 | | |
| ENVT0852-2 | Data collection and processing
|
|
| Duration : | 36h Th, 36h Pr |
|
| Number of credits : |
|
|
| Lecturer : | Jean‑Jacques Boreux, Jacques Nicolas, Eric Pirard |
|
| Coordinator : | Jacques Nicolas |
|
Language(s) of instruction :
|
| French language |
|
Course contents :
|
| The lecture includes 5 chapters, given by 3 different professors.
Chapter 1 : Methodology of environmental data collection (Jacques NICOLAS-2 credits)
Chapter 2 : Sampling (Jacques NICOLAS-1 credit)
Chapter 3 : Geostatistics (Eric PIRARD-1 credit)
Chapter 4 : Mathematical and statistical tools (Jacques NICOLAS-1 credit)
Chapter5 : Introduction to the practice of the Bayesian calculation (Jean-Jacques BOREUX-1 credit)
Outline of chapter 1 : Definition of problem and goals of data collection. Concepts of global indicators and of signature. Preparation of instrumentation scheme, break down into different steps: selecting measurement variables, selecting methodology, selecting instruments, signal transport and conditioning, discussion of results.
Outline of chapter 2 : This chapter is one of the methodological features (chapter 1), it presents more particularly the challenge and theoretical aspects of sampling and their application for an environmental study: basic options of measurement campaign, preliminary site investigation, selection of measurement points in space and time. General principles of sample uptake.
Outline of chapter 3 : Spatial analysis of data. Notion and utilisation of variograms, stationarity hypothesis, deterministic and topo-probabilistic digital terrain models.
Outline of chapter 4 : Review of mathematical, graphical and statistical tools applied to environmental data and practised on data files issued from real studies.
Outline of chapter 5 : This part focuses on the risks for the environment management decisions. A real problem, the tragedy of Montroc, motives the student for Bayesian statistics analysis which the rational principles provide a innovative and relevant point of view in the management of environmental risks. This course provides an introduction to the graphical modeling which promotes a multidisciplinary approach and, thus, the emergence of realistic models. The computational difficulties inherent in this type of approach are extremely reduced by an introduction to Monte Carlo by Markov chain methods via free software WinBUGS. A few parametric statistical models common and useful in environmental sciences slash the nail. |
|
Learning outcomes of the course :
|
| The lecture objective is to provide students with very clear ideas on the various notions of data collection and processing as well as on parametric statistical model. The lecture aims at sharpen their critical sense on the uses of those tools in real studies, in particular in the frame of environmental risk management. |
|
Prerequisites and co-requisites/ Recommended optional programme components :
|
| The mathematical level required is the one of secondary education (e.g. deriving and integrating a single real function). Additional skills on elementary probability and statistics should be recommended. The student should be able to use basic computer tools, such as Excel. |
|
Planned learning activities and teaching methods :
|
| Workshop including presentation of actual cases of instrumentation schemes, application of sampling principles on noise measurement in Arlon's city, solving problems of increasing difficulties using Excel, WinBugs, Statistica or Arc View in the computer room or at home. |
|
Mode of delivery (face-to-face ; distance-learning) :
|
| The lecture is given in modules covering each a half-day of 4 hours (1 credit=3 half-days), in an independent way for each lecturer, but within a condensed time period. Practical works are included in the timetable and in planned credits. |
|
Recommended or required readings :
|
| Lecture notes available for the 4 first chapters. Available on-line for chapters 1, 2 and 4. Moreover, on-line interactive course for chapter 4 and reference book for chpaters 1,2 and 5.
Nicolas, J. 1998 Surveillance de l'Environnement-Méthodologie-Bruylant-Academia
References for chapter 5 :
Bernier, J., Parent, E., Boreux, JJ. 2000. Statistique Pour L'environnement . Traitement Bayésien de L'incertitude. Tec et Doc, Lavoisier.
Boreux, JJ., Parent, E., Bernier, J. 2009 (sortie en librairie en octobre 2009). Pratique du calcul bayésien, Springer Verlag France. |
|
Assessment methods and criteria :
|
| Each Professor queries in its own way, but the overall score is an average of evaluations of each chapter.
Knowledge of chapters 1 and 2 is found to be based on a concrete case by students, as well as presentation by a report of practical work. That of Chapter 3 is based on a question written on the matter. Chapter 4, the students deal with a concrete case proposed by Professor during a full day at the computer room. For Chapter 5, the evaluation is to resolve a given decision problem. To this end, the student comes with his laptop (WinBUGS) and the exam is open book. Eventually, an oral interview confirms the written review. |
|
Contacts :
|
| Jacques NICOLAS
Département des Sciences et Gestion de l'Environnement
Université de Liège-Campus d'Arlon
Avenue de Longwy, 185
B-6700 ARLON
Belgium
Tel: 32 (0) 63 23 08 57
Fax: 32 (0)63 23 08 00
Email: j.nicolas@ulg.ac.be
Jean-Jacques BOREUX
Département des Sciences et Gestion de l'Environnement
Université de Liège-Campus d'Arlon
Avenue de Longwy, 185
B-6700 ARLON
Belgium
Tel: 32 (0) 63 23 08 56
Fax: 32 (0)63 23 08 00
Email: jj.boreux@ulg.ac.be
Eric PIRARD
Département Argenco
Université de Liège
Bât. B52/3 Géoressources minérales & Imagerie géologique
chemin des Chevreuils 1
4000 Liège 1
Belgique
Tél. +32 (0)4 3669528
Fax: +32 (0)4 3669520
Email: Eric.Pirard@ulg.ac.be |
|
|
| |
