IS428 2019-1920 T1 Assign LeeSunho Interactive Visualization

Interactive Visualization

Problem & Motivation

Data Exploration & Transformation Process

Interactive Visualization

Insight

Taks one

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.

As the emergency responders of St. Himark have limited resources, it is imperative that they are able to prioritize certain areas in the city during the natural disaster. They can do so by basing their initial response on the earthquake shake map. Using visual analytics, I have created two live dashboards which will report the various data regarding the earthquakes. Through this dashboard, the emergency responders will be able to answer several important questions such as: “How to prioritize which neighbourhoods for response?” and “Which parts of the city are hardest hit?”

Live Dashboard The first dashboard provides a more micro view of the data by showing the user the earthquake intensities grouped by the different regions on specific dates and times. On the top right corner of the dashboard, emergency responders are able to adjust the date to see the intensity of the earthquake on those specific days. This dashboard currently is split into five different sections: line chart, cartogram, intensity counter, heat map, and box plot. I will now be explaining each of the different sections in detail.

Overll Summary The second dashboard provides a more macro view of the data which shows the summary of the overall reported intensity faced by St. Himark for all the days instead of a specific one. This dashboard is also divided into 5 different sections – line graph, cartogram, table, bar chart, and box plot – in which I will also be explaining in detail.

Explanation	Visual Representation
Cartogram to indicate priority </font the cartogram is basically a map that shows the intensity of the earthquakes at a specific time of day. Through this, users will be able to quickly understand which regions need to be prioritized as the darker hues of red signify higher intensity earthquakes. From the chart, it is evident that Old Town and Safe Town need the most help, followed by Pepper Mill and Wilson Forest.
Intensity Counter the intensity counter is a list which indicates the specific numerical intensity that these regions experienced. In addition, this counter only reports those regions that have experienced an intensity of 6 or higher, or the more dangerous earthquakes. This section of the dashboard perfectly complements the cartogram since the cartogram uses colour to indicate intensity while the list uses an actual numerical amount.
Line chart the line chart is the shake intensities of the different regions of St. Himark. Having 19 different regions within its borders, St. Himark is bound to experience varying levels of earthquake intensity during the natural disaster, thus it is important to separate the effects of the intensity by region. The x-axis represents time in hours while the y-axis represents the intensity of the earthquake For all the of the line graphs, there is a gray horizontal line which signifies the reference line. This is to help indicate the intensity level of 6, which is the threshold between a catastrophic earthquake and a minor shake; if the line is below the reference line, it will be blue, indicating that the earthquake is still moderately safe but a red line indicates danger. Hence, based on the chart, users can easily identify which region needs the most help at certain times of the data on a specific date. For example, from the chart, it is evident that Broadview, Cheddarford, East Parton, Pepper Mill, Safe Town, Old Town, Terrapin Springs, West Parton, ad Weston have reportedly high intensities between 9 am until 4 pm.
Heat Map The heat map indicates the cumulative shake intensity of St. Himark by the hour. In other words, it will help the users understand which hours experience higher levels of intensity in the whole region. By looking at the heatmap, it is evident that St. Himark is hit with earthquakes after 8am.
Box and whiskers plot the box and whiskers plot shows the reported intensity faced by the various public facilities and amenities. This is to help address the question of “Which public good needs to be repaired the most?” By observing this plot, users can quickly see the mean, median, and range of the reported intensities.
Total Summary Intensity level from 6th April to 11th April The line graph indicates the shake intensity from April 6 to 11. Similar to the first dashboard, when the line graph turns red, it means that the earthquake is more intense and dangerous. By observing the chart, it is evident that the 8th and the 9th are the most severely impacted by the earthquakes. During the 8th, the whole region aggregately faced consistently high intensity levels after 8am while during the 9th, there was a slight shock in the morning which then reverted to a safe level, only to increase again after 2pm and remain at that level until the night.
Total Summary by region The cartogram indicates the region that has been most affected by the earthquake. It shows which region has been the most affected overall by the earthquake in all 6 days by using colors to indicate the regions with the highest number of reported intensities with darker hues of red.
Total Summary data table The table reveals the average reported intensity faced by the facilities in the different regions for all the days. If the average aggregate reported intensity is greater than 6, the number is colored in red; for example, the sewer and water facilities in Broadview exceed 6, thus they are prominently colored.
Total Summary for Class The box plot shows the overall minimum, maximum, and median intensities faced by the different facilities in St. Himark. Through this method of visualization, users are able to quickly gauge the range and medians of the intensities in each section.
Total Summary different level of intensity The bar charts convey the amount of people who reported those specific intensity levels. Weak includes an intensity level between 1 and 3, moderate is 4, strong is 5, and extreme is any intensity level greater than 6. Thus, for example, there were 35,000 reports of a strong earthquake felt in St. Himark.

Taks Two 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. As aforementioned, because St. Himark has seismic sensors throughout the cities, they are able to obtain the actual magnitude of the disaster. However, it would be unwise to simply deploy their people there just because there’s a high intensity earthquake because that earthquake might not cause that much damage. It is thus important to compare the actual seismic reading with the reported damage in each of the separate amenities. Explanation Visual Representation Heat map Next, this visual representation shows the number of reported intensity values in each region at specific times of the data. It is evident that some regions are less consistent in providing up-to-date information regarding the earthquakes facing. Although some regions in the city are facing reparations in the electricity department, it was stated in the summary of the city that there should only be power outages which last about 30mins to 60mins in both Old Town and Southwest. However, the data in Old Town clearly indicates a lack of data on the 8th after 12pm. By using consistency as a reliability measure, we can state that Broadview, Chapparal, Old Town, and Scenic Vista are less reliable neighborhoods in terms of their reporting. During the major intensity After the major intensity Bar Chart This bar chart clearly shows the difference in the level between the shake intensity or seismic readings, which is represented in red, with the reported amount of damage, which is represented in the various different colors. This data is presented from a macro perspective, meaning that these are the aggregate average intensity values of a specific region for two whole days. The separation between the specific dates was done intentionally because the 8th and the 9th of April have significantly more earthquakes than the 10th and 11th; thus it would be beneficial to split the data with many earthquakes occurring on the left panel while a fewer amount occurring on the right. Through this visual representation, we are able to see the uncertainty in the data in various regions. On the right panel, some of the regions face an extremely low magnitude, however the reported damage in the various amenities is significantly greater. For example, Broadview and Chapparal are regions that experienced an average magnitude of 2, however the reported damage in those various regions are all greater than 5. Referring back to the earthquake intensity scale chart, a 2 would represent a slight swaying while a 5 would represent a more intense tremble with windows and doors swinging. While on the left panel, due to the fact that less earthquakes occur on these two days, the seismic readings or the red bars are all significantly small. However, the reported damage in all of the amenities is significantly greater in which there are two possible explanations for. First, because the citizens have recently experienced numerous earthquakes in the past few days, they might be more sensitive when experiencing new ones. Perhaps due to the heightened sense of fear and bad experiences in the past, they are overemphasizing the effect of the damage. Second, because there was a relatively short time gap between these two panels, the emergency responders might have not had the opportunity to actually repair all of the damages reported by the citizens. So because of the earthquakes in the past, extensive damage could have been done. However, because they have not yet been fixed, the amenities still remain broken. Thus, the citizens probably correlated all the damage done in the past by the more severe earthquakes to any slight shake of the earth which aren’t actually causing any more harm. During the major intensity After the major intensity Data Table an interesting thing to note is that St. Himark has an exceptional network of hospitals located in 6 different regions. This means that the remaining 12 other regions do not have a hospital located in the area. Above is a table showing those remaining 12 other regions and the reported damage in the hospitals despite there not being one in the vicinity with the sum of those errors listed on the right panel. It is possible that when inputting the data, citizens accidently selected the wrong facilities and forgot to double-check before submitting. However, it is evident that Scenic Vista has a shockingly high level of mistakes. This shows that this neighbourhood is filled with people who are more careless and thus their reports are less reliable. Taks Three 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. As the days pass, more and more citizens input their observations and reports into the mobile application. Hence, in theory, the data should become more accurate and reliable. However, over time, this is not the case. Even though there is large number of reported observations, consistency and reliability cannot be found. Furthermore,After each quake, we had more reports and hence more evidence to certainly support the posterior distribution. We then explored the distribution average intensity reported over time, We observed the same pattern that the reports became certain right after the quakes and gradually spread out. Explanation Visual Representation Region reported data analysis The top chart shows the data pattern when inconsistent regions are filtered out. However, a similar pattern still exist where the reported value is still very high compare to the shake intensity. The bottom chart is a scatterchart with a box chart in it which indicates the non-medical regions reported intensity in each of the days at specific timings. As can be seen through the chart, all of the days except the 8th has some significant outliers while the 9th has a few outliers but not as many as the other days. This means that when the earthquakes were occurring, the data reported becomes more consistent and the other periods have a much larger variability in their data. In other words, the uncertainty in the earthquake period decreases while the uncertainty in the other periods increases. Medical Certainty As aforementioned in the second question, some of the regions do not have hospitals yet many of its citizens have reported medical intensity level. Furthermore, some regions are not consistency in reporting their data such as Wilson Forest, Scenic Vista, Old Town and Chapparal, which significantly causes the reliability to drop. In other words, these regions have periods of greater uncertainty when the citizens in those regions do not actively update the mobile application. Hence, those regions who did not report timely information or reported faulty information can be filtered out. For the medical regions, the data can be more trusted and users can further filter based on the cases and issues to make it more accurate. Medical and shake intensity Region that has hospital