MODIS phenology-derived, multi-year distribution of conterminous U.S. crop types

Richard Massey, Temuulen T. Sankey, Russell G. Congalton, Kamini Yadav, Prasad S. Thenkabail, Mutlu Ozdogan, Andrew J Sanchez Meador

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

Innovative, open, and rapid methods to map crop types over large areas are needed for long-term cropland monitoring. We developed two novel and automated decision tree classification approaches to map crop types across the conterminous United States (U.S.) using MODIS 250 m resolution data: 1) generalized, and 2) year-specific classification. The classification approaches use similarities and dissimilarities in crop type phenology derived from NDVI time-series data for the two approaches. The year-specific approach uses the training samples from one year and classifies crop types for that year only, whereas the generalized classification approach uses above-average, average, and below-average precipitation years for training to produce crop type maps for one or multiple years more robustly. We produced annual crop type maps using the generalized classification approach for 2001–2014 and the year-specific approach for 2008, 2010, 2011 and 2012. The year-specific classification had overall accuracies > 78%, while the generalized classifier had accuracies > 75% for the conterminous U.S. for 2008, 2010, 2011, and 2012. The generalized classifier enables automated and routine crop type mapping without repeated and expensive ground sample collection year after year. The resulting crop type maps for years prior to 2007 are new and especially important for long-term cropland monitoring and food security analysis because no other map products are currently available for 2001–2007.

Original languageEnglish (US)
Pages (from-to)490-503
Number of pages14
JournalRemote Sensing of Environment
Volume198
DOIs
StatePublished - Sep 1 2017

Fingerprint

moderate resolution imaging spectroradiometer
phenology
MODIS
Crops
crop
crops
taxonomy
Classifiers
distribution
Monitoring
monitoring
Decision trees
food security
rapid methods
NDVI
Time series
time series analysis
time series
sampling

Keywords

  • Adjusted kappa
  • Cropland data layer
  • Cropland mapping
  • Generalized classifier
  • MOD09Q1
  • NDVI time-series
  • United States croplands

ASJC Scopus subject areas

  • Soil Science
  • Geology
  • Computers in Earth Sciences

Cite this

MODIS phenology-derived, multi-year distribution of conterminous U.S. crop types. / Massey, Richard; Sankey, Temuulen T.; Congalton, Russell G.; Yadav, Kamini; Thenkabail, Prasad S.; Ozdogan, Mutlu; Sanchez Meador, Andrew J.

In: Remote Sensing of Environment, Vol. 198, 01.09.2017, p. 490-503.

Research output: Contribution to journalArticle

Massey R, Sankey TT, Congalton RG, Yadav K, Thenkabail PS, Ozdogan M et al. MODIS phenology-derived, multi-year distribution of conterminous U.S. crop types. Remote Sensing of Environment. 2017 Sep 1;198:490-503. https://doi.org/10.1016/j.rse.2017.06.033
Massey, Richard ; Sankey, Temuulen T. ; Congalton, Russell G. ; Yadav, Kamini ; Thenkabail, Prasad S. ; Ozdogan, Mutlu ; Sanchez Meador, Andrew J. / MODIS phenology-derived, multi-year distribution of conterminous U.S. crop types. In: Remote Sensing of Environment. 2017 ; Vol. 198. pp. 490-503.
@article{10b977e57b984d47b324572b82ee85ad,
title = "MODIS phenology-derived, multi-year distribution of conterminous U.S. crop types",
abstract = "Innovative, open, and rapid methods to map crop types over large areas are needed for long-term cropland monitoring. We developed two novel and automated decision tree classification approaches to map crop types across the conterminous United States (U.S.) using MODIS 250 m resolution data: 1) generalized, and 2) year-specific classification. The classification approaches use similarities and dissimilarities in crop type phenology derived from NDVI time-series data for the two approaches. The year-specific approach uses the training samples from one year and classifies crop types for that year only, whereas the generalized classification approach uses above-average, average, and below-average precipitation years for training to produce crop type maps for one or multiple years more robustly. We produced annual crop type maps using the generalized classification approach for 2001–2014 and the year-specific approach for 2008, 2010, 2011 and 2012. The year-specific classification had overall accuracies > 78{\%}, while the generalized classifier had accuracies > 75{\%} for the conterminous U.S. for 2008, 2010, 2011, and 2012. The generalized classifier enables automated and routine crop type mapping without repeated and expensive ground sample collection year after year. The resulting crop type maps for years prior to 2007 are new and especially important for long-term cropland monitoring and food security analysis because no other map products are currently available for 2001–2007.",
keywords = "Adjusted kappa, Cropland data layer, Cropland mapping, Generalized classifier, MOD09Q1, NDVI time-series, United States croplands",
author = "Richard Massey and Sankey, {Temuulen T.} and Congalton, {Russell G.} and Kamini Yadav and Thenkabail, {Prasad S.} and Mutlu Ozdogan and {Sanchez Meador}, {Andrew J}",
year = "2017",
month = "9",
day = "1",
doi = "10.1016/j.rse.2017.06.033",
language = "English (US)",
volume = "198",
pages = "490--503",
journal = "Remote Sensing of Environment",
issn = "0034-4257",
publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - MODIS phenology-derived, multi-year distribution of conterminous U.S. crop types

AU - Massey, Richard

AU - Sankey, Temuulen T.

AU - Congalton, Russell G.

AU - Yadav, Kamini

AU - Thenkabail, Prasad S.

AU - Ozdogan, Mutlu

AU - Sanchez Meador, Andrew J

PY - 2017/9/1

Y1 - 2017/9/1

N2 - Innovative, open, and rapid methods to map crop types over large areas are needed for long-term cropland monitoring. We developed two novel and automated decision tree classification approaches to map crop types across the conterminous United States (U.S.) using MODIS 250 m resolution data: 1) generalized, and 2) year-specific classification. The classification approaches use similarities and dissimilarities in crop type phenology derived from NDVI time-series data for the two approaches. The year-specific approach uses the training samples from one year and classifies crop types for that year only, whereas the generalized classification approach uses above-average, average, and below-average precipitation years for training to produce crop type maps for one or multiple years more robustly. We produced annual crop type maps using the generalized classification approach for 2001–2014 and the year-specific approach for 2008, 2010, 2011 and 2012. The year-specific classification had overall accuracies > 78%, while the generalized classifier had accuracies > 75% for the conterminous U.S. for 2008, 2010, 2011, and 2012. The generalized classifier enables automated and routine crop type mapping without repeated and expensive ground sample collection year after year. The resulting crop type maps for years prior to 2007 are new and especially important for long-term cropland monitoring and food security analysis because no other map products are currently available for 2001–2007.

AB - Innovative, open, and rapid methods to map crop types over large areas are needed for long-term cropland monitoring. We developed two novel and automated decision tree classification approaches to map crop types across the conterminous United States (U.S.) using MODIS 250 m resolution data: 1) generalized, and 2) year-specific classification. The classification approaches use similarities and dissimilarities in crop type phenology derived from NDVI time-series data for the two approaches. The year-specific approach uses the training samples from one year and classifies crop types for that year only, whereas the generalized classification approach uses above-average, average, and below-average precipitation years for training to produce crop type maps for one or multiple years more robustly. We produced annual crop type maps using the generalized classification approach for 2001–2014 and the year-specific approach for 2008, 2010, 2011 and 2012. The year-specific classification had overall accuracies > 78%, while the generalized classifier had accuracies > 75% for the conterminous U.S. for 2008, 2010, 2011, and 2012. The generalized classifier enables automated and routine crop type mapping without repeated and expensive ground sample collection year after year. The resulting crop type maps for years prior to 2007 are new and especially important for long-term cropland monitoring and food security analysis because no other map products are currently available for 2001–2007.

KW - Adjusted kappa

KW - Cropland data layer

KW - Cropland mapping

KW - Generalized classifier

KW - MOD09Q1

KW - NDVI time-series

KW - United States croplands

UR - http://www.scopus.com/inward/record.url?scp=85021759504&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021759504&partnerID=8YFLogxK

U2 - 10.1016/j.rse.2017.06.033

DO - 10.1016/j.rse.2017.06.033

M3 - Article

AN - SCOPUS:85021759504

VL - 198

SP - 490

EP - 503

JO - Remote Sensing of Environment

JF - Remote Sensing of Environment

SN - 0034-4257

ER -