Evaluating the Potential of Narrow-Band Indices to Predict
Soybean (Glycine Max L. Merr) Grain Yield in The Free
State and Mpumalanga of South Africa
Volume 3 - Issue 1
Siphokazi R Gcayi1,2*, George J Chirima1, Samuel A Adelabu2, Elhadi Adam3 and Khaled Abutaleb1
-
Author Information
Open or Close
- 1Geoinformatics Division, Agricultural Research Council Soil Water and Climate, Pretoria, South Africa
- 2Department of Geography, Faculty of Natural and Agricultural Sciences, University of the Free State, Phuthaditjhaba, South Africa
- 3Geography Department, Faculty of Science, University of the Witwatersrand, Johannesburg, South Africa
*Corresponding author:
Siphokazi R Gcayi, Department of Geography, Faculty of Natural and Agricultural Sciences, University of the
Free State, Phuthaditjhaba, South Africa
Received: June 14, 2019; Published: June 24, 2019
DOI: 10.32474/OAJESS.2019.02.000153
Full Text
PDF
To view the Full Article Peer-reviewed Article PDF
Abstract
Yield predictions allow for decision making regarding management of agricultural yield before and after harvest by government
and decision-makers. Traditional approaches to collect yield statistics such as manual field surveys and physical computation of
yield are costly and take a long time for information to be available. Remote sensing platforms such as hyperspectral data provide
real-time, fast, and reliable statistics that can be used to derive yield information. Vegetation indices are ratios used to combine
multiple band observations of the hyperspectral data into one index and applied to derive soybean grain yield. The objective of this
study was to evaluate the potential of vegetation indices derived from hyperspectral data to predict soybean grain yield. Soybean
hyperspectral data was acquired using a handheld spectroradiometer with a spectral range of 350 to 2500 nm in March and April of
the summer season of 2017. The random forest regression algorithm was used to predict the soybean grain yield. NDVI, SR and EVI
were calculated from the hyperspectral data for all probable bands situated in the 400 nm and 2399 regions. The results showed
that relevant wavelengths in predicting soybean were combinations situated in the red-edge (680-750 nm), NIR and the MIR (1300
to 2399 nm) of the electromagnetic spectrum. Furthermore, regression results showed that SR better predicted the soybean grain
yield (R2
= 0.843) compared to NDVI (R2
= 0.841) and EVI (R2
= 0.537). In overall, the results of this study suggest that narrow-band
indices have the potential to predict soybean grain yield.
Keywords: Soybean Yield; Hyperspectral Data; Vegetation Indices
Abbreviations: NDVI: Normalised Difference Vegetation Index; SR: Simple Ratio; EVI: Enhanced Vegetation Index; RF: Random
Forest; RMSE: Root Mean Square Error; FS: Free State; MP: Mpumalanga
Abstract|
Introduction|
Materials and Methods|
Data analysiss|
Results|
Discussion|
Conclusion|
Acknowledgement|
References|