Pablo Santoro, pablorsantoro@gmail.com
Ruben Flecha, ruben.flecha@gmail.com
Juan Pablo Pilorget, jpilorget@gmail.com
Student Team:
YES
Tableau
R
Python
Approximately how many hours were spent working on this submission in total?
45 hours
May we post your submission in the Visual Analytics Benchmark Repository after VAST Challenge 2018 is complete?
YES
Video
https://www.youtube.com/watch?v=0ywtj_1bvr8&feature=youtu.be
Questions
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? Your submission for this questions should contain no more than 10 images and 1000 words.
There seems to be a trend upwards in some stations for several chemicals from 2015 onwards. This is also coincident with the increase in the amount of meetings observed in the information retrieved from VAST Mini Challenge 3.
As it is evident in the case of Methylosmoline in the next figure, stations Somchair and Kohsoom present significantly higher values of this chemical both comparing to previous dates and other stations.
To have a better understanding of the hydrological data we have ordered the stations according to the sub-basin to which they belong (e.g. Boonsri, Kohsoom, Busarakhan and Chai). Looking at the sub-basin near the approximate location of waste dumping, we can see an increase of Methylosmoline in Kohsoom and a station downstream, Chai, yet there is no clear causal relationship between them.
What is interesting is that the scatterplot allows us to appreciate the multiple double measures in the same period for the Chai station, which could be related to trying to lower the mean value of chemical concentration.
There is also a higher value for Boonsri station in late 2015. It is worth noting that, for the sake of clarity in the visualisation, we limited the y-axis to 57 so the increases in Chai and Boonsri could be visible.
An interesting feature that caught our attention was the sudden change in Methylosmoline values in Somchair station (upcoming figure) which used to have noisy and random-like behaviours in all its measurements. In this case, despite of the smoothing applied, it is clear that there was no change whatsoever in the measurements. To have a deeper understanding of this phenomenon would require further investigation and we will address this while analysing the sample sizes.
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? Your submission for this question should contain no more than 6 images and 500 words.
Sampling sizes and periodicity are one of the most relevant subjects since they could show anomalies in the sampling strategy that could hint a desire to hide changes in measurements. In order to visualise both units simultaneously we added to the tile plot an option to select an aggregation metric of samples (e.g. count, sum, average and median). By using different aggregation metrics to analyse the sampling scheme it becomes quite clear that the Hidrology Department isn’t collecting sufficient data to understand the comprehensive situation across the Preserve. Figure 2 (as numbered in the paper presented) shows how there is a sampling gap right in the moment before the Methylosmoline values soars. This not only happens in the two stations where there are higher values - Kohsoom and Somchair – but in almost every one. In the following figure we can observe the different patterns that arise by using different metrics to analyse the samples by month:
By simply looking at sample sizes count and median values it becomes quite clear the upwards trend in Methylosmoline concentration in both Somchair and Kohsoom station and we can rely on the scatterplot to have a more refined understanding of what the specific values are.
Also by looking into this patterns we notice several gaps in monthly sampling which relate to the second question of the Mini Challenge: it is clear that the Department is not collecting sufficient data from stations where Methylosmoline values are higher yet, in the same stations - as we could see in Figure 4 of the paper– other chemicals seem to be sample significantly more often.
Another issue that arises from this date ranges and relates to the second question is the different lifespans of chemical measurements. While some seem to be started in recent years (from around 2012) there are others for whom the values were discontinued long ago. This does not contribute to the clarity of the chemical readings, which could be correlated to the desire of hiding relevant information. Further investigation in this area is required..
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? Your submission for this question should contain no more than 6 images and 500 words.
It is mentioned that Methylosmoline presents a significant threat to the Pipits’ survival, so we will focus in this chemical for the remainder of this answer. In the case of Methylosmoline, in the stations Somchair and Kohsoom we can observe significantly higher values of this chemical both comparing to previous dates and other stations.
The visualisation has the advantage of fitting a LOESS regression so we can have a smoothed trend of the chemical concentration evolution In this case, the values of Methylosmoline appear to stabilise across 2016 following a soaring increase in the previous year. So, regarding the first question of the Mini Challenge, one could say that this could be a trend of possible interest in the investigation.
To have a better understanding of the hydrological data we have ordered the stations according to the sub-basin to which they belong (e.g. Boonsri, Kohsoom, Busarakhan and Chai). If we check that sub-basin, the one nearest the approximate location of waste dumping, then we can see an increase in Kohsoom and a station downstream, Chai, yet there is no clear causal relationship between them.
What is interesting is that the scatterplot allows us to appreciate the multiple double measures in the same period for the Chai station, which could be related to trying to lower the mean value of chemical concentration.
There also seems to be a higher value for Boonsri station in late 2015. It is worth noting that, for the sake of clarity in the visualisation, we limited the y-axis in 57 so the increases in Chai and Boonsri could be visible. One thing that caught our attention was the sudden change in Methylosmoline values in Somchair station which used to have noisy and random-like behaviours in all its measurements. In this case, despite of the smoothed trend we used previously, it is clear that there was no change whatsoever in the measurements. To have a deeper understanding of this phenomenon would require further investigation and we will address this while analysing the sample sizes and frequencies.