PowerPoint Presentation Hincks COV4 Ash Exposure Study

1
Long term exposure to respirable volcanic ash on
Montserrat a time series simulation T. Hincks,
R.S.J. Sparks University of Bristol W.P.
Aspinall Aspinall and Associates and
University of Bristol P.J. Baxter Dept of Public
Health and Primary Care, Univ. Cambridge A. Searl
Institute of Occupational Medicine, Edinburgh G.
Woo Risk Management Solutions
2

• Aim
• Estimate risk of silicosis from cumulative
  exposure to cristobalite
• Volcanic activity
• Ash composition
• Deposition and erosion
• Weather conditions
• Human activity / occupation

3
Risk assessment for Montserrat
1996 - DfID and DoH research on health risks
associated with volcanic ash
3 - 24 weight crystalline silica in lt10mm
(inhalable) fraction of ground deposits Baxter
et al. (1999) preferentially fractionated in PM4
(lt4mm) Horwell et al. (2003)
? silicosis, lung cancer, pulmonary tuberculosis,
autoimmune diseases

• Active since 1995 - high cumulative exposures?

• Health risks?
• Exposure to date
• Continuing volcanic activity

4
Uncertainty and expert judgment

• Difficult to assess risk in populations
• Information highly uncertain and poorly
  constrained
• ??probabilistic risk assessment

eruptive history ash isopach data air quality
monitoring weather data
Observations
ash dispersal weather simulation
Numerical models
Expert elicitation
formal and unbiased process for obtaining
information from experts limited data full
extent of system behavior?
Aleatory uncertainty
Observables represented by probability density
functions
5
Time series simulation

• Code generates daily exposures for PM10 (lt10mm)
  and cristobalite crystalline silica
• 6 occupation groups
• 4 sites
• Multiple runs (10,000 runs/simulation)
• Sample from PDF for each parameter
• Correlated sampling

6
Model structure
7
Model structure
8
Dome growth

• Periodic dome growth function controls
  replacement of material
• ? frequency of collapses
• ? increased probability of vulcanian explosions
  at high growth rates

Growth rate m3s-1
9
Rainfall

• Rainfall time series simulated as two part
  process
• Incidence of rain (true/false)
• Quantity of rain (24h)

Mean
Rain depth Lognormal distribution with time
dependent parameters
Standard deviation
10
Volcanic activity

• 6 event categories are considered ? significant
  ash deposits
• 3 - 10 x 106 m3 dome collapse
• 10 - 30 x 106 m3 dome collapse
• 30 - 50 x 106 m3 dome collapse
• 50 - 75 x 106 m3 dome collapse
• gt75 x 106 m3 dome collapse
• Series of 0.4 x 106 m3 vulcanian explosions
• Assume event magnitudes and frequencies similar
  to past 10 years activity ? daily P(event)
• Probability of Vulcanian explosions increases
  after major dome collapses and during periods of
  high extrusion rate

11
Ash deposition
Ash deposits generated with HAZMAP 2-D advection
diffusion model for ash transport (Bonadonna et
al. 2002)

• 3 years of daily wind data
• Dome collapse pyroclastic flows down 5 valleys
• Single source Vulcanian explosions
• ?correlated lognormal deposits distributions for
  4 locations

12
Ash removal (wind and rain)

• Approximate with 4 deposit levels
• Use beta distribution to represent variation in
  deposit lifetime
• Expert elicitation for mean, upper and lower
  bounds

13
Individual exposure
High exposure Gardeners Public works department
Low exposure indoor occupations elderly

• PDF VARIATION IN EXPOSURE
• Beta distribution function of
• deposit depth
• cristobalite content of ash
• occupation
• Modified to account for rainfall

dust trak data
14
Individual exposure
Sum daily exposure values over 5, 10 and 20 years
? estimate cumulative exposure ? risk of
silicosis
15
Simulated time series
16
Simulated time series
17
Simulated time series
UK HSE recommended maximum occupational exposure
to crystalline silica 0.3 mg m-3 Suggested
limit 0.1 mg m-3 8h time weighted average
US NIOSH recommended limit 0.05 mg m-3 time
weighted average for up to 10 hour work day
during 40 h working week
UK HSE (2003) suggested limit
US NIOSH recommended limit
18
Results 20 year cumulative cristobalite exposure
f
w
trials exceeding exposure
s
c
cumulative cristobalite exposure mg.m-3.year
cumulative cristobalite exposure mg.m-3.year
19
Estimating risk exposure-response functions for
silicosis

• Upper limit of risk
• Buchanan et al. 2003
• Study of silicosis in Scottish coalminers
• High intensity exposure gt 0.1 mg m-3
• Heavy ash fall areas only

20
Estimating risk exposure-response functions for
silicosis
Upper limit of risk Buchanan et al. 2003 Study
of silicosis in Scottish coalminers Most
analogous Hughes et al. 1998 occupational
exposure for diatomaceous earth workers
gt 0.5 mg m-3
0.5 mg m-3
exposure intensity affects risk
21
20 year risk of silicosis

• 20 years continuous exposure
• Hughes et al. (1998) risk function

TYPICAL ADULT
CMO risk scale
Estimated exposures lie within bottom 20 of
Hughes cohort (lt100 cases) RISK?
22
20 year risk of silicosis

• 20 years continuous exposure
• Hughes et al. (1998) risk function

TYPICAL ADULT
CMO risk scale
OUTDOOR WORKER
23
Validation future work

• Medical studies
• 2000 x-ray survey of 421 high risk workers
  showed no evidence of chest abnormalities (lt 5
  years exposure)
• X-ray survey after 10 years exposure
• Risk to children highly uncertain
• limited data - better estimates of cumulative
  exposure?
• applicability of exposure response functions?
• Field data
• Continuous PM10 measurement
• weather data
• Personal exposure sampling
• Ash erosion rates - very poorly constrained
• Duration of hazard
• Implications for lahar and flood hazard assessment

24
Validation future work

• Exposure control measures
• dust masks for outdoor workers in ash affected
  areas
• minimize exposure during cleanup operations
• minimize childrens exposure (clear sports play
  areas after ash fall)

25
Further applications

• Ash-leachates
• water contamination
• risk to livestock
• crop damage

Guadeloupe concerns about contamination of
aquifer

• Popocatépetl
• PM10, ash leachates

26
Acknowledgements

• Thanks to my PhD supervisors Steve Sparks, Willy
  Aspinall and Gordon Woo
• Constanza Bonadonna for reconfiguring and running
  HAZMAP ash dispersal code
• DATA
• Ash data from Clare Horwell, Univ. Cambridge
• Personal exposure DustTrak data from The
  Institute of Occupational Medicine
• Montserrat Antigua rainfall data from the
  Montserrat Volcano Observatory and IOM
• Guadeloupe rainfall data from the Hong Kong
  Observatory www.hko.gov.hk
• CODE
• SCYTHE C Statistical Library GNU GPL ?2001
  A.D. Martin and K.M. Quinn
• MT19937 Mersenne Twister random number generator
  ?2002 T. Nishimura and M. Matsumoto

27
References

• Buchanan, D., B. G. Miller, et al. (2003).
  "Quantitative relations between exposure to
  respirable quartz and risk of silicosis."
  Occupational and Environmental Medicine 60(3)
  159-164
• Burmaster, D. E. and P. D. Anderson (1994).
  "Principles of Good Practice for the Use of Monte
  Carlo Techniques in Human Health and Ecological
  Risk Assessments." Risk Analysis 14(4) 477-481
• Cooke, R. M. (1991) Experts in Uncertainty
  Opinion and Subjective Probability in Science.
  Environmental Ethics and Science Policy Series.
  Oxford University Press, New York.
• Hughes et al. (1998) Radiographic Evidence of
  Silicosis Risk in the Diatomaceous Earth
  Industry. Am. J. Respir. Crit. Care Med., Volume
  158, Number 3, 807-814
• Horwell, C.J., Sparks, R.S.J., Brewer, T.S.,
  Llewellin, E.W., and Williamson, B.J. (2003). The
  characterisation of respirable volcanic ash from
  the Soufrière Hills Volcano, Montserrat, with
  implications for health hazard. Bull. Volcanol.,
  DOI 10.1007/S00445-002-0266-6.
• National Institute for Occupational Safety and
  Health (2002). NIOSH Hazard Review Health
  effects of occupational exposure to respirable
  crystalline silica.