Hydrological risks and land use development
Introduction to statistical hydrology and risk management (36 hr, 3 credits).
-
Lecture notes (last update: Sep. 2021, in French). Data (used in annexe D) are available here.
-
chapter 1 : Land use and risk.
chapter 2 : Climate and hydrological hazards.
chapter 3 : Risk management.
chapter 4, part 1 : Introduction to extreme value theory, fundamental in statistics and probability theory
chapter 4, part 2 : Introduction to extreme value theory, extreme value distributions and applications to hydrology.
chapter 5 : Floods in poorly documented watersheds.
chapter 6 : Flood and sediment transport.
Project: rainfall and water discharge inthe Lonza River (VS)
You have to study a watershed in Valais (Lonza, area of 77.8 km²) to ensure the protection of the village. You want to determine the design flood’s features. In this project, these features are deduced from the daily peak flow; a 100-year return period is selected. As you only have 25 years of rainfall data on the watershed, 56 years of river flow data, you have to be imaginative. A detailed statistical study is therefore required to arrive at the best possible estimate of peak discharges.
Projects are to be submitted individually. The project report can be written in one of the national languages (preferably French, but German and Italian are accepted) or in English or Spanish. The deadline for submissions is Friday, January 3, 2020, 12:00 p.m. (noon). Projects should be sent by e-mail in PDF format (no Word document that poses too many conversion problems will be accepted) to Mr Tomás Trewhela. Annexes can be provided in another format (xls sheet, matlab or R script, Mathematica notebook, handwritten notes, etc.).
For more information on the Lonza catchment area, please refer to the FOEN website. Rainfall and flow data for the project can be downloaded here. The project statement is also available for download.
Recommended books
- A.C. Davison, Statistical Models, (Cambridge University Press, Cambridge, 2003).
- S. G. Coles, An Introduction to Statistical Modeling of Extreme Values, (Springer, London, 2001).
See the lecture notes for further recommandations
Computational tools
-
Mathematica for which it is now possible (for students enrolled at EPFL) to have a license
Langage R.
Matlab and its clones (octave and Scilab). Exercises are made using Matlab.
- Jupyter notebooks offer the possibility to use Python, R, and Julia.