ISSS608 2016-17 T3 Assign Jan Patrick Mabilangan Insight

VAST Challenge 2018: Suspense at the Wildlife Preserve

ISSS608_2017-18_T3_Assign_Jan_Patrick_Mabilangan_Gosioco

Visualizations and Insights

Chemical Trends

Patterns	Visualization
1. General Pattern across the Years An overview of the water sensor readings across the years will give us a starting point and rough idea of how the waterways are faring. Figure 1.1A shows the average water sensor reading values by year of all the measures and locations. Immediately the Iron content of the water in the year 2003 becomes apparent having the highest value among all the measures throughout the years. By excluding this record, figure 1.1B zooms into the rest of the measures, showing Total Coliforms, Total Dissolved Salts, Bicarbonates, and Total Hardness as having the next top values after Iron. Next, in figure 1.1C, the water sensor readings are divided by location. The values of the measures form patterns that can be observed throughout all the locations: (1) Consistently Present Measures such as those of Amonium and Nitrites are consistently present throughout the years in all the locations. Values may vary with different measures across time but the measures seem to always be present in the readings. (2) Cyclical Measures such as those of Water Temperature and Dissolved Oxygen form a cyclical pattern. Values of the measures have a recurring increase and decrease across time. (3) Present for a certain period of time Measures such as those of Alachor and Aldrin are observed to be only present in a period of time or in certain location. Measures of these pattern may hint at contamination that will be further examined. (4) Rarely observed Measures such as those of Naphthalene and PAHS are rarely observed. Only a few instances were recorded by the water sensors. Measures of this pattern may still hint at a possible contamination though with such little data, it would be difficult to prove otherwise.	800px
2. Measure Insight - Aluminum Contamination Water sensor readings have shown Aluminum to be present in 2008-2010 in 2 of the waterways: from Busarakha, Chai, to Kannika and from Somchair to Sakda. In all 5 of the locations, Aluminum value rose to over 400 ug/l in the month of April 2009. Aluminum values were observed to be the greatest in April 17, 18, and 25, all within the span of a week and occurring on the same day for pairs of the locations.
3. Measure Insight - Total Chlorodine and Hardness Groups Total Hardness and Chlorodinine have been observed to have very distinct group of values before and after a certain period across the locations in the waterway. Total Hardness is shown to have a pattern of: (1) Average values (before 2010) (2) Period of missing or low values (2010-2014) (3) Higher values (2014 onwards) Total Chlorodinine is shown to have a pattern of: (1) Higher values (before 2016) (2) Lower values (2016 onward) Such consistent trends merit interest and are possible for further investigation as to the change in general average of the measures/properties of water.
4. Measure Insight - Growing presence of Arsenic In figure 1.4, Arsenic has been observed to have increasing value in the recent years across most of the locations in the reserve and as far is the data is concerned, has shown no signs of stopping. Further investigation as to the possible cause of such continuous growth would be advised as to keep it from getting into dangerous levels.
5. Location Insight - Kohsoom Fecal Fears While several of the locations have their share of fecal coliforms, possibly excrements of the wildlife in the area, Kohsoom has significantly more traces of fecal streptococci, peaking at 330mg/l in April 27, 2011, seen in figure 1.5. Streptococci is a strain of bacteria that can cause urinary tract infections, meningitis, and respiratory infections to animals. While the values for the measures have dropped ever since, it would be wise to further study the contamination event, as the bacteria present in the animal feces is indicative of the animal’s health and they may have indeed been infected at the time.
6. Location Insight - November 22, 2014 in Tanasee As seen in the figure 1.6A, among the locations, Tanasee was observed to have a recent period of high ammonium that peaked with 15 mg/l on November 22, 2014. After examining the rest of the data for Tanasee around that date, figure 1.6B shows that Chlorides, Nitrites, and Total Dissolved Salts have also been shown to have the same pattern of heightened increase. This suggests that contamination may have occurred prompting the increased values for the measures.
7. The incident of August 15, 2003 After earlier identifying Iron as having the top value of all the measures that were read, figure 1.8A was created to see the trend of Iron across all locations. With the exception of Achara, Decha, and Tanasee, locations that have just recently had traces of Iron, Iron had been consistently low for Boonsri, Busarakhan, Chai, Kannika, Kohsoom, Sakda, and Somchair. However, a spike in the value of Iron is observed among the latter group of locations. On average, without the spike, Iron values for these group of locations is around 1 mg/l. For Boonsri, the value had jumped to 30mg/l on October 15, 2002, 30 times more. For the rest of the locations, all having the spike on the same day, August 15, 2003, the value had reached 5 digits ranging from 17,970 in Chai to 37,959 in Kohsoom. Taking the sum of the values on the spike, the iron observed on the day accounts for 99.07% of total iron values collected from the reserve. Investigating further on August 15, 2003, figure 1.8C shows that values for Chromium and Copper have also spiked during the same day for the locations. This is especially true for Kohsoom, having the peak of all 3 measures on the same day. While we cannot say entirely for certain the cause, the correlation of such values seem to suggest a contamination incident around the said date.

Dataset Anomalies

Patterns	Visualization
1. Multiple Readings and Varying Values As seen in figure 2.1A, several waterway sensor readings were taken for the same measure on the same day and the same location. The dataset did not specify whether these readings were from different sensors in the same location, a retake on the reading at a different time, or some error on data collection. Regardless of the case, such readings could affect the accuracy of analysis as there are instances, such as those in figure 2.1B, where the values greatly differ. Taking one or the other or an average for multiple of these values would influence the correctness of interpretation.
2. Incomplete Location Data The waterway sensor readings data collection starts differently for the locations, making it especially difficult to compare the last three with the rest of the locations. The Gantt chart in figure 2.2 highlights the beginning of data collection for each location. a) Boonsri, Chai, and Kannika have data since January 1998 b) Kohsoom, Sakda, and Somchair have data since April 1998 c) Busarakhan has data since October 1998 d) Achara, Decha, and Tansanee have data January 2009
3. Missing data for specific years Aside from data not being present initially, there are also missing data in throughout he years for measures that are seemingly consistent. Such gaps can occur consistently across locations and measures, such as those between 2012 and 2014 in the example of figure 2.3B. But there are also those that have no particular pattern such as those found earlier in figure 1.1C. Missing data would best be rectified to have a complete historical view of the waterways to examine.
4. Inconsistent Frequency of Data Collection Inconsistent frequency of data collection, like those in figure 2.4, make it difficult to compare measure values between locations and form distinct patterns, since values in one measure may be frequent enough to form a line, while values in another may be infrequent enough to just form some points. Increasing the frequency of data collection will allow the Hydrology Department to have a much clearer understanding of the data. In figure 2.4b, the increased frequency of recording for water temperature in location Chai in 2016 has lead to a more defined increase, peak, and decrease.

Wildlife Hazard

Patterns	Visualization
1. Methylosmolene Found While soil samples taken from the dumping site have been inconclusive in detecting Methylosmolene (the toxic manufacturing chemical that was suspected to have been dumped by Kasios), figure 3.1a shows that the chemical has made its way to the waterways of the preserve. While traces of Methylosmolene have been present in 2014 in multiple locations, a significant increase in value is observed in Kohsoom and Somchair starting January of 2016. For Kohsoom, the value had started off gradual at around 50 ug/l, had spiked to 145 ug/l on August 26, 2016, and then gone back down to 50 ug/l afterwards. For Somchair, the value had been consistent 130.5 ug/l. Considering the location of the Kohsoom in relation to the approximate location of waste dumping, seen in figure 3.1b, it may not be as far a stretch to say that the recent excavation and building activities going on is a cover-up for Kasios to dump the soil into the waterways. Since Somchair is further away, the consistent 130.5 ug/l may be the work of a vehicle with a consistent load limit. Of course, this is but a speculation. While the true source of the measure may not be determined until further investigation, the existence of Methylosmolene in the waters of the preserve causes danger to the wildlife.
2. Other Toxins and their Safety Levels Aside from Methylosmolene and Fecal Streptococcus that have already previously been shown to be hazardous to wildlife, other measures in Section 1 General Pattern across the Years are also toxic given the excessive amounts observed. The table in figure 3.2a recaps the measures, their potential health effects, and prescribed safety levels. Data for the table was taken from the Water Quality Association International Headquarters and Laboratory.(https://www.wqa.org/learn-about-water/common-contaminants) Figure 3.2b shows the measures with the prescribed safety levels as a reference line. From it, we can see values for the different measures have exceeded the safety levels multiple times in different locations, causing additional wellbeing concerns to the waterways situation.
3. Sampling and Understanding Figure 3.3 shows the Methylosmoline values across all years, focused on the location Somchair. From this, we can see that there is missing data between October 2015 and January 2016. This is where sampling ties in with understanding. If data were present on the gap, the Hydrology Department would have been alerted earlier of the existence of the toxic chemical, be able to monitor the growth of the values – either abrupt or gradual, and possibly respond sooner to the situation.