Remote Sensing of Global Croplands for Food Security

Monitoring of global croplands (GCs) is imperative for ensuring sustainable water and food security for the people of the world in the Twenty-first Century. However, the currently available cropland products suffer from major limitations such as: (1) Absence of precise spatial location of the cropped areas; (b) Coarse resolution nature of the map products and their significant uncertainties in areas, locations, and detail; (b) Uncertainties in differentiating irrigated areas from rainfed areas; (c) Absence of crop types and cropping intensities; and (e) Absence of a dedicated web-based data portal for dissemination of the cropland map products. This research aims overcome the above mentioned limitations through development of a set of Global Food Security-support analysis data at 30 m (GFSAD30) resolution consisting of four distinct products: Cropland extent/area, Crop types with focus on the 8 types that occupy 70% of the global cropland areas, Irrigated versus rainfed croplands, and Cropping intensities: single, double, triple, and continuous cropping. These products are produced using multi-resolution time-series remotely sensed data and a suite of automated and\or semi-automated cropland mapping algorithms (ACMAs). Data include Moderate Resolution Imaging Spectroradiometer (MODIS) time-series, and Landsat Time-series from various epochs. Methods include spectral matching techniques (SMTs), automated cropland classification algorithms (ACCA’s), decision tree algorithms (DTAs), and linear discriminant analysis algorithms (LDAA). Massively large big data (MLBD) are computed over several platforms that include parallel computing over NASA NEX supercomputers, and Google Earth Engine (GEE). Large volumes of ground data are sourced through various crowdsourcing mechanisms and integrated on a web platform: croplands.org.https://commons.und.edu/ss-colloquium/1056/thumbnail.jp