Accurate and up-to-date land cover information is fundamental for sustainable natural resource management, climate change adaptation, disaster risk reduction, biodiversity conservation, agriculture, urban planning, and environmental monitoring. As landscapes continue to change due to rapid urbanization, deforestation, agricultural expansion, and climate-related impacts, governments and organizations require efficient and scalable methods to monitor land cover dynamics and support evidence-based decision-making.

Recent advances in cloud computing, Earth Observation (EO), and artificial intelligence have transformed the way land cover mapping is conducted. Google Earth Engine (GEE) provides an open-access, cloud-based geospatial analysis platform that enables users to process and analyze petabytes of satellite imagery without requiring high-performance computing infrastructure. When combined with Machine Learning (ML) algorithms such as Random Forest, Classification and Regression Trees (CART), Support Vector Machines (SVM), and Gradient Boosting, GEE enables rapid, accurate, and reproducible land cover classification and change detection across local, national, and regional scales.

Recognizing the increasing demand for digital geospatial skills, the Asian Disaster Preparedness Center (ADPC) has developed this three-day training course to strengthen participants' capacity in applying Google Earth Engine and machine learning techniques for land cover mapping. The course provides participants with practical knowledge of satellite data processing, image classification, accuracy assessment, and land cover change analysis using cloud-based geospatial technologies. Through lectures, demonstrations, and hands-on exercises, participants will gain the knowledge and skills required to produce reliable land cover information that supports disaster risk reduction, climate resilience, ecosystem management, and sustainable development.

The training aims to strengthen participants' technical capacity to use Google Earth Engine and machine learning techniques for efficient land cover mapping and geospatial analysis. By the end of the training, participants will be able to:

• Understand the fundamentals of Earth Observation, remote sensing, and land cover mapping.

• Gain practical knowledge of the Google Earth Engine cloud computing platform.
• Access, process, and visualize multi-source satellite imagery for land cover applications.
• Understand machine learning concepts and commonly used classification algorithms for land cover mapping.
• Develop land cover classification models using Google Earth Engine.
• Perform accuracy assessment and validate land cover classification results.
• Apply change detection techniques for monitoring land cover dynamics.
• Explore applications of land cover information in disaster risk reduction, environmental management, and climate resilience.

The training is designed as a three-day face-to-face capacity development program (09:00–17:00 hrs daily) that combines technical presentations, guided hands-on exercises, case studies, and interactive discussions. The course follows a progressive learning approach, beginning with the fundamentals of Earth Observation and Google Earth Engine, advancing to machine learning-based land cover classification techniques, and concluding with change detection, accuracy assessment, and real-world applications.

Participants will work with openly available satellite datasets and cloud-based geospatial tools to gain experience in developing land cover maps and interpreting classification results. Throughout the course, regional case studies and practical examples will illustrate how geospatial technologies support disaster risk reduction, natural resource management, and climate adaptation.

The program emphasizes learning and peer exchange through regional case studies, interactive discussions and hands on sessions.

GROUP ASSISGMENT

There will be group work component during the training. It will be conducted based on case study reflecting around the training topic.

Day 1: Introduction to Earth Observation and Google Earth Engine

Theme: Building Foundations for Cloud-Based Geospatial Analysis

The first day introduces participants to the principles of remote sensing, Earth Observation, and land cover mapping. Participants will gain an understanding of satellite imagery, commonly used EO datasets, and the capabilities of Google Earth Engine as a cloud-based geospatial analysis platform. The sessions will familiarize participants with the GEE interface, JavaScript environment, data catalog, and basic image visualization and processing techniques.

Day 2: Machine Learning for Land Cover Classification

Theme: Applying Artificial Intelligence for Land Cover Mapping

The second day focuses on the application of machine learning algorithms for land cover classification using Google Earth Engine. Participants will learn the concepts of supervised classification, training sample preparation, feature selection, and the implementation of commonly used machine learning algorithms. The sessions will also introduce accuracy assessment techniques and best practices for producing reliable land cover maps.

Day 3: Change Detection and Applications

Theme: Monitoring Land Cover Dynamics for Sustainable Development

The final day focuses on land cover change detection techniques and the application of classified maps to support disaster risk reduction, environmental monitoring, and climate resilience. Participants will learn methods for comparing multi-temporal datasets, identifying landscape changes, and communicating results through maps and visualizations. The sessions conclude with discussions on emerging technologies, including cloud computing, artificial intelligence, and future directions in Earth Observation.

The course follows a "Code-Along" pedagogical model, emphasizing tactile learning through real-time script development.

• Hands-on Lab Sessions: The core of the course. Each theoretical concept is immediately followed by a guided practical exercise in the GEE Code Editor.
• Interactive Technical Briefings: Short, high-impact presentations on the logic of machine learning and the physics of spectral bands.
• Problem-Based Learning: Participants work in small "Sprints" to debug code and solve common classification challenges, such as class imbalance or spectral overlap.
• Clinics & Peer Review: Daily "Script Clinics" where instructors provide one-on-one troubleshooting for participants' specific regional datasets.
• Visual Demonstration: Live walkthroughs of complex workflows, from data ingestion to the final export of a GeoTIFF.

•National and sub- national environmental agencies
•Scientists, researchers, and analysts
•Meteorological and hydrological department staff
•Urban planners and environmental engineers
•Academics andstudents
•Development professionals and practitioners

$1,140 (without accommodation)
$1,464 [with accommodation (4 nights)]

Fees are inclusive of course materials (soft copy), cost of instructions and course certificate. For face-to-face training, fee is inclusive of morning and afternoon snacks and lunch during the course.

Interested individuals and organizations can register online at www.adpc.net/apply.

For more information about the course, you may also contact ApibarlBunchongraksa at apibarl@adpc.net and telephone numbers +66 22980681 to 92 ext. 132.