Predicting paddy yield at spatial scale using optical and Synthetic Aperture Radar (SAR) based satellite data in conjunction with field-based Crop Cutting Experiment (CCE) data

ABSTRACT

The accurate, reliable, and robust information on crop yield has great importance in food security measures. In this study, both optical (Moderate Resolution Imaging Spectroradiometer (MODIS)-derived Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI)) and Synthetic Aperture Radar (SAR) (Sentinel-1A) satellite datasets were used for predicting paddy yield at a spatial scale in conjunction with Crop Cutting Experiment (CCE) data in Sahibganj district of the Jharkhand state (India) during the monsoon season 2017. The yield prediction models were developed from linear (LR) and multiple regression (MR) analysis. The AquaCrop model was also employed to simulate the paddy yield. The key findings showed that the MR-based yield model developed from SAR (Vertical transmission and Vertical reception (VV) + Vertical transmission and Horizontal reception (VH) polarizations) data was more accurate, reasonable, and reliable as compared to the optical-based (NDVI + EVI) MR yield model. The SAR-based MR yield model showed better accuracy between predicted and observed yield (CCE) as evaluated using Nash-Sutcliffe Efficiency (NSE = 0.68). However, the optical-based MR yield model showed relatively lower accuracy (NSE = 0.62). The relative deviation (RD) of the predicted yield from the SAR and optical-based MR model was nearly 3% and 4%, respectively. Using the AquaCrop model, the simulated yield was underestimated by approximately 4%. We conclude that the SAR-based MR-yield model outperformed the optical-based model to predict paddy yield at a spatial scale, and the adopted methodology can beneficial for decision-makers withing agriculture monitoring. Hence, satellite imagery has always been a reliable source for yield forecasting and crop assessment at a regional and national scale.

Acknowledgements

This research was the part of M. Tech project of the first author (AKR) carried out at Department of Geoinformatics, Central University of Jharkhand (India). Authors express gratitude to the farmers who have helped to conduct the crop cutting experiment (CCE) and provided the related information. Authors also thanks to Anniket Rajak, Amarjeet Rajak, and Amit Rajak for their help during field data collection. We thanks to the anonymous reviewers, Editor-in-Chief, and technical editor for their constructive comments which have certainly enhanced the overall quality of the manuscript. We sincerely acknowledge the United States Geological Survey (USGS) and European Space Agency (ESA) for providing MODIS products (MOD13Q1) and Sentinel-2B & 1A satellite data, respectively. We also acknowledge the mission scientists and Principal Investigators (Giovanni, NASA GES DISC; NASA POWER) who provided the data (air temperature, wind speed, insolation, relative humidity, and rainfall, respectively) used in this research effort.

Authors’ contributions

Conceived, designed research, analyzed data, and wrote the manuscript: AKR and BRP. All authors have read and approved the final version of manuscript for publication.

Disclosure statement

Authors declare no potential conflict of interest.

Ethical statement

All ethical practices have been followed during this research.

Supplementary material

Supplemental data for this article can be accessed here.

Additional information

Funding

This research was supported by the Science and Engineering Research Board (SERB), Department of Science & Technology (DST) project grant no. YSS/2015/000801.