Skip to content

pexposure.md

Latest commit

 

History

History
141 lines (106 loc) · 5.24 KB

pexposure.md

File metadata and controls

141 lines (106 loc) · 5.24 KB

How can we compare a Modelled outcome with a Personal Measurement outcome?

Hyesop Shin 2019-05-18

Abstract




Rationale

  • Ignoring the spatiotemporal variability of environmental risk factors and human mobility may lead to misleading results in exposure assessment (See Yoo Min Park and Mei-Po Kwan) \

  • Where people live is often not the only important factor in determining their exposure to environmental factors

  • Rather, where people visit and how much time they spend at a particualr location are more relevant to assessing the effects of environmental factors on people’s health behaviours or outcomes




Personal measurement of PM10 Exposures

exposure_summer %>% 
    mutate(loc1_1nm = case_when(loc1_1 == 10 ~ "Restaurant",
            loc1_1 == 11 ~ "Cafe",
            loc1_1 == 12 ~ "BBQ grill",
            loc1_1 == 13 ~ "Bar",
            loc1_1 == 14 ~ "Office",
            loc1_1 == 15 ~ "Traditional market",
            loc1_1 == 16 ~ "Superstore",
            loc1_1 == 17 ~ "Department store",
            loc1_1 == 18 ~ "Shopping complex",
            loc1_1 == 19 ~ "Other shops",
            loc1_1 == 20 ~ "Workplace",
            loc1_1 == 21 ~ "Bank",
            loc1_1 == 22 ~ "School",
            loc1_1 == 23 ~ "Academy",
            loc1_1 == 24 ~ "Bookshop",
            loc1_1 == 25 ~ "Senior centre",
            loc1_1 == 26 ~ "Stroll",
            loc1_1 == 27 ~ "Walking",
            loc1_1 == 28 ~ "Bus",
            loc1_1 == 29 ~ "Subway",
            loc1_1 == 30 ~ "Taxi",
            loc1_1 == 31 ~ "Vehicle",
            loc1_1 == 32 ~ "Home",
            loc1_1 == 999 ~ "Missing data"
            ))




Overview of Trajectories(Cont.)

  • 16142 records of 5 individual backpack sensors measured by minutes




Clusters of footprints

  • Points are the footprints for each researcher
  • The ellipsoid represents the 95% confidence of the distribution of points on the map
  • The footprint of researchers are centered in Gwanak district, particularly situated in the University Campus.

Exposure levels by researchers






Highly polluted areas?



loc1_1 loc1_1nm mean_pm10 sd_pm10 min_pm10 max_pm10 count
10 Restaurant 53.59 41.30 9.7 270.4 713
11 Cafe 35.31 15.14 14.7 98.3 184
12 BBQ grill 259.02 276.20 24.3 1170.3 295
13 Bar 124.69 114.32 20.3 771.5 1178
14 Office 22.83 15.93 4.5 147.2 5397
15 Traditional market 76.64 30.84 25.4 152.2 60
16 Superstore 70.39 35.34 22.7 136.9 10
18 Shopping complex 112.53 75.63 27.1 320.4 14
19 Other shops 37.33 13.69 16.2 172.0 1771
20 Workplace 36.45 19.95 5.2 113.6 713
21 Bank 26.69 15.68 14.4 91.9 41
22 School 37.04 21.25 2.4 239.9 880
23 Academy 11.09 5.20 4.8 46.7 254
24 Bookshop 86.90 NA 86.9 86.9 1
25 Senior centre 22.18 37.04 4.9 799.5 954
26 Stroll 43.72 37.09 12.7 265.0 259
27 Walking 62.85 61.09 1.0 954.1 1365
28 Bus 37.64 90.66 3.5 2565.9 860
29 Subway 32.18 21.46 8.3 112.7 118
30 Taxi 45.21 67.63 1.1 581.0 373
31 Vehicle 36.92 40.21 8.6 802.1 645
999 Missing data 63.15 47.10 7.2 402.0 394




PM10 Exposure by Transport Modes




(Near) Future works

  • Use walking and strolling info to compare modelled outcome
  • Consider time scale from minutes to 12 hour aggregation