Salt-affected soils represent a major constraint to agricultural productivity and environmental sustainability in many regions of the world. Effective monitoring and management of soil salinity require advanced approaches capable of capturing its spatial and temporal variability. Recent advances in geospatial technologies—such as remote sensing, LiDAR, cloud-based platforms like Google Earth Engine, and machine learning—have significantly enhanced our ability to map and analyze soil salinity dynamics. However, there remains a need to strengthen technical capacity among researchers, practitioners, and decision-makers to effectively apply these tools in real-world contexts.

This webinar series aims to address this gap by combining theoretical foundations with practical applications, enabling participants to develop skills in multitemporal and three-dimensional soil salinity mapping for improved soil management and decision-making.

The third webinar covers machine learning approaches for analyzing spatial and temporal variability of soil salinity, including model evaluation, variable selection, and prediction with error and uncertainty quantification.

AGENDA

16:00–16:05 – Opening remarks

Jorge Batlle-Sales, Chair, International Network of Salt-Affected Soils

16:05–16:25 – Introduction to machine learning in soil salinity

Mario Guevara Santamaría, National Autonomous University of Mexico; Monica Aviles, Autonomous University of Baja California; Arnau Riba, University of Almeria, Spain

16:25–16:45 – Model evaluation and variable selection

Mario Guevara Santamaría, National Autonomous University of Mexico; Monica Aviles, Autonomous University of Baja California; Arnau Riba, University of Almeria, Spain

16:45–17:10 – Prediction with error and uncertainty quantification

Mario Guevara Santamaría, National Autonomous University of Mexico; Monica Aviles, Autonomous University of Baja California; Arnau Riba, University of Almeria, Spain

17:10–17:25 – Questions and answers

Moderated by Maria Konyushkova, FAO

17:25–17:30 – Concluding remarks

Jorge Batlle-Sales, Chair, International Network of Salt-Affected Soils

Speakers and trainers

Mario Guevara Santamaría, National Autonomous University of Mexico

Dr. Mario Guevara (UNAM–Geosciences) is a leading expert in digital soil mapping (DSM) and transdisciplinary sustainability research. Over the past five years, he has authored 21 high-impact peer‑reviewed articles and 4 book chapters, while establishing the first UNAM‑accredited DSM program in Spanish, training 126 specialists across 12 countries. He currently supervises six graduate students at UNAM and co‑advises two international PhD candidates. His work integrates Open Data Science with strategic global partnerships—such as the FAO Global Soil Partnership—to improve soil data quality and accessibility, advancing sustainable land management worldwide.

Monica Aviles, Autonomous University of Baja California

Dra. Mónica Aviles (UABC-Soil Science) is a soil scientist and academic leader specializing in soil science and regenerative agriculture. As a professor and researcher at UABC, she leads international projects focused on soil salinity, carbon sequestration, and nitrogen dynamics in arid zones. Her research centers on the water-soil-nutrient nexus, strategies for the efficient use of fertilizers, and the optimization of irrigation in arid and semi-arid climates. She leads transdisciplinary initiatives for the recovery of saline agricultural systems, training technicians and producers. A member of Ibero-American cooperation networks, she works on soil health as a fundamental pillar of climate resilience for sustainable food security.

Arnau Riba, University of Almeria, Spain

Ing. Arnau Riba is an agronomist specializing in precision agriculture, remote sensing, and digital soil mapping. With a background in agricultural water management and biophysical modeling, his research spans diverse environments, including an academic exchange at Aarhus University (Denmark) and recent fieldwork in the arid zones of Almería, Spain. Arnau has experience utilizing both proximal and remote sensors for environmental assessments and energy balance data. His recent technical contributions include supporting digital soil mapping initiatives in Ecuador. He holds an MS in Precision Agriculture and focuses on applying spatial data to agricultural and soil-related challenges.