Team Members:

1.       Benjamin Weidner || Student Researcher || weidnerb6@students.rowan.edu || Primary Contact

2.       Dr. Bo Sun || Doctor/Supervisor || sunb@rowan.edu

Student Team

Analytical Tools:

1.       Tableau 2018.1

2.       Unity 2018.1.1f1 (64-bit)

3.       Oculus

Estimated Work Hours:

-          6/6 –  6/8: 25 Hours

-          6/11 – 6/15: 25 Hours

-          6/18 – 6/22: 25 Hours

-          6/25 – 6/28: 25 Hours

-          7/2 – 7/6: 20 Hours

-          7/9 – 7/13: 20 Hours

-          Estimated Total Hours: 140 Hours

Figure 1

Public Accessibility: Granted

Video: https://www.youtube.com/watch?v=RP6wjnQZc7U&feature=youtu.be

Questions:

1. Characterize the past and most recent situation with respect to chemical contamination in the Boonsong Lekagul waterways. Do you see any trends of possible interest in this investigation?

Many trends of possible interest exist within the past and recent situations with respect to the chemical contamination in the Boonsong Lekagul waterways. While many of the chemicals tested had steady trends, certain measures had much more drastic changes in some years and locations over the other measures.

One instance of the said drastic changes was the chemical measurement Iron. Iron was tested in every location of the Boonsong Lekagul waterways, and was tested from the years 1998 – 2015. Although the measurement wasn’t tested in 2016 like many others, a spike in readings can be seen from 2002 – 2003. After 2003, the readings level back down to similar testing prior to this specific year. The visualization to the right (figure 1) shows a timeline of the Iron measurement trends (left) as well as a calendar view (right) of the specific year in question. Both visualizations show measurement through the size of the square, however the calendar view shows two numbers on the squares: The topmost number shows the measurement in the chemical’s respective units, and the bottommost number shows the number of records that were taken during that year. Even if Iron is not harmful to the waterways, this spike shows some form of change within the environment during 2003.

Another instance in which the data seemed to spike was the chemical measure Aluminum. Aluminum was only tested in five of the ten locations, and was only tested from 2008 – 2014. However, a visible jump in the readings occurred from 2008 – 2009, then sank back down during 2010. This trend was very similar to that of the Iron spike in 2003, the measures increased for one year but then immediately died down during the next year. The visualization to the right (figure 2) shows the readings of Aluminum over the years it was tested. In the timeline graph (left) it is seen that 2009 had much larger readings than in the other years. In the calendar view (right) the squares that represent aluminum are highlighted to show that, even though their readings weren’t as high as many other chemicals, they had a presence during the time that their readings spiked in 2009. Any noticeable spike is a point of interest within the data set, no matter the comparison to the other measures.

Bicarbonates were tested in every location from 1999-2016. In 2010, the only location that it was tested was Boonsri, Kohsoom, and Tansanee. Other disparities in the data are present, in 2012 & 2013 test were held only in the same three locations, and no readings were found from 2014 – 2016 in Busarakhan and Chai. The graph below (figure 3) shows the trends within the Bicarbonates readings. Spaces that are void of a square show that there are no readings for that given year or location. The smallest squares shown are readings that have the smallest value in the given set, whether it be zero or not. The data shown below shows that the trends of Bicarbonate readings have remained consistent over the years, however, the absence of certain readings could have left out vital information on possible spikes in a specific location or year.

Figure 2

Figure 3

Total Dissolved Salts was tested in every location, but only tested from 2005 - 2016, however, it seemed to be the dominant chemical in terms of value throughout each of these years. Any increase or decrease was gradual throughout a few years. It’s important to note these gradual trends to compare to any other environmental changes to see if there is a correlation (for example, temperature change, human activity, etc.). The graph below (Figure 4) shows the dominance of the chemical compared to the other readings. The number displayed portrays the number of recordings taken, the pink square represents the Total Dissolved Salt measure. The absence of data from prior to 2005 could raise some questions as to if this chemical was dominant in those prior years as well. The absence of data seems to also be a present trend within the given dataset.

2. What anomalies do you find in the waterway samples dataset? How do these affect your analysis of potential problems to the environment? Is the Hydrology Department collecting sufficient data to understand the comprehensive situation across the Preserve? What changes would you propose to make in the sampling approach to best understand the situation?

Many of the anomalies found within the dataset involve some form of missing data. Whether it be a chemical that was untested in a location, or in any given year, there weren’t many measures that were taken in every given year and location. Of the 106 provided chemical measures, only 38 were tested in every single location, but only 23 of those same chemicals were tested in every year and location. So only about a fifth of the chemicals were tested in every location and every year. This absence of data could leave out important information that could tell of other chemical spikes and dips in absent years and locations. The visualization below (figure 5) shows the measures of the 23 chemicals over the years throughout every location. These measures can be accurately analyzed due to the consistency of their recordings.

Looking again at the visualization above (figure 5), it is seen that the locations Achara, Decha, and Tansanee did not begin testing until 2009, as there is no information displayed for any year prior to 2009 for those given locations. For whatever reason, these three locations weren’t tested at the same time as the others. This leaves a hole in the data, a portion of the story that is missing in those given years.

Another anomaly presented within the missing data was the trends of the chemicals taken at the locations Boonsri and Kohsoom. Many other chemicals were tested within those locations; however, a few specific measures were exclusively taken in Boonsri and Kohsoom. The visualization below (figure 6) shows the chemicals that were specifically tested in those two locations. All chemicals listed were tested in the years 2008 and 2009, except for fluorene which was tested exclusively in 2009 in the two locations. Other information could be told if these chemicals were tested more often, and hidden trends could be present due to the absence of said data.

Much of the data that has been presented seemed to have missing parts, and the question is raised; what if recordings were made, but the values of said measures were zero and therefore not recorded at all in the data? This question was disproven by looking at the data itself. The graph below (figure 7) shows the measures of Lead and its recordings on a day-by-day basis. The number atop represents the specific measure value and the bottommost number represents the number of recordings. As it can be seen, multiple recordings have been taken with a value of zero. This is just one example that missing data is missing data, and not a value of zero.

The graph below (figure 8) shows a specific shot of the three locations that did not begin testing until 2009. Some recordings do indeed have a value of zero, and no data was left out in any recording that was taken.

3. After reviewing the data, do any of your findings cause particular concern for the Pipit or other wildlife? Would you suggest any changes in the sampling strategy to better understand the waterways situation in the Preserve?

Regarding the concern to the Pipit and other wildlife living in the preserve, no information is given about which chemicals may harm the wildlife and which will not. With the spike of Iron measurements in 2003, the alarming rate of chemicals is concerning, however no information is shown that supports that an increased amount of iron in the water will do any harm to the wildlife. Regardless of if it harms the environment, these spikes and dips are something to keep an eye out for.

The sampling strategy could be improved to not only understand the trends of the chemicals in a better way, but to understand the geographic relationship these chemical readings have with the waterways around them. Below is an image (left) of the waterways provided with the data set (figure 9.1). Highlighted are the rivers that seem to be connected in their own ways. The far-left river houses the Decha location, the next holds Tansanee, the next travels through Sakda and splits off to meet Somchair and Achara, and the final river passes through Kannika to Chai, branching off to find Busarakhan and Kohssom, and finally reaches Boonsri towards the top of the map. All the rivers seem to converge towards the bottom of the map, however the picture is cut off before the rivers can either diverge or converge. If these rivers are connected, it could explain some of the patterns in the data provided. The next image below (middle) is a map of the 2003 Iron readings (figure 9.2, which had a noticeable spike. The circles in these geographic maps represent the values of the measurements in the given year. The measures between each river remained consistent (except for the three locations not tested), and it is seen that in the rightmost river, the readings seem to be smaller as the river travels upward. The final image (right) shows the same map but representing Total Dissolved Salts in 2016. (figure 9.3). Some of the values, like Achara, Kohsoom, and Busarakhan, have a reduced value and are either farther north or branched off the main part of the rivers they reside in. If all these rivers come from one source, it would be easier to analyze the trends of these chemical contaminations. Based on the information below, the rivers most likely do come from a common source, and based off of the chemical readings being more minute in the northern part of the map, the rivers most likely flow upward as well.

To better understand the waterways situation in the Preserve, more concrete data is necessary. The missing information from above does put holes in the story the data is trying to tell, and will every chemical being tested every year it can be monitored so that no random spikes happen like with the Iron and Aluminum readings. Another piece of information that could help to better understand the waterways is more geographical information. A compass for directional input, a larger map to show how the rivers connect more in depth, and perhaps an elevation scale to see if the chemicals are influenced by the flow and location of the rivers. With more data, a much more thorough analysis can be made.

4. VR Visualization Prototype

The following images are examples of the VR prototype that’s being developed to show the information in a much more comprehensive way.

Figure 10.1

An example of the Leap Motion Controller’s hand tracking technology portrayed in Unity.

Figure 10.2

When the palm faces the user, a control panel appears. This control panel contains a year slider (top), chemical slider (bottom), add (blue) and remove (red) chemical buttons, detach (white bottom) and reattach (white top) from the hand buttons, and to the right is a visualization of the waterways maps with buttons (green) on each of their respective geographical areas. This map is attached by the blue object between the control panel and the map, an anchorable object (blue) within the anchor (white wireframe).

Figure 10.3

An example of the detached control panel.

Figure 10.4

When the green buttons are pressed, a bar graph for each area appears above the panel. The time and chemical slider can alter the time being viewed on the information panel below the bar graph.

Figure 10.5

When the anchorable object (blue) is pulled away from its anchor and placed into the space around the user, it expands into a 3D map with landmarks representing each location. Each colored circle is a representative chemical, and the size represents their value during any given year. The control panel changes functions, where the blue button will add as many chemicals as the user picks out to represent, and the time slider affects the time of this map as well.