Combining Sentinel-1 and Sentinel-2 images to monitor irrigation in sugar cane

摘要

This article analyzed the application of Sentinel-2 and Sentinel-1 for monitoring and controlling irrigation in sugarcane production by vinasse (residue of industrial ethanol-sugar manufacturing), a sub-product highly rich in organic matter and minerals, which is used as fertilizer after harvesting, during the months of drought for the plant to regrow. Irrigation with vinasse is a complex process and its lack of use or excess can causes important losses in tonnage of sugarcane. The insufficient spraying of vinasse was identified in the Sentinel-2 image of April 2017 in a farm plantation as the principal cause of stress in the development of the plant in an area representing 21% of the plantation (100 hectares). This represented a total loss of 4335 tons of cane (～S90000). The identification of the cause of the reduced growth was made possible through analyzing a Sentinel-2 image of July 2016 that showed clearly that no vinasse was spread in the same 100 hectares with cane growing problem. The presence/lack of vinasse was easily detectable using the visible bands of the S-2 image but the difference in growth/stress was better related to the red edge and near infrared bands in the January image. To thoroughly complete our investigation we also acquired a Sentinel-1 radar image of April 2017 where the lack of vinasse could not directly be identified but the effect on the growth can be readily interpreted. This evaluation was extended to other farming facilities and a coefficient of determination of -0.7 was obtained between the production rate per hectare and the plantation where lack of vinasse could be identified in 41 fields. We proposed a systematic approach to monitor the spreading of vinasse with Sentinel-2 images and Sentinel-1 radar images in an effort to increase efficiency.