Entry Name:  "Individual-HanYing-MC1"

VAST Challenge 2019
Mini-Challenge 1

 

 

Team Members:

Ong Han Ying, Individual Contributor, mynameishanying@gmail.com PRIMARY

 

Student Team:  No

 

Tools Used:

Tableau

 

Approximately how many hours were spent working on this submission in total?

36 Hour

 

May we post your submission in the Visual Analytics Benchmark Repository after VAST Challenge 2019 is complete? Yes

 

Video

https://vimeo.com/347870023

https://youtu.be/O0GSR51fETM

 [video size ~300 MB]

 

Tableau Public Link:

https://public.tableau.com/profile/ong.han.ying#!/vizhome/MiniChallenge1/0_Reported_Case

 

Questions 1 – Emergency responders will base their initial response on the earthquake shake map. Use visual analytics to determine how their response should change based on damage reports from citizens on the ground. How would you prioritize neighborhoods for response? Which parts of the city are hardest hit? Limit your response to 1000 words and 10 images.

 

[Total: 945 words, 8 images.]

 

Assumptions:

       i.         earthquake effects are estimated with reference to the “Richter Magnitude Scale” (reference:  https://en.wikipedia.org/wiki/Richter_magnitude_scale)

     ii.         all shake and damage reports are relatively reliable for the purpose of analysis in this question.

 

1.     Volume of reports received by neighborhood is not correlated to the distance of the earthquake’s epicenter: Based on the earthquake shake map, emergency responders will likely focus on neighborhood near epicenters.

 

However, out of those which have received high volumes of reports; (Figure 1)

 

a.     Old Town (16%, rank 2) is the only neighborhood that received high volume of reported cases, and located in an area where “light” to “moderate” shakes are felt. Areas around Old Town such as Easton, Northwest, and Weston have received much reported cases, but with “no” to “weak” shakes only.

 

b.     Scenic Vista (17%, rank 1) & Broadview (11%, rank 3) do not receive any shake of the earthquake but have received much higher volume of reported cases instead.

 

Instead, neighborhood found with light and moderate shaking, such as Safe Town, Pepper Mill & Wilson Forest; did not receive much reports. (Figure 2)

 

Hence, emergency responders should not focus on neighborhood based on the shaked map only but should make reference to the responses too.

 

Figure 1 – Q1: Image 1

 

Figure 2– Q1: Image 2

2.     Prioritized neighborhood based on severity of the shake: The first priority is to save lives from the damaged areas, especially for areas with children and elderly. According to Richter Magnitude Scale, we should focus on area with >3.99, since < 3.99 rarely caused damages.

 

Indeed, reported areas with shake intensity of > 3.99 are found in neighborhood nearer to the epicenter of the earthquake (Figure 3) from 8th April 2020, 8AM onwards. In all,

 

a.     Priority 1: Old Town; with 45% reported cases.

 

b.     Priority 2: Easton, next to Old Town, with next highest reported cases (15%.) It is also an area with a higher population of children (schools and homes).

 

c.      Priority 3: while Safe Town (9%) has lower reported case than Easton & East Parton, it is located between Old Town & Pepper Mill, an area with older single homes. Hence, it might be under-reported and should be focused.

 

Figure 3– Q1: Image 3

3.     Next priority on medical services: Based on the responses, we identify areas with hospitals (Figure 4), and with the least damages in terms of shake, power, road & bridges & medical, (reported of < 3.99) to facilitate traveling to the hospital. This is ranked by Downtown, Southton, and Broadway. We then took a deep-dive (Figure 5) on the medical services and

 

a.     Emergency responders can send those seriously injured victims to TRAUMA Hospital located in Downtown, the safest and nearest hospital to Old Town with Trauma Centre,

 

b.     followed by EAGLEPEAK Hospital in Southton with critical care.

 

c.      The less injured victims can be sent further to the COMMUNITY HOSPITAL in Broadway further away, however; there were significant reports of 5.0 damage to the medical services, and thus; may not be fully equipped.

 

d.     As such, emergency responders can consider GOLDCARE HOSPITAL in Terrapin Spring with Medical Services with less damages but note that the power might not be available.

 

e.     Do not sent any patients to OLD TOWN HOSPITAL as high volume of strong & major damages to the medical services are reported. (Figure 5)

 

Figure 4– Q1: Image 4

Figure 5 – Q1: Image 5

4.     Next priority on repairing work on power, sewage and water: This will help to facilitate daily living to continue as usual. With reference to Figure 6, we should focus on neighborhoods with Power and Sewer & Water with reported damage of >= 6.0. These neighborhoods are identified in the order of;

 

a.     Priority 1: Scenic Vista (33%) & Broadway (13%), with Broadway located next to Scenic Vista and with more elderly in the area.  

 

b.     Priority 2: Old Town (29%) & Easton (10%), with Easton located next to Old Town and with schools in the area.   

 

Figure 6 – Q1: Image 6

 

5.     Last priority on repairing work Roads & Bridges and Buildings: This tends to take longer time, and priority should still go to neighborhoods with >=6.0 damages in these aspects, as shown in Figure 7;

 

a.     Priority 1: Scenic Vista (30%), Broadway (22%), & Chapparal (5%). with Broadway & Chapparal located next to Scenic Vista only.

 

b.     Priority 2: Old Town (28%) & Easton (4%), with Easton next to Old Town only.

 

Figure 7 – Q1: Image 7

6.     Overall, Scenic Vista & Old Town are the hardest hit in the city, with a median reported overall score of 7. However, the most hit neighborhood is Old Town; based on our observation from Figure 8, as below;

 

a.     Only neighborhood on the North to Northeast has feedbacked on significance shake intensity, such as Old Town, Safe Town, Pepper Mill, and Wilson Forest.

 

b.     All areas except Downtown & Southwest are affected by damaged / missing medical services.

 

c.      In the short term, except for Old Town; the South to East of the cities are hit most by damages to Power, Sewer & Water – despite not having high feedback on significant shake intensity.  Examples of neighborhood include Broadview, Chapparal, Scenic Vista & Wilson Forest.  

 

d.     In the long term, except for Old Town; the South to East of the cities are also most hit most by damages to Buildings & Roads and Bridges too.  Examples of neighborhood include Broadview, Chapparal, Scenic Vista & Wilson Forest.  

 

Thus, Old Town is hit by the higher Shake Intensity, and suffer significant damages in both short & long term. This is followed by those neighborhoods in the South to East, though not suffer from higher shake intensity; but seem to have received damages that required repair work in both short & long term.

 

Figure 8 – Q1: Image 8

 

Questions 2 – Use visual analytics to show uncertainty in the data. Compare the reliability of neighborhood reports. Which neighborhoods are providing reliable reports? Provide a rationale for your response. Limit your response to 1000 words and 10 images.

 

[996 words, 7 images]

 

Assumptions:

 

       i.         Report reliability matters to neighborhoods with

 

a.     high volume of reports and/or

 

b.     suspected significance damages to the neighborhood.

 

These will determine if the emergency responders should react to them. From question1, we will examine the reliability of reports from;

 

a.     Scenic Vista, Old Town, Broadview – the top 3 “most hit” neighborhoods, with the highest number of reports received.

 

b.     Safe Town & Pepper Mill - the populated housing estates & tourist locations, nearest to the epicenter of the earthquake

 

c.      Any ‘potential’ highly damaged locations that have been missed out;

 

d.     Any ‘potential’ highly shaky locations that have been missed out;

 

1.     Firstly, we examine Old Town - the only neighborhood in the North with both high volume of reports & receive the most hit. As shown in Figure 9, Old Town has shown much reliability in the report because;  

 

a.     Received high shake intensity of >=5.0, and resulting in

 

b.     significant damages (>= 5.0) in all areas - Power, Roads & Bridges, Buildings, Medical, Sewer & Water.

 

c.      Even through according to the shaked map, the area should have (Figure 1 – Q1) received <= 4.0 of shake intensities, on-going construction work in Old Town such as old water supply lines, modernizing the electrical distribution network will have worsen the damages, and results in over-reporting.  

 

However, note that there was uncertainty shown in the data, where there are likely 2 cut-off in power supply in Old Town (after the earthquake);

 

a.     between 08April-10AM to 09April-12AM – resulting in a delay & high volume of reports received on 09April-1AM

 

b.     between 09April-4PM to 10April-11AM – resulting in a delay & high volume of reports received on 10April-12PM

 

and these reports are likely delayed and are likely referring to reports to damage during the earthquake, rather than much later.

 

Figure 9 - Q2: Image 1

 

2.     Next, we will examine Scenic Vista and Broadway – located in the South with high volume of reports & receive the most hit. As shown in Figure 10 (Scenic Vista) & Figure 11 (Broadway), the reports were likely less reliable as compared to Old Town because;

 

a.     Both neighborhoods did not receive much significant in shake intensity of more than 4.0,

 

b.     however; both have reported high volume of damages (> 5.0) in all areas except for Medical in Scenic Vista.

 

However, we also noted that reports from Scenic Vista (Figure 10) are more reliable than Broadway because;

 

a.     it is a neighborhood located on the hillside, and therefore; it might be prone to much higher damages, despite lower shake intensity.

 

b.     There is no hospital in this region, and therefore; accurate reporting of >10K missing reports in medical.

 

c.      There are 2 power cut-off time, which similar period as Old Town, as below;

                                               i.     between 08April-1PM to 09April-9AM – resulting in a delay & high volume of reports received on 09April-9AM.

                                              ii.     between 09April-6PM to 10April-2AM – resulting in a delay & high volume of reports received on 10April-2AM.

 

These reports are likely delayed and are likely referred to reports to damage during the earthquake, rather than much later.

 

            On the other hand, reports from Broadway (Figure 11) are less reliable than Scenic Vista because;

 

a.     It is not located at hill area, and therefore; low shake intensity (< 3.0) is unlikely able to result in high damages in facilities.

 

b.     There was only 1 power cut-off time only instead, as below;

                                            i.       08April-1PM to 08April-11PM - resulting in a delay & high volume of reports received on 08April-11PM.

 

Figure 10 - Q2: Image 2

Figure 11 - Q2: Image 3

3.      Also, we will examine both Safe Town and Pepper Mill –located near the epicenter of the earthquake. As shown in Figure 12 (Safe Town) & Figure 13 (Pepper Mill),

 

a.     Both neighborhoods did receive significant in shake intensity of more than 4.0 to 7.0.

 

b.     No/low report on damages to medical (accurate in absence of hospitals in both areas).

 

c.      Both have reported higher volume of damages (> 5.0) in all areas except for Medical.

 

With further examination to Safe Town (Figure 12), we noted that;

 

a.     With a Nuclear Powerplant, reported damage of between 4.0 to 6.0 may have been under-estimated.

 

b.     Damage to Sewer and Water is unusually high of above 5.0, without active construction work around.

 

c.      Damages to Roads & Bridges and Buildings (between 3.0 to. 6.0) are aligned to the shake intensities.

 

While further examination to Pepper Mill (Figure 13), we noted that it is much more reliable that Safe Town because

 

a.     The damages are lower than Safe Town but align to its shake intensity.

 

b.     The results are aligned to the low-density housing – higher damages to the neighborhood (near to epicenter) but lower in volume reports (lower population density).

 

Figure 12 - Q2: Image 4

Figure 13 - Q2: Image 5

4.     No additional town shown significant damages (>= 5.0) as per Figure 14, except;

 

a.     Easton, next to Old Town & Chapparal between Broadview & Scenic Vista. Thus; are likely reliable.

 

b.     Palace Hills – further away from both epicenter of earthquake and high-volume towns. Thus, they are likely unreliable reports.

 

Figure 14 - Q2: Image 6

 

5.     No additional town shown significant shake intensity (>= 5.0) as show per Figure 15, except Easton & East Parton - next to Old Town & Safe Town. Thus; are likely reliable.

 

Figure 15 - Q2; Image 7

Conclusion:

1.     the early reports before the power cut-offs from Old Town, Scenic Vista are relatively reliable and should be responded.

2.     Damages to Safe Town may have been under-estimated and better for emergency responders to react to it.

3.     Easton & Chapparal may have significant damages and should be responded at the same time as Pepper Mills – which likely reliable and should be responded.

4.     Reports from Broadview might be over-estimated and should be lower priority as compared to the other areas.

5.     Significant Damages Reports from Palace Hills are likely over-estimated and can be of lower priority.

 

 

Questions 3 – How do conditions change over time? How does uncertainty in change over time? Describe the key changes you see. Limit your response to 500 words and 8 images.

 

[486 words, 8 images]

 

1.     Overview of Changes in Conditions & Uncertainty Over time:

 

a.     From Figure 16, significant earthquake (>=4.0) is suspected to happen on 8thApril, where the volume of reports peaked in most neighborhoods.

 

b.     From Figure 17, this significant earthquake was likely to happen between 8AM to 12PM on the 8thApril.

 

c.      From Figure 17, Unusually high volume of reports before / after earthquake, as shown;

 

                                               i.      06 April at 4PM: >1000 reported cases of earthquake and damage. This is the only time with high reported cases, before the likely earthquake.

                                              ii.     1st significant spike in volume of reports after the earthquake on 08April-11PM.

                                             iii.     Multiple spikes observed on 09April-1AM & 9AM. A continuous spike from 3PM to 5PM, and dies down afterwards

                                             iv.     Last significant spikes were identified on 10 April at 2AM and 12PM

 

 

Figure 16 - Q3: Image 1

 

Figure 17 - Q3: Image 2

 

 

2.     Key Highlights of Changes in Conditions Over time:

 

a.     From Figure 18, 1st peak of reports was received on 8April-8AM, where;

                                               i.     High volume of 2200 (30%) were received from Old Town, significantly.

                                              ii.     Reports with shake intensity of >4.0 were received too, majority from Old Town too.

                                             iii.     Damage reports of > 4.0 also peaked in Old Town in all areas, followed by Eason (next to Old Town), in all areas except medical.

 

b.     From Figure 19, on 8April-9AM, number of reports peaked from the city, where;

                                               i.     Significant numbers of reports were received from neighboring of Old Town & Eason – such as Weston, Northwest & East Parton.

                                              ii.     High numbers of reports were first noticed from the southern of the city – from Broadway and Scenic Vista.

                                             iii.     Damage reports of > 4.0 were first received from the south and continued from the north.

 

c.      From Figure 20, on 8April-1PM, the first sharp declined of reports were received from the city, where;  

                                               i.     No report received further from neighborhoods affected by earthquakes.

                                              ii.     Power supply may have been cut off during this period.

 

Figure 18 - Q3: Image3

Figure 19 - Q3: Image 4

 

Figure 20 - Q3: Image5

3.     Key Highlights of Changes in Uncertainty Over time:

 

a.     Before the earthquake, From Figure 21;

                                               i.     High number of reported on 6April-4PM, but no significant shake intensity or damages were reported.

                                              ii.     Shakiness on the bridges & buildings might be detected (signs on upcoming earthquake) with most reports in these areas.

 

b.     After the earthquake,

                                               i.     From Figure 22; power supply might have return (after the cut-off from earthquake) gradually in the order of Broadway, Old Town, followed by Scenic Vista. Power supply is returned to most neighborhood on 9thApril-3PM.

                                              ii.     From Figure 23; seems to have a 2nd power supply cut-off to Old Town & Scenic Vista, where power supply returned again on 10th April instead.

                                             iii.     Those reports received during these periods (after the earthquake) are likely delayed due to power supply cut-off.

 

Figure 21 - Q3: Image 6

 

 

Figure 22 - Q3: Image 7

Figure 23 - Q3: Image 8

 

Questions 4 –– The data for this challenge can be analyzed either as a static collection or as a dynamic stream of data, as it would occur in a real emergency. Describe how you analyzed the data - as a static collection or a stream. How do you think this choice affected your analysis? Limit your response to 200 words and 3 images.

 

[197 words, 2 images].

 

Both methods were used, with

 

       i.         An interactive dashboard (Figure 24) to support static collection

     ii.         A motion dashboard (Figure 25) to display the changes to the town & damages reports hourly, between 6April to 10April.

 

1.     An interactive dashboard helps since all the charts within it act as filters.

a.     This helped to provide me with an overview to the situation of earthquake over time, volume of report by neighborhood, shake intensity and damages reports.

b.     Upon detecting interesting insights, I can further deep dive by clicking on the part of the charts, where and the other charts will be updated. This provide me with related insights speedy.

 

2.     A motion dashboard helps where;

a.     It simulated what have happened / been reported in sequence of time, hourly.

b.     Upon playing this dashboard, I can observe the overview, shake intensity, and damages report – overtime; all over the same screen.  

c.      Upon detecting a significant change in color (darken) in any of the charts on this dashboard, I can click “pause” to “stop” the motion of the dashboard, analyze the changes that have happened.

 

Thus, using both methods together provide an iterative way to conduct visual analysis.

Figure 24 - Q4: Image 1

Figure 25 - Q4: Image 2