Developing a Graphical User Interface to Improve Learning of Stochastic Theory for Water Resources i

1
Developing a Graphical User Interface to Improve
Learning of Stochastic Theory for Water Resources
in the Classroom

• Faisal Hossain, Jonathan Schwenk and David
  Huddleston
• Department of Civil and Environmental Engineering
• Tennessee Technological University

2
What is Stochastic Theory?

• Probability Theory
• Stochastic Processes
• Random Variables Random Processes (that can be
  described by a probability distribution)
• Each time the process acts it yields a
  different realization in time and space
• Repeatability is probabilistic

3
Why Stochastic Theory for Water Resources
Engineering?

• Uncertainty is omni-present in natural or
  man-made water resources systems.
• Need understanding of Random functions,
  Probability, Distributions, Time series, Spatial
  trends to model/predict the variability.
• Hydrologic models inherently uncertain
  (uncertain data, model assumptions, scale
  issues).
• ABET requires some concepts of Probability and
  Statistics as part of CE curriculum.

Nothing stays the same FOREVER in Water Resources
4
Examples of Application of Stochastic Theory in
Water Resources

• Reservoir systems analysis forecasting upstream
  inflow, power demand, navigation (optimization)
• Flood Forecasting Rainfall forecasting (time
  series analysis)
• Data assimilation in adaptive schemes for
  real-time decision making (kalman filtering)
• Improving model structure through reduction of
  uncertainty
• Spatial Interpolation of groundwater
  contamination (kriging)
• Flood frequency analysis (extreme value
  distribution)

5
Examples
SYSTEM
TEMPORAL
SPATIAL
6
Emerging Needs related to Stochastic Theory in
Water Resources

• More and more research conducted at graduate
  level involving stochastic theory applications
• Blooms learning level of entering graduate
  students should be understanding or
  application
• Major demand raised on admission criterion for
  graduate applicants
• Graduate students should be prepared a priori on
  the application of stochastic theory

7
Questions we should ask

• Are we doing a good job with instruction of
  stochastic theory in CE/Water resources?
• What do statisticians think?
• Are entering graduate students adequately
  prepared to conduct research involving this
  stochastic aspect in water resources?
• What could we do to improve learning of students
  in classroom?
• Could computer assisted schemes help? (e.g. GUI
  tools)

8
Instruction of Stochastic Theory(What do
Statisticians Think?)
For too long we in the statistics profession
have tolerated poor statistics teaching, which
produces courses that are often rated as the
worst course or the most useless course that
graduates in other fields claim they have ever
taken. We too often teach what appears to the
students a collection of unrelated methods
illustrated by examples taken from coin-tossing,
card-playing and dice-rolling. And then we expect
the students to be able to translate this wide
variety of methods with simple gambling examples
to complex industrial problems involving the
application of a large number of methods".
Godfrey, B. 1986. Future Directions in
Statistics. Report 10 Center for Quality and
Productivity Improvement, University of Madison,
WI, 34-39.
9
OBJECTIVES

• 1. Gauge the current state of instruction of
  Stochastic Theory in Civil Engineering curriculum
  (survey courses).
• 2. Proof of Concept of a GUI-based instruction
  tool for teaching stochastic theory in the
  classroom.

10
Stochastic Theory in Civil Engineering Curriculum

• Survey conducted using the world wide web only.
• Survey method search for keywords from
  course title and description.
• Keywords Stochastic, Probability,
  Numerical, Systems etc.
• ASSUMPTIONS
• Information posted by university course catalog
  or instructors website on the world wide web is
  accurate and up to date.
• All relevant course content information is
  available from the world wide web.
• All courses are actively offered on a routine
  basis by instructors.
• The course has a significant amount of stochastic
  theory component (or a nearest relative
  discipline) delivered as course content.

11
Stochastic Theory in Civil Engineering Curriculum
is calculated by dividing the absolute number
by the total number of courses surveyed (i.e.
241).
12
Preliminary Synopsis on Survey of Curricula on
Stochastic Theory in the Nation

• Current overwhelming representation of graduate
  courses perhaps underscores a current need to
  rethink strategies and strive for a more
  equitable distribution that would facilitate a
  smoother learning experience. For example,
  creating more undergraduate variants of these
  graduate courses and offering them early in a
  students CE education experience are likely to
  further strengthen the appreciation of the
  concepts on stochastic methods by the CE student.

13
Solution?Popularize Stochastic Theory using
Graphical User Interface (GUI) and Active Learning

• A picture is worth a thousand words - Confucius
• A picture is worth a million words if you can
  rapidly visualize the words Anonymous
• GUIs can rapidly visualize any way desired
  Ideal for active learning
• GUIs give full interactive control to manipulate
  and alter concepts and see the effect graphically
  almost immediately
• A lot of GUIs in mathematics education none
  exists (to the best of our knowledge) for
  stochastic theory in water resources education

14
STEVE Stochastic Theory Education through
Visualization Environment
Proof-of-Concept Can it work in a classroom
environment?
Core program is a Stochastic Model Two
Dimensional Satellite Rainfall Error Model
SREM2D SREM2D corrupts true rainfall using
various concepts of Stochastic Theory to simulate
satellite rainfall
SREM2D
15
STEVE Stochastic Theory Education through
Visualization Environment
Entity Dependence Diagram for STEVE GUI
Screen Shot of STEVE 1.0
Coded in Java Native Interfacing No O/S and
compiler requirement!
16
Using GUI to improve student learning
One example Geostatistics, Correlation lengths,
spatial clustering
1. Teach the theory concept of variograms, lag
distance, spatial correlation, modeling
variograms, correlation lengths, interpolation
(say kriging)

• 2. Next, allow students to use STEVE GUI
• Alter parameters on variogram model type
  (exponential)
• Alter correlation lengths (high, low, medium)
• Observe the effect on rainfall visualization
• Pose questions seek answers reconcile actual
  observation with expected observation through
  theory
• Improve learning through trial and error (rapid
  visualization is key to multiple iteration)

17
Using GUI to improve student learning
High Correlation Length
Medium Correlation Length
Low Correlation Length
CORRELATION LENGTHS IN RAINFALL
18
Using GUI to improve student learning
High Probability of Precipitation
Medium Probability of Precipitation
Low Probability of Precipitation
19
Conclusions

• Survey of current curriculum on Stochastic Theory
  in Civil Engineering reveals a dominance of
  graduate courses (84)
• GUI for rapid visualization of stochastic theory
  concepts to pictures has merit in improving
  student learning.
• GUI educational tools for Stochastic Theory in
  Water Resources engineering is absent.
• Technical and Software issues on GUI development
  need to be addressed.

20
Future Directions

• Improve GUIs computational aspect
  visualization, portability, software
• Survey (surveymonkey.com) of instructors on
  demand for such instructional tool.
• Prototype testing in summer research camps using
  a group of control and test students

ACKNOWLEDGEMENTS
This work was funded by the Department of Civil
and Environmental Engineerings DMF
Project Support received from Dr. Ambareen Siraj
on the development of STEVE GUI is greatly
appreciated. Derek Parsons of Computer Science
Department led the development of STEVE