Prediction of Canola Yield in Some of Growth Stages by Using Landsat Satellite, OLI Sensor

Canola is a source of edible oil and its cultivation in Iran and the world is growing. Only few studies have been carried out by remote sensing for canola yield estimation,. In 2017-2018, in order to predict the canola yield by Landsat satellite, OLI sensor, three farms were evaluated. The satellite images were processed in five stages: before flowering, early flowering, peak of flowering, green and dry maturing, and some of vegetation indices were extracted based on the ratio of the bands. The pixel network of each farm was determined and the Real Time Kinematic Global Positioning System (RTKGPS) was used to increase the precision of pixels location in the farms. Sampling was done inside farms pixels during harvesting time and canola yield was measured. Totally, 28 pixels from three studied farms were used to develop and validate the predictive models. Simple and multivariate linear regression models were used to assess the relationship between canola yield and vegetation indices. The results showed that, on the basis of simple linear regression models, among the growth stages, the highest coefficient of determination (R2) in each of the vegetation indices belonged to one of the two stages: the peak of flowering and green maturing. The coefficient of determination in all vegetation indices was low in the before flowering stage (less than 10 percent) and relatively medium (24- 52 percent) in the early flowering and dry maturing stages. According to this model, the NDYI with 67 percent in the peak of flowering stage, and the RVI with 64 percent in the green maturing stage had the highest coefficient of determination compared to other vegetation indices. The stepwise multivariate linear regression models, with four visible and near infrared bands, resulted to the best yield predictive model in the peak of flowering stage, with 78 and 74 percent of coefficient of determination, for its implementation and validation, respectively.