A Novel Approach for Mapping Wheat Areas Using High Resolution Sentinel-2 Images

Abstract

Global wheat production reached 754.8 million tons in 2017, according to the FAO database. While wheat is considered as a staple food for many populations across the globe, mapping wheat could be an effective tool to achieve the SDG2 sustainable development goal—End Hunger and Secure Food Security. In Lebanon, this crop is supported financially, and sometimes technically, by the Lebanese government. However, there is a lack of statistical databases, at both national and regional scales, as well as critical information much needed in the subsidy and compensation system. In this context, this study proposes an innovative approach, named Simple and Effective Wheat Mapping Approach (SEWMA), to map the winter wheat areas grown in the Bekaa plain, the primary wheat production area in Lebanon, in the years of 2016 and 2017. The proposed methodology is a tree-like approach relying on the Normalized Difference Vegetation Index (NDVI) values of four-month period that coincides with several phenological stages of wheat (i.e., tillering, stem extension, heading, flowering and ripening). The usage of the freely available Sentinel-2 imageries, with a high spatial (10 m) and temporal (5 days) resolutions, was necessary, particularly due to the small sized and overlapped plots encountered in the study area. Concerning the wheat areas, results show that there was a decrease from 11,063 ± 1309 ha in 2016 to 7605 ± 1184 in 2017. When SEWMA was applied using 2016 ground truth data, the overall accuracy reached 87.0% on 2017 data, whereas, when implemented using 2017 ground truth data, the overall accuracy was 82.6% on 2016 data. The novelty resides in executing early classification output (up to six weeks before harvest) as well as distinguishing wheat from other winter cereal crops with similar NDVI yearly profiles (i.e., barley and triticale). SEWMA offers a simple, yet effective and budget-saving approach providing early-season classification information, very crucial to decision support systems and the Lebanese government concerning, but not limited to, food production, trade, management and agricultural financial support.

Sentinel-1 Data for Winter Wheat Phenology Monitoring and Mapping

“…Third, (3) on the already classified winter wheat plots in the West Bekaa plain of Lebanon, following a proposed classification approach by Nasrallah et al [24], smoothing and Gaussian fitting were applied on all the winter wheat classified plots (2017-2018 season). Eventually, in step four (4), after the analysis of the temporal profiles of the reference plots, each phenological phase in addition to harvesting (in West Bekaa) was mapped using the most appropriate configuration (polarization and incidence angle).…”

“…As for heading time (Figure 9), the estimated dates showed that around 73% of the wheat plots (Table 4) achieved heading between 2 and 26 April 2018. In the area of interest (the Bekaa plain of Lebanon), this period is a typical heading period [24]. The third mapped phenological phase after heading was the soft dough phase.…”

“…As storage period time increases, the risk of mycotoxins and fungus contamination increases and the need of grain fumigation during storage period would rise [23].Intensive field work and in situ campaigns are known to require large number of personnel, high logistical supply, machinery and equipment, and thus huge budgets, which increase with the increasing area of the study site. In this context, remotely sensing data provide information on the spatial distribution [24] in addition to monitoring crop growth by providing precise and timely information on the phenological status and vegetation development. From field to global scale, remote sensing tools can tackle these information, especially when coupled with crop growth models for assessing crop yields [25][26][27][28] and computing their water cycle budgets [29][30][31][32].By the route of vegetation indices, optical data (e.g., Sentinel-2, Landsat 8, and MODIS) have been widely used by seeking the interconnection between the plant optical properties and its photosynthetic activity characterized by the content of chlorophyll in the leaves [33].…”

Mapping Winter Crops in China with Multi-Source Satellite Imagery and Phenology-Based Algorithm

“…Regional crop-type mapping provides basic data for crop growth monitoring and yield forecasting [4][5][6]. These crop maps can assist decision-makers and end-users in identifying the crop areas and estimating the biomass production, irrigation needs, water productivity, and scheduling management strategies [7]. Winter crops are one of the major grain crops in China, as well as all around the world [8][9][10].…”

“…Winter crops are one of the major grain crops in China, as well as all around the world [8][9][10]. Obtaining a timely and precise map of the winter crop plantings is critical for China and the rest of the globe [7,11]. Global land cover products are available in China (e.g., 500 m Moderate…”