View source for Cs382

This page documents the work of CS382 - Scientific Computing, Fall 2007
----
==enVision Tabletop Groundwater Simulator==

General Instructions
* Setup 
* Teardown and cleaning
* Packing and travelling

Instructions for Demonstrations
* First one
* Second one
* etc.

==Computational Groundwater Simulations==

=== Fitz, Bryan and Mikio===


[[Image:gws.jpg|thumb|right|Confined aquifier simulation]]
[[Image:parabolic.jpg|thumb|right|Parabolic contaminant flow model]]

*Experiments
**Demonstrating porosity
***model water flow unconfined aquifier
**Illustrating groundwater flow in a confined aquifer
***We will use a cellular automata model where at the lowest level, a cell is either fresh water or contaminated. We see this problem split into two concepts - speed and direction. 
****Direction: The illustration to the right demonstrates our assumptions about how the water will move through the material. The simulation will calculate a new direction at each generation based on it's position relative to the known locations of water input and output.
****Speed: Remains constant throughout generations for a given run. The "speed" value represents a combination of speed of water flow and material porosity, and in terms of the simulation is the possibility that a a neighboring cell in the flow direction becomes contaminated.

**Describing recharge, transition and discharge areas
***modeling behavior of water recharge, discharge in wells, lake, etc

*Computational Tools
**C
***+Very fast
***+Libraries are available
***+Good distributed Libraries
***-Potentially difficult to use
***-no graphics libraries
**Netlogo
***+Fancy Graphics
***+Fun to use
***+Available examples/code
***-Slow
***-Small problem size
***-No Distributed processing

=== Peter and Mikio ===
* Experiment
** Describing the model
*** Describing the various parts of the Groundwater Simulator by attaching tags: Key words -- wells, artesian wells, lake, underground storage tank, septic tank, springs, vegetative layer, river/ocean, recharge area, discharge area, aquifers, confining layer, clay layers
** Illustrating and Calculating Porasity of different types of earth materials
** Determining how it is easy for ground water to move in different earth materials.
* Computetional Tool
** NetLogo for computatinal experiment

=== Brad and Nate ===

Our goal is an incremental approach towards illustrating groundwater contamination in a confined aquifer.  The confined aquifer, viewed between wells 1 and 8, offers an environment within the groundwater simulator with the fewest variables.  The first 4 experiments are an effort to illustrate the behavior and underlying science that must be understood and demonstrated in the final experiment.

*Experiments
** [[Cs382/Diffusion_Experiment|Diffusion]]
*** Show diffusion without groundwater movement.
** [[Cs382/Flow_Rate_Experiment|Flow Rate]]
*** Show the leading edge of groundwater contamination as a indicator of flow rate (related to section 5 and 13 in manual) 
** [[Cs382/Plume_Length_Experiment|Contaminant Plume Length]]
*** Determine whether contaminant plume length is affected by flow rate for a given amount of dye
** [[Cs382/Soil_Density_Experiment|Soil Density]]
*** Use displacement method and measurements of aquifer component to determine the density of the soil. We can use this value in silico.
** Illustrate laminar flow in a confined aquifer (Activity 7-1)
*** Show laminar flow between wells 1 and 8.

* Computational Tools
** NetLogo
*** + Visualization built in
*** + Agent and cell based simulation structure built in
*** - Possible limitation on world size / agent count in RAM
*** - Possible run time slower than groundwater simulator at higher flow rates
*** - Not parallel
** Python and MYMPI
*** + Parallelizable
*** + Faster than NetLogo in serial code ?
*** + Visualization software exists
**** TKInter - easy to install; seemingly easy to use
*** - Visualization software must be integrated
*** - MYMPI is untested
**** Need to compile stuff.

=== Plume Tracking ===
* Setup
** physical simulator setup approximately 16 inches away and perpendicular to the line of sight of a web enabled camera.
** A script was used to capture output of the output of the camera from the server at a rate of one every two seconds. A faster rate may be possible, but the current script did not have time to get the image and rename it within a 1 second interval.
* Procedure
** set pump flow rate at maximum and allow water table to equalize
** start image capture script
** inject a full pipette bulb into well number 1
** remove pipette before allowing bulb to reinflate
** allow simulator to run for approximately 5 minutes or until the majority of the dye in the system has been discharged
** stop image capture script

We did three complete runs, each with a different dye colors.  We used blue, purple and green because we thought they would give the most contrast for edge detection.