Main Article Content

Evaluating Terra MODIS Satellite Sensor Data Products for Maize Yield Estimation in South Africa


C Frost
N Thiebaut
T Newby

Abstract

The Free State Province of the Republic of South Africa contains some of the most important maize-producing areas in South Africa. For this reason this province has also been selected as a Joint Experiment for Crop Assessment and Monitoring (JECAM, 2012) site representative of South Africa.
The Terra (EOS AM-1) research satellite carries the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. Two data products which are used in this research, created by the National Aeronautics and Space Administration (NASA) from the MODIS sensor, are the Normalised Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI).
Objective yield points (OYP) are Global Positioning Points (GPS) that fieldworkers visit to record certain yield related data, including a final objective yield. Three research studies utilizing Terra MODIS data were performed. Study one utilised objective yields from the 2001/2002, 2004/2005 and 2005/2006 growth seasons for three provinces in South Africa. In study two, the 2006/2007 growth season OYP were extracted for the Free State province only. NDVI was downloaded from United States Geological Survey (USGS) for study one and two and extracted into a GIS for comparison with the OYP. For the Free State pilot study, study three, OYP from the 2001 to 2010 growth seasons were used. MODIS NDVI and EVI data were obtained from the Wide Area Monitoring Information System WAMIS portal for the data set of 35 points.
Statistical analysis was done on the data of studies two and three. The 2006/2007 dataset yielded a R2 of 0.47 for all 225 points and 0.51 for the medium growth cultivar data only (186 data points), while the 2001 to 2010 growth season yielded an R2 of 0.63 for the medium yield group. Therefore it is concluded that Terra MODIS NDVI and EVI data can be utilized for maize yield estimation on points.

Journal Identifiers


eISSN: 2225-8531