Earth observation by satellite

Teacher(s)

Rousseau Raphaël;

Language

French

Main themes

Prerequisites : Notions of statistics, general physics course. The course has three components: 1: The presentation during lectures of the theoretical and methodological bases of remote sensing; 2: The application of image processing and interpretation methods to Landsat data on a region of Belgium, using image processing software on PC; 3: The exploration of a large range of remote sensing applications and of the methods used in each application, on the basis of CD-ROMs demonstrating case studies. Physical bases of remote sensing: - Definitions: radiant energy, radiant flux, radiant flux density, radiance; - Interactions between energy and the surface of the earth: laws of Stefan-Boltzmann and Wien. - Spectral reflectance curves ; - Atmospheric effects; - Physical interactions with thermal infra-red energy. The sensors used in remote sensing: - Landsat MSS and TM, SPOT; - AVHRR, Vegetation, MODIS; - the new high spatial resolution sensors. Image processing: - Corrections for non-systematic and systematic geometric distortions - Radiometric corrections - Extraction of statistics from images - Contrast enhancement - Spatial filtering - Supervised classification - Unsupervised classification - Classification errors estimation - Change detection methods - Multispectral transformations: Tasseled cap transformation; principal components analysis; - Notions of microwave remote sensing. Practical work: Processing of a Landsat TM image of Belgium: 1st session Introduction to image processing software 2nd session Color composites and contrast enhancement 3rd session Design of a scientific project 4th and 5th sessions Geometric correction 6th session Unsupervised classification 7th session Supervised classification 8th session Accuracy assessment of classification

Learning outcomes

At the end of this learning unit, the student is able to :
1	Knowledge : The students will acquire a good knowledge of the different applications of remote sensing, and a capacity to decide which sensors and which image processing and interpretation methods are most appropriate for a given application. Skills : The students will gain understanding of the bases of remote sensing and will be able to process and interpret satellite data on a given region, using a image processing software on PC. Emphasis is put on optical remote sensing for terrestrial ecosystem applications.

Content

Earth observation data play a central role in many fields, including the monitoring of environmental changes, natural resource management, disaster monitoring, and climate research. This course provides the scientific and methodological foundations required to use these data rigorously.
The course follows the complete workflow of a remote sensing analysis, from signal acquisition to the production and interpretation of geographic information. After an introduction to the physical principles (Part 1), students will be introduced to the main Earth observation systems (Part 2), and will then implement the main image preprocessing and analysis steps (Part 3) before exploring several environmental applications (Part 4).
Part 1: Physical Foundations of Remote Sensing

• Definitions: radiant energy, radiant flux, radiant flux density, radiance
• Interactions between energy and the Earth's surface: Stefan–Boltzmann and Wien's laws
• Electromagnetic radiation and spectral signatures
• Atmospheric effects
• Satellite orbits and operating principles
• Interactions between radiation and the Earth's surface in the thermal domain

Part 2: Main Sensors Used in Remote Sensing

• Optical sensors
• Radar
• LiDAR
• Hyperspectral sensors

Part 3: Satellite Image Processing

• Preprocessing: contrast enhancement, spatial filters, radiometric and geometric corrections, extraction and analysis of image statistics
• Image analysis: classification methods (unsupervised, supervised, object-based), random forest, change detection, spectral separability measures, time series
• Derived variables: Tasseled Cap transformation, spectral indices (NDVI, SAVI, …), temperature, biophysical variables
• Introduction to modern processing methods: Google Earth Engine, automation, reproducibility

Part 4: Applications of Remote Sensing
Practical Sessions
Session 1: Introduction to the software, satellite imagery, and image preprocessing. Session 2: Unsupervised classification. Session 3: Supervised classification. Session 4: Object-based classification. Session 5: Field data collection. Session 6: Accuracy assessment and improvement of classifications. Session 7: Multispectral transformations. Sessions 8 and 9: Change detection.

Teaching methods

This course unit is taught in French. It consists of complementary lectures and practical sessions. Attendance at both lectures and practical sessions is mandatory.
The lectures provide the theoretical foundations required to understand how remote sensing data are generated, processed, and used in various Earth observation applications. A mid-semester assessment will be conducted and will account for 10% of the final grade. Several practical applications will be presented by the students during the final course session.
The practical sessions will enable students to work directly with satellite remote sensing data on computers using ArcGIS Pro. This requires students to have assimilated the corresponding theoretical material.

Evaluation methods

The final grade is the weighted average of the theoretical examination (40%), the practical examination (45%), the mid-semester assessment (10%), and entry tests for selected practical sessions (5%).
A grade below 8/20 in either the theoretical examination or the practical examination results in a failing grade for the entire course unit. In the second examination session, students must retake both the theoretical and practical examinations. The final grade for the second examination session will consist of 50% of the theoretical examination grade and 50% of the practical examination grade.
The theoretical examination is a written examination consisting of true-or-false questions and open-ended questions requiring students to provide a detailed response. The practical examination takes place on a computer on the same day as the theoretical examination.
The mid-semester assessment is, like the theoretical examination, a written examination. However, it covers only the material taught up to that point in the semester.
Any use of generative artificial intelligence (AI) must be responsible and comply with academic and scientific integrity practices and with the RGEE (see, among others, Article 107). Students are asked to consult the document entitled Guidelines for the Responsible Use of Generative Artificial Intelligence in the Faculty of Science.

Bibliography

Le cours est principalement basé sur les deux références suivantes (des exemplaires sont disponibles en prêt à la Bibliothèque des sciences et technologie, BST). Les autres sources sont renseignées dans les slides du cours.

• Campbell J. B., Wynne R. H. et Thomas V. A. (2023). Introduction to remote sensing. Gulford, 6ème edition, 634 pages ISBN 978-1462549405.
• Richards J. et Jia X. (2013). Remote Sensing Digital Image Analysis, Springer-Verlag, 5ème édition, 494 pages. ISBN 3-540-64860-7

Teaching materials

• matériel sur moodle

Faculty or entity

> GEOG