The theory and practice of time series modelling are
presented for systematically studying observations sampled over time from water
resources, environmental, economical, energy and other kinds of systems. By
understanding, for example, how inputs to a given system dynamically affect
various outputs, better decisions can be made regarding the design and
operation of the system.
The rich variety of time series models that are defined,
explained and illustrated in the course include ARMA, nonstationary ARIMA, long
memory FARMA, seasonal ARIMA, deseasonalized, periodic, transfer function-noise
(multiple inputs - single output), intervention and multivariate ARMA (multiple
inputs - multiple outputs) models. Extensive hydrological, water quality, environmental,
and other applications are given for clearly demonstrating how the various
kinds of time series models can be systematically and conveniently fitted to
real world data sets by following the identification, estimation and diagnostic
check stages of model construction. Moreover, a major emphasis of the course is
the use of exploratory data analysis graphs, intervention analysis,
nonparametric trend tests and regression analysis in the detection and
estimation of trends in environmental impact assessment studies. Other topics
covered include time series analysis in decision making, estimating missing
observations, simulation, the Hurst Phenomenon, forecasting experiments and
causality.
At least one university course in probability and
statistics.
Course assignments, a project selected according to a given student’s interest, and a final examination will account for 10%, 30% and 60%, respectively, of the final grade.
Every Monday during the fall term from 1:30 p.m. to 4:30
p.m. in E2 1307C.
Please purchase a paper copy of the overheads from Graphic
Services in Room 2022 CEIT (Centre for Environmental & Information
Technology) and kindly bring this material to every class so you can write your
own additional notes directly on your paper copy.
If you have any questions please contact the class
instructor, Professor Keith W. Hipel, in the Department of Systems Design
Engineering (Room DWE-2518B; 519-888 4567, extension 32830;
kwhipel@uwaterloo.ca). Students are most welcome to take the course for credit
or audit. Certain components of the courses will be tailored to meet the
background and interests of the students.
Time Series Modelling of Water Resources and Environmental
Systems, by Keith W. Hipel and A. Ian McLeod, published by Elsevier, Amsterdam,
1994 (ISBN: 0 444 89270-2).
To download a copy of the book please visit:
http://www.systems.uwaterloo.ca/Faculty/Hipel/Time%20Series%20Book.htm
The McLeod-Hipel Time Series (MHTS) Package constitutes a
flexible DSS for carrying out comprehensive data analysis studies in order to
obtain meaningful statistical results upon which wise decisions can be made.
Each student will be able to use the MHTS package in course assignments and a
project consisting of data analyses of time series chosen from the field of
interest of the student.
Kindly refer to the website of Prof. A.I. McLeod at:
http://www.stats.uwo.ca/faculty/aim/epubs/mhts/default.htm
A rich variety of time series data can be found at:
http://www-personal.buseco.monash.edu.au/~hyndman/TSDL/
Other statistical software are available at:
http://ist.uwaterloo.ca/ew/software/math_and_statistical/index.html