This document discusses how GIS technologies can be used for environmental hazard and exposure surveillance. It describes categories of GIS applications for hazard surveillance using sensors to model hazard movement, and exposure surveillance by intersecting population and hazard data to consider factors like dose, exposure time, and activity spaces. It also discusses dose and outcomes surveillance, limitations of GIS, and trends toward personalized exposure assessment using sensors on devices, buildings and individuals.

1. Innovations in GIS for
Environmental Hazard and Exposure Surveillance
PH503
Damien Leri
May 1, 2014

2. Categories of GIS to Enviro. Health:
Hazard surveillance
● Sensors in environment
● Modelling of hazard movement such as air
flow
○ Dispersion modelling can be sophisticated, for
example water pollution disperses through plumbing
networks rather than in a simple circular pattern
around the point source.

3. Exposure surveillance
The intersection of population data with hazard data.
Considers several dimensions of exposure such as:
● Dose
● Effective exposure time
● Latent period
● Threshold vs non-threshold toxicants
Considers movement of people in their activity spaces
through the day.

4. Dose surveillance
An extension of exposure surveillance, this is
also a field in its own right.
It is the human biomonitoring of actual dose
received.

5. Outcomes surveillance
Mapping clinical symptoms and outcomes
using screening tools and diagnostic reports.
Examples: Blood lead level screening; illness
hotspotting analysis.

6. Limitations of GIS for env. health
● Often a lack of patient level data (for privacy)
● Lack of nationwide consistency for disease tracking and
much other data
● Some effects show up at only certain levels of aggregation
● Geographic boundaries (such as census ones) are arbitrary
● Geographic data effects such as spatial auto-correlation
● Temporal lag in onset of symptoms for chronic disease
● Poor quality/consistency of outcomes data such as
diagnoses

7. Geography of risk
Geography of susceptibility
Geography of exposure
Geography of risk
A smart design of exposure surveillance will incorporate both susceptibility and
exposure.

8. Cromley & Joy (1995)
Researchers looked at electromagnetic field
(EMF) radiation for exposure modelling for a
specific target population.
● Geography of susceptibility: schools and
homes. Filtered for vacant areas.
● Geography of exposure: predicted EMF
sources including dispersion pattern

9. ● 1980s - 2000s: Satellite imagery. Reflection by particles correlates
with air quality but has limitations.
● Trend toward personalized exposure assessment -- identifying risks
at the level of buildings, families, and individuals.
● 2006: Pigeon Blog project by Beatriz DaCosta -- homing pigeons
carrying GPS and air quality sensors.
● 2010: Personal monitoring through auxiliary devices.
● Future? Using existing sensors such as phones and wearables, at
the personal level, combined with networked data.
Innovations in surveillance

10. The holy grail of exposure surveillance
Personalized exposure assessment:
Identifying risks at the level of buildings, families, and individuals.
Trends that are advancing us toward this goal:
● Increases in sensor volume and quality
● Moving sensors closer to humans
● More sophisticated data analytics

11. Personal exposure: noise
MobGeoSen: facilitating personal geosensor
data collection and visualization using mobile
phones (Kanjo et al., 2008)

12. Personal exposure: air
Development of a method for personal, spatiotemporal
exposure assessment (Adams et al., 2009)

13. Consumer products?
“Visibility” app by USC in 2010 MicroPEM personal sensor by RTI in 2012
These are prototypes of 3 approaches to
personal exposure monitoring of air quality.
Citi-Sense

14. The EPA and Next Generation Air Monitoring
CAir Clip
EPA concept proposals

15. Better analytics: Hystad, et al. (2009)
Hystad, P. U., Setton, E. M., Allen, R. W., Keller, P. C., & Brauer, M.
(2009). Modeling residential fine particulate matter infiltration for
exposure assessment. Journal of Exposure Science and Environmental
Epidemiology, 19(6), 570-579.
The penetration of air pollution indoors was
predicted using publicly available tax property
assessment data combined with weather and
topology data.

16. Better analytics: Miranda et al. (2002)
Miranda, M. L., Dolinoy, D. C., & Overstreet, M. A. (2002). Mapping for
prevention: GIS models for directing childhood lead poisoning prevention
programs. Environmental Health Perspectives, 110(9), 947.
Blood lead risk stratification based on factors including:
● resident SES
● land use (zoning for residential)
● building age and renovations (based on construction
permits)
● occupancy status (owned-occupied or rented)

17. Grassroots data networking
Air Quality Egg

18. damien@bigyellowstar.com (Damien Leri)