An Operational Framework for Mapping Irrigated Areas at Plot Scale Using Sentinel-1 and Sentinel-2 Data

In this study, we present an operational methodology for mapping irrigated areas at plot scale, which overcomes the limitation of terrain data availability, using Sentinel-1 (S1) C-band SAR (synthetic-aperture radar) and Sentinel-2 (S2) optical time series. The method was performed over a study site located near Orléans city north-central France four years (2017 until 2020). First, training non-irrigated plots were selected predefined selection criteria to obtain sufficient samples each year. is based on two irrigation metrics; first one SAR-based metric derived from S1 series second optical-based NDVI (normalized difference vegetation index) S2 data. Using newly developed event detection model (IEDM) applied all in VV (Vertical-Vertical) VH (Vertical-Horizontal) polarizations, weight calculated plot. series, maximum value achieved crop cycle considered as metric. By fixing threshold values both metrics, dataset constructed Later, random forest classifier (RF) built year order map summer agricultural into irrigated/non-irrigated. classification uses plots. Finally, proposed validated real situ collected results show that, procedure, overall accuracy reaches 84.3%, 93.0%, 81.8%, 72.8% 2020, 2019, 2018, 2017, respectively. comparison between our approach RF directly showed that nearly similar obtained classifiers with not exceeding 6.2%. analysis accuracies precipitation revealed higher rainfall amounts during crop-growing season (irrigation period) had lower (72.8% 2017) whereas encountering drier very good (93.0% 2019).