IMPROVING METHODS FOR CROP YIELD FORECAST AND EARLY WARNING SYSTEMS

BRUNO BASSO - MICHIGAN STATE UNIVERSITY - USA

• Crop simulation models (CSM) – types of models • Remote Sensing (RS): Principles and Vegetation

Indices • Yield Gap Analysis • Linking CSM with RS for yield forecast • Early Warning Systems (EWS) • Research needs for improving Ag Statistics in

developing countries

Outline - Research Topics

Yield Forecasting Methods

0 200 400 600 800

Growing-season rainfall (mm)

0

2

4

6

8

Gra

in y

ield

(t

ha

-1)

District averageyields

Rainfall alone cannot be used to predict yield

Evapotranspiration (mm)

0 100 200 300 400 500 600

Yie

ld (

t/h

a)

0

1

2

3

4

5

6

n = 691

China Loess Plateau

Mediterranean Basin

North American Great Plains

SE Australia

Lobell et al., 2013 Nature Climate Change

Sadras and Angus, 2006 Aust. J Ag. Res

• Types of models: Deterministic and Stochastic • Deterministic: Statistical, Mechanistic and Functional

Crop Simulation Models

Input Data

Weather

Soil

Management

Crop

Crop Yield

Soil Biochem, N , C, P dynamics

Soil Water Balance

Crop Growth Modules

Soil organic matter and Nutrient

Module

Soil Water Balance Module

Example of a functional model (SALUS)

SUB-TOPICS REQUIRING FURTHER RESEARCH

NASA POWER

Soil WISE Database

Spatial and Temporal Variability of Crop Yield

Spatial Variability Temporal Variability

Simulation and Predictions

0

2

4

6

8

10

12

14

2007 2008 2009 2010 2011 2012 2012 lowyielding

zone(higher

elevation)

2013 tillJuly 5,

then wetyear like

2011

2013 tillJuly 5,

then dryyear like

2012

Wet yrwith 10days of37C andno rain

for 1 mo

Dry yearwith

deficitirrigation

Co

rn G

rain

Yie

lds

(Mg/

ha)

Years

SIMULATED

MEASURED

AgMIP project

Rosenzeweig et al., 2013, Ag. For. Meteo; Asseng et al., 2013 Nature Climate Change

Remote Sensing

NDVI

NDVI = (NIR-RED) / (NIR+RED)

RS can be used for yield estimation directly using VI or indirectly by incorporating data into simulation model for crop growth and development either as within season calibration checks of model output (LAI, biomass) or in a feedback loop used to adjust model starting conditions (Maas, 1988).

ND

VI

Leaf Area Index (m2/m2)

Weekly changes of NDVI

Early Warning Systems

• Early Warning System (EWS) is defined as an integrated system for monitoring, data collection, analysis, and communicating to people in order to make early decisions to protect peoples and the environment (Davies et al. 1991).

• FAO Global Information & Early Warning System (GIEWS) on Food and Agriculture (ISDR, 2009)

• Most of the EWS use RS for crop monitoring and early screening for signs of drought, flood, ENSO, climate change

Research for improving Ag Statistics

• RS - Create a database with availability and usefulness of satellite to exploit its spatial, temporal and spectral resolutions and its cost

• Test functional CSM using observed weather and soil properties along different scenarios of management to reproduce what RS has indicated

• Run pilot projects where CSM are integrated with RS to test the capability of the procedure under stratified agriculture to improve Ag Stats in developing world

Thank you

Bruno Basso

Michigan State University, USA

basso@msu.edu

The first Scientific Advisory Committee

FAO Headquarters, 18 and 19 July 2013