)approach

Prediction Systems:statistical vs. dynamical system

ADVANTAGES

Based on actual, real-worldobserved data. Knowledge ofphysical processes not needed.

Many climate relationshipsquasi-linear, quasi-Gaussian------------------------------------Uses proven laws of physics.Quality observational data not required (but helpful for val-idation). Can handle casesthat have never occurred.

DISADVANTAGES

Depends on quality and length of observed data

Does not fully account for climate change, or new climate situations.------------------------------ Some physical laws must be abbreviated or statis- tically estimated, leading to errors and biases.

Computer intensive.

Stati-stical

-------

Dyna-mical

Dynamical Prediction System:2-tiered vs. 1-tiered forecast system ADVANTAGES

Two-way air-sea interaction,as in real world (required Where fluxes are as important as large scale ocean dynamics)

-------------------------------------More stable, reliable SST inthe prediction; lack of driftthat can appear in 1-tier system

Reasonably effective for regionsimpacted most directly by ENSO

DISADVANTAGES

Model biases amplify (drift); flux corrections

Computationally expensive------------------------------ Flawed (1-way) physics, especially unacceptable in tropical Atlantic and Indian oceans (monsoon)

1-tier

-------

2-tier

30

12

30

12

10

24

10

24

10

FORECAST SST

TROP. PACIFIC: THREE (multi-models, dynamical and statistical)

TROP. ATL, INDIAN (ONE statistical)

EXTRATROPICAL (damped persistence)

GLOBAL ATMOSPHERIC

MODELS

ECPC(Scripps)

ECHAM4.5(MPI)

CCM3.6(NCAR)

NCEP(MRF9)

NSIPP(NASA)

COLA2

GFDL

ForecastSST

Ensembles3/6 Mo. lead

PersistedSST

Ensembles3 Mo. lead

IRI DYNAMICAL CLIMATE FORECAST SYSTEM

POSTPROCESSING

MULTIMODELENSEMBLING

PERSISTED

GLOBAL

SST

ANOMALY

2-tiered OCEAN ATMOSPHERE

30

(1) NCEP CFS(2) LDEO

(3) CPC Constr AnalogIRI CCA CPTEC

CCA

dampedpersistence

damped persistence

0.25--------------------------------------------------

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

Method of Forming 3 SST Predictions for Climate Predictions

Method of Forming 4th SST Prediction for Climate Predictions

(4)Anomaly

persistence

from most

recently

observed

month

(all oceans)

 

Atmospheric General Circulation Models Used in the IRI's Seasonal Forecasts, for Superensembles

 Name Where Model Was Developed Where Model Is Run

NCEP MRF-9 NCEP, Washington, DC QDNR, Queensland, Australia

ECHAM 4.5 MPI, Hamburg, Germany IRI, Palisades, New York

NSIPP NASA/GSFC, Greenbelt, MD NASA/GSFC, Greenbelt, MD

COLA COLA, Calverton, MD COLA, Calverton, MD

ECPC SIO, La Jolla, CA SIO, La Jolla, CA

CCM3.6 NCAR, Boulder, CO IRI, Palisades, New York

GFDL GFDL, Princeton, NJ GFDL, Princeton, NJ

Collaboration on Input to Forecast Production

Sources of the Global Sea Surface Temperature Forecasts 

Tropical Pacific Tropical Atlantic Indian Ocean Extratropical Oceans

NCEP Coupled CPTEC Statistical IRI Statistical Damped PersistenceLDEO Coupled Constr Analogue 

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 | || ||| ||||.| || | | || | | | . | | | | | | | | | |

Rainfall Amount (mm)

Below| Near | Below| Near | Above Below| Near |

The tercile-based category system:Below, near, and above normal

(30 years of historical data for a particular location & season)

Forecasts of the climate

Data:

33% 33% 33%Probability:

JUL | Aug-Sep-Oct* Sep-Oct-Nov

Oct-Nov-DecNov-Dec-Jan

Monthly issued probability forecasts of seasonal global precipitation and temperature

*probabilities of extreme (low and high) 15% issued also

Four lead times - example:

OND 1997 OND 1998

verygoodskill

goodskill

OND 1999 OND 2000

somewhatnegative

skillfairlygoodskill

OND 2001 OND 2002

Neutral(zero)skill good

skill

OND 2003 OND 2004

negativeskill

neutral(zero)skill

Historical skill using actually observed SST

Ensemble mean (10 members)

Historical skill using actually observed SST

Ensemble mean (24 members)

OND

JFM

AMJ

JAS

ONDPrecip1950-2000

SouthAmerica

Europe North America

TropicalAmericas

Indonesia

Australia/Indonesia

Some Themes in Customer Demands:

1. Higher skills in existing products: -Tighter range of possibilities2. More usable and easily understood product3. Expansion of repertoire of products beyond seasonally averaged climate

Specifically, users consistently request:More concrete forecast format Direct use of physical units (inches, degr C) Best guess quantity, with comparison to average More flexibility in probabilistic forecast format

More temporal detail than 3 month averages Outlooks for single months, even at long lead Rainy season (or monsoon) onset time Weather features within season (e.g. dry spells) Chance of extreme event (storm, flood, drought)

Climate change nowcast (the new “normal”)

Almost all users want reliable probabilities (probabilities that would match observed frequencies of occurrence over a large sample of identical forecasts).

Even if reliable, users for some applications are not impressed with slight changes from the climatological precipitation probabilities. They are interested inforecasts having probabilities that deviate strongly from the neutral, climatological probabilities.

OND 1997

Ranked Probability Skill Score (RPSS)

3 prob prob

RPSfcst = (Fcsticat – Obsicat)2

icat=1

icat ranges from 1 (below normal) to 3 (above normal)

icat is cumulative, from 1 to icat

RPSS = 1 - (RPSfcst / RPSclim)

RPSclim is RPS of forecasts of climatology (33%,33%,33%)

Real-timeForecast

Skill(from Goddard

et al. 2003,BAMS, p1761)

Real-timeForecast

Skill(from Goddard

et al. 2003,BAMS, p1761)

Real-timeForecast

Skill(from Goddard

et al. 2003,BAMS, p1761

1998-2001 Reliability Diagram

from Barnston et al. 2003, BAMS, p1783

Reliability Diagram

longer “AMIP”period

from Goddardet al. 2003

(EGS-AGU-EUG)

IRI Forecast Information Pages on the Web

IRI Home Page: http://iri.columbia.edu/

ENSO Quick Look:http://iri.columbia.edu/climate/ENSO/currentinfo/QuickLook.html

IRI Probabilistic ENSO Forecast:http://iri.columbia.edu/climate/ENSO/currentinfo/figure3.html

Current Individual Numerical Model Climate Forecastshttp://iri.columbia.edu/forecast/climate/prec03.html

Individual Numerical Model Hindcast Skill Mapshttp://iri.columbia.edu/forecast/climate/skill/SkillMap.html

Current Multi-model Climate Forecastshttp://iri.columbia.edu/forecast/climate/multi_model2003.html