ISSS608 2017-18 T1 Assign RACHEL TONG DiseaseSpreadHypothesis
|
|
|
|
|
|
|
|
Part II - Transmission |
This section of the notes is dedicated to hypotheses about the disease transmission modes. Snapshots of geospatial exploration plots have been labelled chronologically in ascending numerical order.
Transmission Mode Hypotheses
Over the course of an average day, in a metropolis like Smartpolis, it is reasonable to expect some movement within regions, and varying levels of human traffic in and out of the different regions. In general, this is due to daily work or school commuter patterns. The association between symptom patterns and commuter travel patterns may thus shed some insight about possible modes of disease transmission.
Flu-like Symptoms
With the flu-like symptoms, we see some oscillation in related message locations throughout a working day. Upon the first instance of the symptoms on the morning of 18th May at the central region (east of the bridge region of highway 610), it may be observed that the concentration of messages with similar symptoms are localised at this region, and begin to spread towards the east into the Eastside region (snapshots 3 and 4).
In contrast, such ‘sick’ messages were diffused to the peripheral regions at 11.45pm that night (snapshot 5). During the morning hours of 19th May, we again see the concentration of ‘sick’ messages return back to central/Eastside area, away from peripheral towns (snapshot 6). This is likely reflecting the travel patterns of the daily commuter crowd. Similarly, the ‘sick’ message concentration remains localised to the central region during the daylight hours. Although some other ‘sick’ messages do appear to a smaller degree in peripheral regions, this may be the effect of some affected individuals remaining at home to rest. It could also be due to some background illness which was already present before the obvious advent of the more drastic onset on the 18th (snapshot 2). As such, it is unlikely that the flu-like illness is virulent through human-to-human contact, symptomatic messages do not appear as rapidly in peripheral regions (home to day commuters) as compared to the rate at which the symptoms developed upon onset at the central region on the morning of 18th May.
(Specific to the situation in snapshot 2, the symptoms already present include fever and cold, which are reasonably common symptoms that could be due to bad weather or other common illnesses. This implies that fevers and colds are less-discerning symptoms of this particular mystery illness.)
In contrast, by the morning of 20th May (snapshot 9), a more dispersed pattern is observed at the central area. Here, it appears that unlike the pattern observed at the same time over the previous 2 days, we do not see a distinctive commuting pattern, indicating that it is likely that more people are remaining at home due to the illness, and less people are commuting to work.
Over the course of the day (snapshots 9 and 10), the concentrations of symptomatic messages are diffused; and we observe that people are likely visiting the hospitals given the clusters of symptomatic messages that night (snapshot 11, above). By the late night (snapshot 12), we observe that peripheral clusters are completely dispersed, and the central region is dramatically de-concentrated, especially when compared to the patterns at the same timings over the previous 2 nights (snapshots 5 and 8).
Cross referencing snapshots above with the wind patterns, we can infer that the flu-like illness is likely to be airborne.
Considering that the truck spill was on the 17th, and days with observable patterns were on the 18th and 19th, we observe that the wind was blowing from the west from the 17th to 18th, with a mid-range wind speed. During the same period, the concentration of symptomatic messages clustered at the central region with a gradual spread towards Eastside is consistent with such wind patterns and speeds. From snapshots 5 to 7 (the night of the 18th to the morning and afternoon of the 19th), a very slight spread of the concentration along the south may also be observed. Again, looking at the wind directions, we observe a corresponding slight shift in wind direction source from the West to the West-Northwest. (Weaker grounds exist to corroborate the wind direction on the 20th with observable patterns, due to low wind speeds and the fact that a city-wide diffusion in pattern was observed.)
For gastrointestinal-related symptoms, affected regions were localized to Plainville, Smogtown, and Westside (and slight spillover to south of Villa); in the areas near the Vast River. Unlike the flu-like illness, the message locations generally do not oscillate between regions during different times of the day (snapshots 2 to 4, and 5 to 8 below), which indicates that these cases are likely contracted by people nearby the river banks, who are likely to be native residents of these regions. As such, the gastrointestinal-related symptoms are likely water-borne.
Further considering the later geospatial patterns (snapshots 5 to 8), it may be inferred from the distinctly localised clusters that people entering or leaving the affected regions did not develop similar symptoms, neither did they bring their symptoms out of the affected areas. This makes human-to-human transmission highly unlikely for this illness.
Finally, the likelihood of air-borne transmission of the illness with these symptoms is very slight, as the wind direction would have had to originate from the North-easterly direction. Considering that the wind direction from 05/14/2011 until the 05/20/2011 had either originated from the West/North-west/West-northwest, this pattern of symptoms is unlikely to be indicative of these symptoms.
|
|
|
|
|
|
|
