July 22, 2016

Summer Research Institute has come to an end, library books may have been returned, PokemonGo Gym may have been conquered, and presentations have definitely been given.

Here's the proof: (And great job!)

Permeameters!

Time-lapse and Simulation Quantification!

We're DONE!!!!!!!!

Picture time!

July 21, 2016

Guest Appearances: Suzanne Wiese

Today Caleb and Nicole finished working on their presentation before going to lunch with all of the other groups in the Summer Research Institute. After lunch, Caleb and Nicole practiced the presentation exactly 10 times, one was with Suzanne Wiese in the audience.

For tomorrow:
Present
Put stuff back where it belongs

July 20, 2016

Guest appearances: Justin Keister

Today Nicole continued taking data from Humboldt and Nicole permeameter. We were only able to collect data from the Humboldt once because it ran out of water due to the pump's tubing falling out of place. We collected various data from the Nicole permeameter and were able to develop a hypothesis explaining why our K values are different from each head. We believe it is caused by vibrations from when we unscrew the head levels. Nicole also worked on her poster.

Caleb finished his first draft of the presentation and fixed an error in his code.

To do:
Not practice the presentation ten times

July 18, 2016

Guest Appearances: Justin Keister, Dr. Forrest Stonedahl, Kerry Rogers MD

Today Caleb and Dr. Forrest Stonedahl worked on fixing an error in the Marked Image Comparison code that determines how well the Blue Detection is working. The second run seems to be not going as well, and there seems to be a method that may work better for it, because of lighting changes in the Timelapse pictures. We might add that in as a special case if we have enough time to do so.

Nicole took Humboldt measurements and also set up the Humboldt and Nicole Permeameters for tomorrow, and has also worked a lot on her part of the presentation.

For Tomorrow:
Caleb will work on getting the thresholds for the program.
Caleb will work on his presentation.
Nicole will take measurements

July 15, 2016

Guest appearances: Beau Howes, Dr. Forrest Stonedahl, and Justin Keister

Today was a big day. After taking data from the Humboldt permeameter we started the ant farm. The ant farm required us to take Q data points EVERY TEN MINUTES. While waiting Nicole worked on graphing each day's data from the Nicole permeameter so that the slope gave us the K value for the day. From that, we were able to calculate the standard error, standard deviation, and average.

Beginning of ant farm

Lots of minutes into the ant farm

Dr. Stonedahl working hard

Beau helping us research

Device we used to measure Q

To do:
collect Humboldt data
work on presentation

July 14, 2016

Today Nicole continued to take data from the Humboldt and the Nicole permeameter. Dr. Stonedahl began setting up the ant farm so that it can be run tomorrow morning. We had a short research day today because we visited Augustana College to discuss our research and learn about other undergraduate research.

Group research picture taken after our weekly lunch 

Discussing our research for about the eight time

July 13, 2016

Guest appearances:

Justin Keister, Noah Moraski, Shawn Jackson, Dr. JJ Straton, Mackenziw Clawson, Paul O'Connor, Dr. Robert Mitchell, Dr. Shyam Seetharaman

We took more permeameter measurements and celebrated Justin's birthday, 

Nicole worked on figuring out statistics stuff and finishing bar graph.

Caleb worked on Augustana slides, marked images for round 3, started a really long program and then power went out before it could finish.

Giving a hydrology lab tour!

Free Pie day at Village Inn!

Tomorrow:

Get Antfarm ready for Friday!
run Humboldt
run Nicole backwards

July 12, 2016

Guest Appearances from: Justin Keister

Today Caleb worked on getting run two and run three setup for the program before leaving early to go spelunking and catch Geodudes. Run two is ready for the blue comparison program to find the parameters for it in the simulation comparison program, but the third run still needs a set of marked images before this can be done.

Today Nicole collected the final data points from the falling head permeameter. She also collected the last set of data from the Nicole permeameter, but continued to run the permeameter at the different head levels again to see if K decreased as time went on. We are still unsure if the decrease in K values are due to head differences or the actual amount of time the permeameter has been running (by the time we collect data from A the permeameter has been overflowing for three hours). Nicole and Dr. Stonedahl also collected data from the Humboldt permeameter.

Dr. Stonedahl with the ant farm

Dr. Stonedahl continued playing with the ant farm. We are tentatively planning on setting it up on Friday. Along with cleaning the ant farm, Dr. Stonedahl helped Nicole analyze our data. 

The relationship between volumetric discharge multiplied by the length between the head difference of the Nicole permeameter and the area times the change in head levels. The slope of the graph gives us our hydraulic conductivity (K).

To do:
Get powerpoint for Augie ready
Take more perm data
Continue data analysis
Work on presentation
Probably work on the program some more

July 11, 2016

Guest appearances: Cole Epping

Today Nicole was brought up to speed on everything she had missed. She and Dr. Stonedahl set up the Nicole permeameter, the falling head permeameter, and the Humboldt permeameter. We decided that we need one more day of data collection on both the Nicole permeameter and the falling head permeameter. Along with setting up each permeameter, Nicole started data analysis. She created many charts to go with the data we currently have. Right now we are searching for answers as to why the Nicole permeameter is giving us different K values at the different overflow spots.

Today Caleb wrote the code so that the top row does not go all the way up the window. This means that from a certain pixel we are now only looking left, right, and down now. We are doing this because our pictures had a reflection of blue at the top. Caleb also started looking at the specific spots where the program detected the points incorrectly to see if we can figure out other things that might be wrong with our method.

To do:
Go to star party
Take permeameter data
Continue looking at points
Work on presentation for Augie
Work on posters
Do art farm stuff

July 8, 2016

Today Caleb and Dr. Stonedahl took measurements from the Humboldt and Fallinghead permeameters, we were able to get three measurements from the Humboldt again, but on the first one it was giving a much higher reading on our h0 value, which ended up giving us a much lower k-value on that run.

Caleb also worked on marking images for the next set while Dr. Stonedahl ran new simulations for this set, because we have new k-values, and also did not have all of the images for the set, anyway. After this, we started looking at how the Timelapse pictures were matching with our marked pictures from the first run in each frame, so we could get a better idea of what could be causing the error. Such as the images being marked differently, or maybe lighting issues.

Figure for how often each point was marked incorrectly

For next week:
Work on presentation. (Finish presentation for Augustana)
Finish program.
Finish falling head measurements.

July 7, 2016

Today Caleb and Dr. Stonedahl worked on collecting data from the permeameters. We were able to finish collecting data for the Nicole permeameter, and also got measurements from the Humboldt permeameter at two different levels, without it stopping. The falling head and Humboldt are setup for tomorrow.

Caleb also looked at cutoffs and windows for the program, and also worked on setting up the data for the second run, while files were being backed up on an external harddrive.

We also brought the dry sand up from downstairs, and disposed of the sand with the broken glass in it.

July 6, 2016

Today's messages are brought to you by: ?
Guest Appearance from: Dr. Forrest Stonedahl

Today Caleb and Dr. Stonedahl began running the permeameters again, and once again the Humboldt permeameter stopped running after the first test.

The Humboldt Permeameter before it stopped running

Caleb also worked some more on the program, and when Dr. Forrest Stonedahl showed up, all three of us looked through the code and fixed some problems, and Dr. Forrest Stonedahl helped to improve it some more, so we will be able to see how much is being classified as a false positive or not being detected as blue. We'll be able to look more into cutoff values and window sizes later to see if we can get a better match. Currently the Light Sand is matching at about 97% and the Dark Sand is matching at about 94% correctly.

For tomorrow:
Continue running permeameters.
Look at windows and thresholds in program.
Bring sand upstairs.
Begin working on presentation.
Dispose of sand with broken glass in it.

July 5, 2016

Today's messages are brought to you by the letter: D

Today Caleb and Dr. Stonedahl setup the permeameters so that they will be ready to be used for collecting data tomorrow. Mostly Caleb worked on the program, and found a method for detecting the blue dye in the darker sand. This method uses the RGB colorspace, unlike the method for the lighter sand which we are using the HSV colorspace, so after modifying the code so that the new method would be able to work in the RGB colorspace, we ran the code used to compare the blue detection method to some marked images, we were able to get matches of 98.09% on the light sand, and 98.90% on the dark sand, so we should be able to use this method in the code where we are comparing the timelapse images to the simulated images.

For tomorrow:
Collect data from all three permeameters.
Run program to see how it has improved.
Move sand upstairs.

July 1, 2016

Today's messages are brought to you by the letter: h
Guest appearances by: Justin Keister and Noah Moraski

Today Caleb and Dr. Stonedahl setup and ran the two permeameters. The Humboldt one stopped running before we could get data from it the second time, again. The new glass cylinder for the falling head permeameter also showed up today. Caleb also worked on the program, so that it can use different methods of detecting blue depending on if it is in the dark sand or the light sand. Currently that hasn't yielded better results, but we need to look more into how to detect the blue in the darker sand.

The broken glass cylinder on top of the new one

For Next Week:
Continue running permeameters.
Continue working on program.
Remove Herobrine