The hydrology group is composed of two students, Nicole (Exercise Science) and Caleb (Computer Science), and Dr. Stonedahl. Each student has a different focus, looking at two components of the project. To read about our daily exploits, please visit: Our group’s blog.
Nicole is focusing on constructing a cost effective, less heavy PVC pipe permeameter than the one built in the Spring of 2016. A permeameter is an apparatus used to measure hydraulic conductivity of a porous material as fluid flows through it. Hydraulic conductivity is directly proportional to how fast water flows through sand, which is important for modeling water flow under rivers and streams. She plans to compare and evaluate each of the permeameters: 4” PVC pipe (built Spring of 2016), 2” PVC pipe (constructed this summer), falling head permeameter (used most often in the field), and Humboldt permeameter (accepted permeameter).
We have begun taking data from the 2” PVC pipe permeameter, but have run into various problems that had made our data for the day incomplete. Our procedure requires the 2” PVC pipe permeameter to sit overnight with the sand and water in it to ensure the sand settles. Also, before every data collection our protocol requires a constant head of water for an hour which is created by an overflow of water. After that hour we collect approximately 45 mL of water from the head and record the time it takes. We take five readings from each head level. The 2” PVC pipe permeameter has three different head levels. After the first head level data points have been collected, we then open the next head level and wait another hour until we proceed. We continue this process until we have collected five readings from each of the three head levels. Once we have completed the readings for the day we empty the permeameter and clean out the sand. Then we make sure all of the head levels are close so that we can fill the permeameter with water. After the permeameter is filled with water we are able to add sand to the apparatus. Having a clear PVC pipe in the 2” permeameter allows us to better monitor the sand level. Once it is filled, we let it sit overnight.
Nicole has also been working on collecting data from another method called a falling head permeameter. This method is usually performed in the field. However, we are performing the falling head permeameter test in a 9”x24” glass vase filled with 77 pounds of wet sand. A clear tube with a ruler attached to it is placed inside of the sand at a depth of approximately 17 inches below the water level. Like the 2” PVC pipe permeameter, the falling head must sit overnight to let the sand settle. When we are ready to collect data we add water to the top of the vase to replace the water that has evaporated overnight. Then we fill the clear tube with water above the 31 inch mark. We start our stop watch when the water level reaches the 30” mark and stop our stopwatch when the water level reaches 27”. Using the time it takes the water to fall, along with other measured system characteristics collected from the apparatus we are able to calculate the hydraulic conductivity of the sand.
For the first two weeks of this summer, Caleb has been working on a program to help quantify parts of the data from last summer’s hydrology research. The program will take specified frames from the Physical and Simulated runs of last year’s data, and compare them in each cell to determine whether or not specified areas are matching.
Currently the program is comparing four windows of pixels each centered on the midpoint of the center of the grid cell and each corner.
We tried using RGB data, but have now switched to having the program use the hue, saturation, and luminosity values of the pixels in each window, and compare them to the initial images to determine if they are now considered to be blue. If they are blue, the corresponding bit in its bitmap will be changed from a 1 to a 0 to indicate the change, and then the Physical and Simulated bitmaps can be XOR’d together to show which bits are matching and which are not. This final map can be used to compare the Physical and Simulated runs to determine how close the two are matching.