Entry Name:
"PKU-Yuan-MC2"
VAST
Challenge 2018
Mini-Challenge 2
Team Members:
Wentao Zhang, Peking University, zhwt@pku.edu.cn PRIMARY
Hao Wu, Peking University, 1500012715@pku.edu.cn
Guozheng Li, Peking University, guozheng.li@pku.edu.cn
Xiaoru Yuan, Peking University, xiaoru.yuan@pku.edu.cn (faculty.advisor)
Student Team: YES
Tools Used:
Tableau
Excel
Pixel River: Interactive Visulizing and Exploring Large Number of Time-series Data through Pixel-based Approach developed by students of the MC2 Group in Visualization Lab of Peking University, and adapted by the student team for the challenge.
Approximately how
many hours were spent working on this submission in total?
200 hours.
May we post your
submission in the Visual Analytics Benchmark Repository after VAST Challenge
2018 is complete? YES
Video
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.
A) The Overall Information:
To comprehend the overall information concerning the chemical contaminations in
the Boonsong Lekagul waterways, we provide a pixel-based overview for time-series with
all the hydrological data collected. In the overview, each row is the whole
series of a measure at one location, and every value is encoded as a pixel of different colors. Red stands for values quite higher than average;
grey for values around average; blue for much lower than average. (technically speaking, we have applied the three-sigma rule of thumb)
In every interval, the color changes gradually. Thus, we can explore the sample data in the waterways by interactions.
B) The Past Situation:
i) Heavy Metals Pollution Around 2000
Figure Q1-1
The Chemicals in frame are Biochemical Oxygen, cadmium, chromium, gamma-Hexachlorocyclohexane,
lead, p,p-DDT, Potassium, sulphates. Their patterns are roughly the same, with a slight rise
around 2000 and a drop around 2001. From the overall distribution, these chemicals'
peaks of value lie in the range from 1998 to 2002, and values are lower than average after 2002.
Among those chemicals, most are heavy metals. Therefore, we make a conclusion that 20 years ago,
the waterways in the Preserve were polluted by heavy metals but the pollution situation was almost under control after 2002.
ii) Herbicides and Pesticides Pollution Around 2006
Figure Q1-2
The chemicals in frame are Aldrin, alpha-Hexachlorocyclohexane, beta-Hexaxchlorocyclohexane,
Dieldrin, Endosulfan (alpha), Endosulfan (beta), Endrin, Metolachlor, p,p-DDT,
p,p-DDE, p,p-DDD, Simazine. Their patterns are roughly the same, with a slight rise around
2005 and a drop around 2006. Besides the similar distributions, all of those chemicals
have the same start and suspend time of sampling. Among these chemicals, most of the contaminations are herbicides,
pesticides or their ingredients. What's more, when the sampling of herbicides and
pesticides ended, some new chemicals started to be sampled at the same time.
These chemicals mainly fall into three categories. The first one is
harmful carcinogen in coal tar; the second one is related to new kinds of
herbicides and insecticides; the last one is probably used as rodenticides.
Using pesticides or herbicides in the Preserve is legal, so sampling the data
of pesticides may be a part of detection of whether or not the amount of usage in
the Preserve exceeds the upper limit. Thus, it explains the reason why the sampling
of a kind of pesticide exactly started just as another sampling suddenly stopped.
We conjecture that the Preserve changed the kind of pesticides they used.
After 2009, the Preserve stopped sampling the pesticides to detect the exceeding.
Thus, we speculate that the Preserve may turn to the non-polluting pesticides
or other deworming methods.
iii) Periodic Pattern of Water
Figure Q1-3
Two series in frame have a periodic pattern on the overview graph. These two measures are water temperature
and oxygen dissolved in water. Their cycles are close to one year, which is consistent with common
knowledge that such measures are mainly determined by the climate and other natural influences. From the
distinguishable changes in water temperature, we conclude that the Preserve has four distinct seasons.
C) The Recent Situation:
In recent years, most of the sample values are stable and
under the average values. In other words, the amounts of most chemical contaminations
are within the normal range. However, from the end of 2014 to 2016, there were
strange changes in three chemicals: Chlorodinine, AGOC-3A and Methylosmoline. The sample values
of Chlorodinine and AGOC-3A stayed high before 2016, but they fell dramatically at the end of
2015 in all sampling sites. On the contrary, Methylosmoline, which is the toxic
manufacturing chemical in the suspected dumping, stayed low before the end of 2014,
and started increasing gradually first at Chai, but at the end of 2015 the sample
values of Methylosmoline decreased at Chai, Boonsri, Busarakhan, Sakda and Kannika,
but increased dramatically at Kohsoom and Somchair. Combining the location of Somchair
and Kohsoom with the location of waste dumping, obviously Kohsoom is the closest
monitoring station to the waste dumping place, which somehow explain why the chemical
contamination increased so abnormally.
Figure Q1-4
Line chart of Methylosmoline
Figure Q1-5
Line chart of Chlorodinine and AGOC-3A. The color of line is consistent with the color of the location spot in the map below.
However, there is an interesting relationship between Chai and Kohsoom concerning the amount of Methylosmoline. Just as the
amount of Methylosmoline at Chai started to decline, values at Kohsoom began to
increase steeply. And what is more important is that Kohsoom is on the upstream of Chai. Such coincidence raises our
suspect that probably Kasios Furniture Company at first tried to dump waste nearby
Chai, and then moved to a more upstream place near Kohsoom to avoid being discovered.
So we import the map graph to help understanding of the upstream and downstream information and the recent situation of waterways.
What's more, the decrease of amount at the end of 2016 at Kohsoom may also be
related to the dumping, since the Kasios may have changed their waste dumping
place to another location. As for Somchair, it is built on a totally different
river, and somehow a little far away from the waste dumping place. So the situation
around Somchair is worth further investigating.
Figure Q1-6
By circling out the over-polluted monitor sites on the map, the recent situation is discovered
that the southwest part of Preserve have seldom pollution. However, the further northeast one goes, the more serious the
pollution situation is. This situation is consistent with the given information that
the waste dumping location is on the northeast corner of the Preserve. On such
basis, we conjecture that the Kasios may change their waste dumping point to a
less polluted and less concerned place to avoid being discovered their illegal
and immoral behaviors.
D) Some Interesting Trends and Findings:
i) There are some Herbicides(Alachlor, Metolachlor, Simazine) and
insecticides(Aldrin, Dieldrin, Endosulfan(alpha&beta), Endrin, Heptachlor,
Hexachlorocyclohexane(alpha, beta and gamma, gamma-Hexachlorocyclohexane
is called Lindane), p,p-DDT, p,p-DDD, p,p-DDE ) which have the same trend.
Figure Q1-7
The Stockholm Convention banned 12 POPs, nicknamed "the dirty dozen". These
include: aldicarb, toxaphene, chlordane and heptachlor, chlordimeform, chlorobenzilate,
DBCP, DDT, "drins" (aldrin, dieldrin and endrin), EDB, HCH and lindane, paraquat,
parathion and methyl parathion, pentachlorophenol, and 2,4,5-T. This took
force on 17 May 2004. Since 2006, use of alachlor as a herbicide has been banned in the European Union.
In 2007, international steps were taken to restrict the use and trade of endosulfan.
Figure Q1-8
DDT was banned in most developed countries before 1990, and around 2000 in developing
countries. After that, malaria rates of the developing countries that gradually
decreased its DDT use had large increases, due to the lack of good Alternatives.
In 2006 WHO reversed a longstanding policy against DDT by recommending that it be
used as an indoor pesticide in regions where malaria is a major problem.
Simazine and Metolachlor have never been banned, but they show similar trends to
previous chemicals. It seems that people's environmental awareness rose
sharply at that time.
ii) By sorting the overview by location, we find a fact that there are three stations built up
about 10 years later than other monitor sites. They are Tonsanee, Decha, and Achara. And the number of
chemical categories measured at these three stations are also less than others. We find that the sample of herbicides and
pesticides are hardly measured at these three sites. Therefore, we conjecture that the southwest corner of the Preserve, where
these three sites are located are still undeveloped or under-developing areas. In this way, it explains why the Kasios
may change their waste dumping place to Somchair or somewhere nearby.
Figure Q1-9
- 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.
A) Anomalies:
i) Outliers: Iron, Coliforms, etc.
Figure Q2-1
Besides Iron, Coliforms and Manganese, there are some other possible anomalies: copper, fecal coliforms, zinc, aluminium.
At Chai, the sampling value of iron is more than 20,000 on Aug. 15th, 2003. But
it turns to 0.19 in the next day. The unit of iron sampling value is mg/L, so the
concentration is 0.32-0.67mol/L. Because of the pH of nature water is 6-9, the
concentration of iron ions in rivers cannot be 0.3 mol/L. Since the standard of
nature water is less than 0.3mg/L, So these can be considered as wrong values.
Just like the leveling effect in chemistry, the other values are pulled to the
same level, which makes us unable to know the details of other dates.
ii) Several Samples One Day
Sometimes there are more than one sample a day at a single location,
such as the situation shown in the table below.
Figure Q2-2
Here, the numerical value 4.175000 and 4,23100 are close to the peek value
throughout the recent three years, which 0.65900 is only a little higher than
the average value of the amount of Methylosmoline. Such tremendous change in
value reveals that the toxic chemical contamination is dumped on
this single day 2015/4/18. Thus, this kind of anomalies are of great value of analysis to us.
iii) Strange Days
Figure Q2-3
B) Insufficient Data:
The Hydrology Department have not collected sufficient data to
understand the comprehensive situation across the Preserve in
two aspects respectively.
i) Lack of spatial scale:
Somchair (considered as a new dumping location by us) and the nearest downstream monitoring
station Sakda are too far away. The long time drifting within the river can degrade
the toxic chemical contamination, so that it becomes hard to discover useful information at Sakda.
Besides, Decha and Tansanee are the only monitoring station on the each river. It
is absolutely not possible to have a comprehensive enough understanding of the
whole river with only the sample data at single place.
To sum up, the scarcity of station leads to data insufficiency.
ii) Lack of time scale:
As mentioned in the Question one,
the set up time of every measuring station differs. Three places were built up about 10 years later than other
monitor stations. And these three stations, Tansanee, Decha, and Achara, only occupy about 20% of all the measures.
From the overview , it is obvious that only half of all the weeks has sampling values. Like the herbicides and insecticides
mentioned before, we do not know the situations after 2009. Under the circumstances of huge absence of samplings, sometimes
there will even be two consecutive days of sampling. Thus, the distribution of the sampling days is irregular too.
As a result,
we cannot make reasonable analysis of comprehensive situation across the Preserve.
C) Changes in the sampling approach:
Actually, we discover that as time goes by, the sampling become more frequential
and regular. Still we suggest that observers should take a more scientific
sampling approach. The sampling cycle should be regulated as per week or per day
at all the locations. Besides, with respect to different chemicals, the sampling
frequency should be determined by the dissolution rate and degradation rate.
When an anomalies occurs or the measure data is deviant, the sampling frequency
should correspondingly increased to one time a day or several times a day,
so that we can have a more comprehensive information about the situation
happening at that point.
- 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.
A) Changes in the Sampling Strategy
Considering all the data are hydrological and the measuring locations are along
the river, thus the information of upstream and downstream are crucial for the
understanding of the waterways situation. We suggest that more measuring stations
should be added on the upstream and downstream of the east two rivers, since the
other two rivers that Decha and Tansanee are on seldom have pollution. First,
we will discuss the river containing Somchair, Sakda and Achara. From the map,
Sakda is on the downstream of both Somchair and Achara, while Somchair and Achara
are on the different tributaries. The toxic chemical contamination, Methylosmoline,
was detected increasing steeply at Somchair in recent years, but did not change
so much at Sakda and Achara, so surely we need to add the measuring stations both
on the upstream and downstream of the Somchair in order to clarify the range of
effect of pollutant and the real dumping location of chemicals. However, we need
to add measuring stations on the downstream of Achara. It is quite a long way from
Achara to Sakda. Although the values at Achara and Sakda are within the normal
range, we cannot deny that there may be another dumping point on the downstream
of Achara and the chemicals continue degrading along the way down to Sakda.
Therefore, measuring stations on the downstream of Achara are also very essential.
Same as the river discussed above, the river containing Boonsri, Kohsoom, Chai,
Busarakhan and Kannika need to add measuring stations on the upstream of Kannika,
Chai and Kohsoom, and on the downstream of Boonsri, Chai, and Busarakhan.
B) Concern for Pipits
If we observe closely the map of Boonsong Lekagul waterways, there are some white
lines on the background of the map. After comparison with the map given in the
Mini Challenge One, we find that these road lines are exactly driveways. Then,
we can combine the information of pollution and the migration of wild birds in
the past 20 years (analysis from Mini Challenge One) to unearth the influence of
pollution to the Pipits and other wildlife.
Here is the bird distribution map, according to the roads in Mini Challenge One.
Figure Q3-1
Compared to the methylosmoline distribution map before, birds in methylosmoline
polluted locations are much less than the other locations. There are nearly no
birds in the middle of the map, and there is no monitoring station nearby. In
the previous analysis, we suspect that dumping point changes to Somchair and Kohsoom
from Chai in 2016. So we need to build a new monitoring station
in the middle of the map.
In the year 2015, corresponding to the steep increase of the amount of toxic
chemical contamination, Methlosmoline, at Kohsoom, the number of Rose-crested
Blue Pipits nearby Kohsoom fell dramatically. What's more, the Pipit group started
to migrate southeast to keep aloof from Kohsoom, where is close to the waste
dumping place. Thus, we assume that the pollutants in the waterways really influence
survive of Pipit and maybe some other wildlife. Then, to go one step further,
according to our discovery of the unnatural increase of the amount of Methylosmoline
at Somchair, we have reasonable concern for the subsist of pinkfinch, purple Tooting
Tout, Queenscoat, Bent-beak Riffraff, Vermillion Trillion and Qax, since these birds'
habitats are distributed around Somchair and the upstream or downstream of Somchair,
the suspected new dumping place.