Efficient and Sustainable Crop Intensification: An Assessment of Phenofit Algorithm and Envelope Crop Classification Method for its Monitoring

Abstract

To optimize use of land, farmers need to make decisions regarding grain varieties, rotation, different crop management systems, and whether to sow a single or double crop in a calendar year. In Córdoba (Argentina), the predominant crops are wheat, soybean and maize, sown as single crop (SC) or double crop (DC) sequences (wheat–soybean or wheat–maize). The objective of this work was to compare Phenofit algorithm and Envelope Crop Classification (ECC) method to identify the presence of SC or DC using MODIS-NDVI temporal series. Calibration and validation were carried out using field data acquired from 2015 to 2018. NDVI signatures of each plot were compared with SC and DC temporal NDVI profiles and the class membership was determined when at least 50% of values fell inside of one profile and the difference between classes was positive. The results showed that the ECC/Phenofit present overall accuracy between 96/90 and 98/92% and Kappa coefficients from 91/82 to 97/95%, respectively. On average, when the ECC was applied, the percentages of the study area detected as DC were between 18.3 and 28.7%, for the considered periods, while the area occupied with SC decreased from 64 to 49.5%. ECC and Phenofit are very good methods for detecting double crop.