Using Precipitation and Temperature to Model Agriculture Conditions in Africa Eric Wolvovsky NOAA/FEWS-NET July 1, 2008.

Using Precipitation and Using Precipitation and Temperature to Model Temperature to Model

Agriculture Conditions in Agriculture Conditions in AfricaAfrica

Eric WolvovskyEric WolvovskyNOAA/FEWS-NETNOAA/FEWS-NET

July 1, 2008July 1, 2008

OverviewOverview

Introduction to FEWS-NETIntroduction to FEWS-NET MethodologyMethodology OutputOutput ApplicationsApplications Potential Future WorkPotential Future Work ConclusionConclusion

IntroductionIntroduction

Famine Early Warning System NetworkFamine Early Warning System Network Early warning on food security concernsEarly warning on food security concerns US Agencies involvedUS Agencies involved

• USAID (Lead)USAID (Lead)• USDAUSDA• USGSUSGS• NASANASA• NOAANOAA• ChemonicsChemonics

USGS and Chemonics have staff in countryUSGS and Chemonics have staff in country

IntroductionIntroduction

IntroductionIntroduction

NOAAs role in FEWS-NETNOAAs role in FEWS-NET Analyze and track meteorological Analyze and track meteorological

phenomenon as it relates to food securityphenomenon as it relates to food security• Tropical cyclonesTropical cyclones• Large scale severe weatherLarge scale severe weather• Extreme heatExtreme heat• FreezesFreezes• Rainfall for crops, pastures and drinking waterRainfall for crops, pastures and drinking water

IntroductionIntroduction

Goals for model:Goals for model: Analyze individual cropsAnalyze individual crops Analyze regionallyAnalyze regionally High resolutionHigh resolution Simple metricSimple metric Light weightLight weight Relates temperature and rainfallRelates temperature and rainfall

MethodologyMethodology

Blaney-Criddle FormulaBlaney-Criddle Formula

E is seasonal moisture requiredK is crop coefficientTai is mean monthly temperaturedi is monthly fraction of annual daylight hoursn is number of months

MethodologyMethodology

Data ChallengesData Challenges Of the 1000 weather Of the 1000 weather

stations in Africa ~500 stations in Africa ~500 report dailyreport daily

Data is not filteredData is not filtered• May have bad dataMay have bad data• May have reported May have reported

-999.0-999.0

MethodologyMethodology

CPC RFE 2.0CPC RFE 2.0 Uses 3 satellite inputs and Uses 3 satellite inputs and

daily station datadaily station data Daily temporal resolutionDaily temporal resolution 0.1 degree spatial 0.1 degree spatial

resolutionresolution StrugglesStruggles

• CoastsCoasts• MountainsMountains• Areas with few station reportsAreas with few station reports

MethodologyMethodology

NCEP/NCAR ReanalysisNCEP/NCAR Reanalysis Uses:Uses:

• StationStation• ShipShip• AircraftAircraft• SatelliteSatellite

Monthly Temporal ResolutionMonthly Temporal Resolution 2.5 degree spatial resolution2.5 degree spatial resolution Temperatures have a warm Temperatures have a warm

bias at higher elevationsbias at higher elevations

MethodologyMethodology

Monthly Fractional Hours Monthly Fractional Hours of Annual Daylightof Annual Daylight Developed as a function Developed as a function

of latitude based on fixed of latitude based on fixed valuesvalues

Monthly temporal Monthly temporal resolution resolution

0.1 degrees resolution0.1 degrees resolution Hours of daylight varies Hours of daylight varies

only with latitudeonly with latitude

MethodologyMethodology

FAO Crop shapefilesFAO Crop shapefiles Monthly temporal resolutionMonthly temporal resolution

Crop CoefficientCrop Coefficient Determined by US Soil Determined by US Soil

Conservation Service field Conservation Service field teststests

Values usedValues used• Maize 2.2Maize 2.2• Sorghum 2Sorghum 2• Wheat 1.8Wheat 1.8• Millet 1.4Millet 1.4

MethodologyMethodology

Blaney-Criddle FormulaBlaney-Criddle Formula

*Crop Coefficient *

MethodologyMethodology

MethodologyMethodology

Conditions are determined by comparing Conditions are determined by comparing required rainfall with received rainfallrequired rainfall with received rainfall

Percent of RequiredPercent of Required RainfallRainfall ClassificationClassification

Less than 50%Less than 50% FailureFailure

Between 50% and 75%Between 50% and 75% PoorPoor

Between 75% and 125%Between 75% and 125% Below AverageBelow Average

Between 125% and 175%Between 125% and 175% AverageAverage

Between 175% and 225%Between 175% and 225% GoodGood

Greater than 225%Greater than 225% ExcellentExcellent

Required Rainfall

CPC RFE 2.0* 100 = Percent of Required Rainfall Received

MethodologyMethodology

OutputOutput

OutputOutput

OutputOutput

OutputOutput

OutputOutput

OutputOutput

ApplicationsApplications

Hazards assessmentsHazards assessments

Weekly weather briefingsWeekly weather briefings

Use by decision makersUse by decision makers

Potential Future WorkPotential Future Work

Beyond AfricaBeyond Africa

Beyond GrainsBeyond Grains

Increase temporal resolutionIncrease temporal resolution

Better method of validationBetter method of validation

ConclusionConclusion

Light weight agriculture modelLight weight agriculture model

Method uses inputs that are knownMethod uses inputs that are known

Method is expandableMethod is expandable

Will support FEWS-NETWill support FEWS-NET

Thank YouThank You

[email protected]@noaa.gov