A data-driven approach improves food insecurity crisis prediction

E. C. Lentz, Hope C Michelson, Katherine R Baylis, Y. Zhou

Research output: Contribution to journalArticle

Abstract

Globally, over 800 million people are food insecure. Current methods for identifying food insecurity crises are not based on statistical models and fail to systematically incorporate readily available data on prices, weather, and demographics. As a result, policymakers cannot rapidly identify food insecure populations. These problems delay responses to mitigate hunger. We develop a replicable, near real-time model incorporating spatially and temporally granular market data, remotely-sensed rainfall and geographic data, and demographic characteristics. We train the model on 2010–2011 data from Malawi and forecast 2013 food security. Our model correctly identifies the food security status of 83 to 99% of the most food insecure village clusters in 2013, depending on the food security measure, while the prevailing approach correctly identifies between 0 and 10%. Our results show the power of modeling food insecurity to provide early warning and suggest model-driven approaches could dramatically improve food insecurity crisis response.

Original languageEnglish (US)
Pages (from-to)399-409
Number of pages11
JournalWorld Development
Volume122
DOIs
StatePublished - Oct 2019

Fingerprint

nutrition situation
food
prediction
food security
hunger
Malawi
train
Prediction
Food insecurity
village
weather
Food security
Food
rainfall
market
modeling

Keywords

  • Crisis
  • Early warning
  • Famine
  • Food insecurity
  • Prediction
  • Sub-Saharan Africa

ASJC Scopus subject areas

  • Geography, Planning and Development
  • Development
  • Sociology and Political Science
  • Economics and Econometrics

Cite this

A data-driven approach improves food insecurity crisis prediction. / Lentz, E. C.; Michelson, Hope C; Baylis, Katherine R; Zhou, Y.

In: World Development, Vol. 122, 10.2019, p. 399-409.

Research output: Contribution to journalArticle

@article{3532994af8ba46c78a6de32c012b11f0,
title = "A data-driven approach improves food insecurity crisis prediction",
abstract = "Globally, over 800 million people are food insecure. Current methods for identifying food insecurity crises are not based on statistical models and fail to systematically incorporate readily available data on prices, weather, and demographics. As a result, policymakers cannot rapidly identify food insecure populations. These problems delay responses to mitigate hunger. We develop a replicable, near real-time model incorporating spatially and temporally granular market data, remotely-sensed rainfall and geographic data, and demographic characteristics. We train the model on 2010–2011 data from Malawi and forecast 2013 food security. Our model correctly identifies the food security status of 83 to 99{\%} of the most food insecure village clusters in 2013, depending on the food security measure, while the prevailing approach correctly identifies between 0 and 10{\%}. Our results show the power of modeling food insecurity to provide early warning and suggest model-driven approaches could dramatically improve food insecurity crisis response.",
keywords = "Crisis, Early warning, Famine, Food insecurity, Prediction, Sub-Saharan Africa",
author = "Lentz, {E. C.} and Michelson, {Hope C} and Baylis, {Katherine R} and Y. Zhou",
year = "2019",
month = "10",
doi = "10.1016/j.worlddev.2019.06.008",
language = "English (US)",
volume = "122",
pages = "399--409",
journal = "World Development",
issn = "1873-5991",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - A data-driven approach improves food insecurity crisis prediction

AU - Lentz, E. C.

AU - Michelson, Hope C

AU - Baylis, Katherine R

AU - Zhou, Y.

PY - 2019/10

Y1 - 2019/10

N2 - Globally, over 800 million people are food insecure. Current methods for identifying food insecurity crises are not based on statistical models and fail to systematically incorporate readily available data on prices, weather, and demographics. As a result, policymakers cannot rapidly identify food insecure populations. These problems delay responses to mitigate hunger. We develop a replicable, near real-time model incorporating spatially and temporally granular market data, remotely-sensed rainfall and geographic data, and demographic characteristics. We train the model on 2010–2011 data from Malawi and forecast 2013 food security. Our model correctly identifies the food security status of 83 to 99% of the most food insecure village clusters in 2013, depending on the food security measure, while the prevailing approach correctly identifies between 0 and 10%. Our results show the power of modeling food insecurity to provide early warning and suggest model-driven approaches could dramatically improve food insecurity crisis response.

AB - Globally, over 800 million people are food insecure. Current methods for identifying food insecurity crises are not based on statistical models and fail to systematically incorporate readily available data on prices, weather, and demographics. As a result, policymakers cannot rapidly identify food insecure populations. These problems delay responses to mitigate hunger. We develop a replicable, near real-time model incorporating spatially and temporally granular market data, remotely-sensed rainfall and geographic data, and demographic characteristics. We train the model on 2010–2011 data from Malawi and forecast 2013 food security. Our model correctly identifies the food security status of 83 to 99% of the most food insecure village clusters in 2013, depending on the food security measure, while the prevailing approach correctly identifies between 0 and 10%. Our results show the power of modeling food insecurity to provide early warning and suggest model-driven approaches could dramatically improve food insecurity crisis response.

KW - Crisis

KW - Early warning

KW - Famine

KW - Food insecurity

KW - Prediction

KW - Sub-Saharan Africa

UR - http://www.scopus.com/inward/record.url?scp=85067472949&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067472949&partnerID=8YFLogxK

U2 - 10.1016/j.worlddev.2019.06.008

DO - 10.1016/j.worlddev.2019.06.008

M3 - Article

AN - SCOPUS:85067472949

VL - 122

SP - 399

EP - 409

JO - World Development

JF - World Development

SN - 1873-5991

ER -