Abordagens inovadoras de monitorização da salinidade do solo em áreas agrícolas

Soil salinization is a major abiotic constraint affecting food production, particularly in arid, semi-arid, and sub-humid regions. However, the large-scale quantification of salt-affected areas remains challenging. AGROSALT seeks to address this research gap by developing monitoring frameworks for the rapid, non-invasive, and cost-effective assessment of the spatial and temporal distribution of salt-affected areas at both field and regional scales. Despite significant advancements in recent decades, challenges persist in proximal and remote sensing applications for soil salinization assessment. These challenges include issues such as data resolution, spatial and temporal coverage, acquisition costs, data processing and storage, as well as the effectiveness of solutions considering the diversity in agricultural systems in areas affected by either primary or secondary salinization processes. Therefore, the overall objective of the AGROSALT project is to develop reliable monitoring approaches for assessing soil salinity in agricultural systems using both proximal and remote sensing technologies. The specific objectives are: 1. To develop and implement robust monitoring approaches for soil salinization assessment in the selected agricultural regions of Lezíria Grande and the Roxo irrigation district in southern Portugal. This involves considering the different processes influencing soil salinization (primary and secondary salinization) and the distinct characteristics of agricultural systems (annual vs. permanent crops) in these areas. 2. To test and compare different monitoring approaches for assessing soil salinization, including proximal sensing, unmanned aerial vehicles (UAVs)-based sensors, and satellite-based methods. 3. To conduct extensive field data collection, surveying soil salinity and relevant soil properties in the study regions, which will serve as a ground-truth dataset for validating the tested approaches. 4. To build robust models relating proximal and remote sensing data with ground-truth dataset using multivariate statistics and/or machine learning algorithms. 5. To identify the main advantages and limitations of each approach used in the field-scale and/or regional-scale assessment of the spatial and temporal distribution of salt-affected areas. 6. To disseminate research findings and tools developed throughout the project to relevant stakeholders, including agronomic and environmental sectors, governmental agencies, and institutions responsible for soil and water resource management in Portugal and beyond. AGROSALT will evaluate the performance of various proximal and remote sensors, along with different approaches, for assessment of soil salinity. Proximal sensors to be tested include electromagnetic induction-based devices such as EM38 (Geonics Ltd, Mississauga, CA), CMD Mini Explorer 6L (GF Instruments, Brno, CZ), and DUALEM-1h (Dualem Inc., Ontario, CA). UAVs-based sensors include those based on multispectral data and the GEM-2 UAV (Geophex Ltd, Raleigh, NC, USA) electromagnetic induction-based sensor. Remote sensing methods will involve the utilization of multispectral and thermal imagery, as well as synthetic aperture radar available in Sentinel and Landsat satellite sensors.