Evaluation of remote sensing vegetation indices for monitoring maize crop condition and yields in Tanzania
Loading...
Date
2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Sokoine University of Agriculture
Abstract
Monitoring smallholder production systems for crop condition and yield through remote
sensing techniques is challenging due to the heterogeneity of farming systems. The study
was conducted at Sokoine University of Agriculture (SUA) and Magadu farmers’ maize
fields in Morogoro Tanzania. This study investigated the relation between vegetation
indices (Vls) and field measurements of crop condition and yield as indicated by maize
growth (biophysical) parameters under mixed farming practices. Unmanned Aerial
Vehicle (UAV) and Landsat 8 (green, red, red-edge, and near infrared spectral bands)
were used to calculate four spectral vegetation indices; the normalized difference
vegetation index (NDVI), wide dynamic range vegetation index (WDRVI), red-edge
chlorophyll index (CIred-edge)» and the green chlorophyll index (CIgreen)- The Clgreen, Clred"
edge and NDVI showed ability to detect differences in maize crop biophysical parameters
With CIgreen and NDVI performing better than the other indices. The CIgreen and NDVI were
the best indices for detection of the effect of fertilizer application (fertilizer and non
fertilizer) and pigeon pea intercropping (with pigeon pea and without pigeon pea
intercrop) on maize at 60 days after sowing due to R2 > 0.50. The best performance of
these indices was on grain yield, where all UAV-VIs detected yield variability by higher
R2 values (> 0.60) for both sole and intercropped maize. The assessment for the usefulness
of satellite remote sensing Vls showed that the satellite derived Vis were significantly
affected more by cloud cover than UAV-Vls. Thus, the NDVI and WDRVI derived from
UAV and Landsat 8 were significantly different (p < 0.05); unlike UAV-CIgreen which
exhibited some consistence with Landsat 8-Clgreen and their differences were not
statistically significant except under heavy cloud cover. These findings clearly
demonstrate the need to use multiple vegetation indices to best monitor maize crop
conditions and yield differences attributed to different farming practices and weather
conditions.
Description
Dissertation
Keywords
Smallholder production systems, Maize fields, Monitoring smallholder, Heterogeneity farming systems