Case Study: Where's the Poop?

Source: data generated by a spatial model of the Neuse River created in the STELLA modeling language.   Model interface, data collection, and case study by Robert R. Gotwals, Morehead Planetarium and Science Center.

Introductory Readings

Background Science

Point source pollution of our streams and rivers is a significant problem in today's environment.  Certainly in the state of North Carolina (with apologies to the Illinois folks!), the amount of pollution entering water systems such as the Neuse River is of great concern to everyone who depends on this source of water.

The data from this model is more qualitative than quantitative, in that it represents a crude approximation of how pollutants are transported in a body of water, in this case a river flowing from Durham North Carolina,terminating in the Atlantic Ocean at the Pamlico Sound.  In this model, the Neuse River has been segmented into three segments:  a Durham segment, a Raleigh segment, and a Pamilco segment.  For purposes of keeping the dataset manageable, only three segments are identified.  In a more sophisticated model, the river would be segmented into a larger number of smaller segments.

At some point in this model, hog waste -- hence the title, "Where's the Poop" -- is being introduced into the river.  The amount of hog effluent in this model is small.  Depending on the flow conditions of the river, the hog waste is transported downstream, eventually finding its way into the Pamlico Sound.  

The model also divides the river into four vertical segments, each approximately one meter in depth.  As such, we have a two dimension dataset, with the three city segments (Durham, Raleigh, Pamlico) as one dimension, and the four depths (1m, 2m, 3m, and 4m) as the second.  
The model is run for 300 hours.  Data is collected on the total amount (in pounds) of hog waste at each of the 12 coordinates (city and depth) every ten hours.  As such, this is time-stamped data, and for the purposes of this case study, time represents the third dimension.

Objectives of the Case Study

In this case study, you will create a time-profile visualization of the hog waste distribution dataset.  Note:  the dataset is not particularly interesting due to its size, intentionally kept small to help visualize the raw and modified datasets.  

About the dataset

This dataset consists of 372 data points -- three segments, four depth measurements, and 31 time intervals (0 to 300, in units of 10).  The dataset is downloadable as a Microsoft Excel spreadsheet with two worksheets -- Raw data and AVSready.  A visual inspection of both worksheets should be undertaken.  Because of the small size of the dataset, simply looking at the raw data provides some insight into where the largest amounts of hog waste are located, and, where it isn't located.  

Building the Visualization

Data Retrieval

The data is downloadable as a Microsoft Excel spreadsheet from the Morehead web page.

Data Filtering

No data filtering is necessary.

Data Rendering

1. Save the AVSready worksheet as Text-tab delimited.  It is a good idea to put this file into your Clyde folder, more specifically into the data folder found in your Clyde folder.
2. Since this is continuous data, you want to use the 3DMesh tool.  In your Clyde folder, find the Clyde icon entitled "3DMesh".  When you double click on it, the 3D mesh set of windows opens.
3. At the top of the control panel, click on "Browse", then find the tab-delimited file. 
4. You need to know your dataset!  If you recall, there are three columns (corresponding to the three city segments), four rows (corresponding to the four vertical depths), and 31 planes (corresponding to the 31 10-hour intervaled time stamps).  Type those values into the appropriate boxes, making sure to hit the Enter key after each entry.
5. Click on the Read File button to get the data into Clyde.
6. Now you can manipulate the visualization to see what you like.  Some considerations:
1. turn off isosurface, color bounds.  You can leave on Show Bounds, this shows the time profile for the visualization.
2. Type in a title, something clever like "Neuse Poop".
3. Using the appropriate controls on the control panel, size your display, move it around, try to see which view is the best.
4. Consider viewing the second axis (Axis 2), and then the 31 different planes.
5. You might want to turn on the isosurface, and move the scroll bar to see if you can learn anything interesting from this feature!
   
7. You are now done rendering your dataset!
   

Analysis of the Visualization