REMOTE SENSING METHODS

Remote sensing methods, combined with GIS, played a key role in advancing the identification of land use, particularly irrigated land. The UPV team utilized advanced GIS analysis and machine learning to develop and calibrate a model for automatic detection of irrigated areas, with efforts focused on refining algorithms to minimize identification errors.

Additionally, an eddy covariance sensor was acquired and deployed in a controlled plot to assess crop water requirements, enabling the calibration of water need algorithms through remote sensing data. The Spanish team also leveraged forecasts from the Copernicus Service to predict the pumping cap for the Requena-Utiel region, aiding farmers in planning their irrigation season in advance.

In the Portuguese case study, satellite imagery analysis was employed to improve the identification, characterization, and quantification of irrigated areas within the aquifer, while also monitoring changes over recent years. This analysis utilized satellite images from Landsat 5 (TM sensor) for data before 2013 and Landsat 8 (OLI-TIRS sensor) from 2013 to 2020, both with a 30x30m resolution. Two classification methods—supervised and unsupervised—were tested to classify and measure changes in irrigated areas. These combined efforts in both case studies enhanced the precision and reliability of the project’s outcomes related to groundwater management.