Modeling Earth Processes using Spreadsheets and Commercial Models

1
Modeling Earth Processes using Spreadsheets and
Commercial Models

• William Locke/David Mogk
• Dept. of Earth Sciences
• Montana State University, Bozeman MT

2
Why Modeling?

• Effective learning is experiential!
• Substitute arithmetic/mathematics for
• Space
• Loess deposition
• Time
• Thrust-belt evolution
• Safety
• Thermohaline circulation
• Money!

3
What do modelers do?

• A major purpose of scientific investigation is
  to describe reality through models Gauch, 1993,
  Am. Sci., 81, 468.
• ..what some researchers are calling the third
  branch of science - usingcomputers to simulate
  physical processes.. Pool, 1993, Science, 260,
  1077.
• ..predict the future state of a system with
  tolerable certainty Denning, 1990, Am. Sci.,
  78, 496.

4
Caveats(Denning, 1990, American Scientist, 78,
498)

• We must come to understand the domains over
  which a given model is reliable..
• Systems whose rules canchange in unpredictable
  ways are unlikely to have a reliablemodel.
• it is easy to substitute the model for reality
  and to confuse our opinions with scientific
  facts supported by the model.

5
Why Spreadsheets?

• Ubiquitous
• AppleWorks, Microsoft Works, Quattro, Lotus
• Visually Powerful Graphics
• 2-D (XY, Line), 3-D (Ribbon, Surface)
• Translucent - not Black Box
• point-and-click cell contents
• What-if?
• real-time graphical response to changing inputs

6
Modeling in ESCI 307 - Principles of
Geomorphology

• Assumption
• One application is Exposure
• Two applications are Familiarization
• Three applications are Competence
• Four applications are Mastery
• How far do you want to get this year?
• Mastery of spreadsheet skills - exposure to and
  familiarization with specific applications

7
LaboratoryMechanics

• Traphagen 100
• 10 computers (in pairs, out-facing) instrux.
• Teams of 2 or 4
• Face in for lecture, map work, specimens
• One instructor and a teaching assistant are
  generally required.

8
SCARP1 - Scarp Evolution

• Exposure to spreadsheet practice
• Block operations
• Graphing
• Theory of slope modeling
• Mass redistribution
• Running mean

9
(No Transcript)
10
Outcomes - SCARP1

• Way-cool graphics!
• I already knew that.
• Oops! I lost it!
• Its not real but Material eroded from the top
  accumulates at the bottom.
• It looks like chemical diffusion.
• Measure all of the possible variables and the
  slope responses.

11
Discussion - Scarp1

• Do you use spreadsheets, personally?
• If so, why - what do you find useful?
• Is a lab teaching spreadsheeting worthwhile?
• Is the graphical capability valuable for other
  applications than modeling?

12
SCARP2 - Scarp Evolution

• Familiarization with
• Block ops. Graphing
• Exposure to
• Absolute cell refs.
• Practice of slope modeling
• Acceleration of gravity
• Disequilibrium - Erosivity

13
Outcomes - SCARP2

• It looks just like SCARP!
• You didnt really include all of the variables.
• How do I
• add a resistant layer?
• mimic a deluge?
• model a stream at the toe?
• ?

14
Discussion - SCARP2

• Is this a good way to learn about the effects of
  gsin??
• Is this a good way to learn about probabilistic
  natural processes?
• Is it worth doing additional exercises?

15
LONGPRO - Stream Evolution

• Competence in block operations graphing
• Familiarization with absolute cell references
• Exposure to equations
• Interrelationships between modeling and
  collection of field data.

16
Outcomes - LONGPRO

• About 70 of teams succeed - matching real and
  modeled profiles (but with unreal parameters -
  e.g., a tiny output of m-scale particles).
• About 40 of teams succeed.
• Success requires changing one variable away from
  the optimal solution, then another one towards it.

17
Discussion - LONGPRO

• Is the integration of field data with modeling
  worth the investment of two laboratory periods?
• Is the failure rate (30 total, 60 partial)
  acceptable given the complexity of natural
  systems?
• Do those who dont succeed, fail?

18
GLACPRO - Glacier Reconstruction

• Mastery of block operations and graphing
• Introduction to research
• Team activity
• Resource allocation
• Interaction with other groups
• Quality control

19
Outcomes - GLACPRO

• A surprising number of results are incorrect.
• No intuitive knowledge of what a glacier should
  look like.
• Great difficulty meeting as teams outside of
  class.
• Inevitable team friction.
• Publishable products (when massaged).

20
Discussion - GLACPRO

• Is the difference between glacial extent and
  modeled glacial extent clear?
• Are team/interactive exercises worth the
  degradation of content?
• Are exercises with no right answer fair to
  students (as for grading purposes)?

21
Data Downloading

• Sources
• USGS
• Other?
• Formats
• Parsing
• Web Tables X
• Outcomes
• Trends
• Statistics

22
Log10 transformation
23
Magnitude/Frequency
24
Stream Rating Curves
25
Flood Recurrence Intervals
26
Discussion - Download

• What numerical data sets exist?
• Does spreadsheet manipulation result in increased
  learning effectiveness?
• Would a field trip to a stream gauge be a more
  effective teaching tool?

27
Virtual MODFLOW

• The standard in professional groundwater modeling
  - 750 (incl. academic discount!)
• The Scientific Software Group http//www.scisoftw
  are.com/
• Downloadable demo http//www.scisoftware.com/prod
  ucts/visual_modflow_demo/visual_modflow_demo.html

28
ModflowGrid
29
Discussion of Visual MODFLOW

• Is it appropriate to use a commercial product for
  a class activity?
• Is black-box modeling a valuable learning tool?

30
General Modeling Issues

• At what academic level is modeling critical?
• Relative importance of physical, theoretical and
  mathematical models?
• Optimal execution - homework, solo lab, team lab,
  class demo?
• Optimal tradeoff against content, field trips,
  map exercises, quizzes?

31
General Issues relating to DLESE

• Who should maintain such models?
• Author? DLESE? Other agency?
• How can DLESE promote this activity?
• Action Items?
• User Group?
• White Paper on modeling pedagogy?
• Testing Service for models?