Transcript Characterizing spatiotemporal patterns of fire season fuel moisture

A Remote Sensing Concept for
Mapping Parameters of
Infectious Disease
Stephen R. Yool, Ph.D.
Associate Professor
Geography and Regional Development
[email protected]
What do we need to model
infectious disease?
•
•
•
•
Solid theory or theories of causality
Data and Methods at causal scale
Unquenched thirst for knowledge
Congenital sense of adventure
Do Satellite Data Support
Infectious Disease Modeling?
General satellite data characteristics
– Collected over long time scales
– Collected at fine spatial scales
– Collected over large geographic areas
The Valley Fever Example
Valley fever (coccidioidomycosis) is a disease
endemic to arid regions in the Western
Hemisphere, and is caused by the soil-dwelling
fungi Coccidioides immitis and Coccidioides
posadasii.
Arizona is currently experiencing an epidemic
with almost 4000 cases annually, greatly
exceeding other climate-related diseases such
hantavirus or West Nile Virus.
Mapping/Modeling Needs Map
Span a Large Geographic Areas
Arizona’s Valley Fever Epidemic
Reported Arizona Coccidioidomycosis Cases
4000
3500
3000
2500
2000
1500
1000
500
0
1990
1992
1994
1996
1998
2000
2002
2004
Coccidioides Life Cycle
Linking Precipitation and Dust to Incidence
(Source: Comrie, 2005)
7
Observed
Predicted
Incidence (per 100,000)
6
5
4
3
2
1
0
1992
1993
1994
1995
1996
1997 1998 1999
Year (Seasons)
2000
2001
2002
2003
The Moisture Stress Index (MSI)
• By converting the NDVI value for each pixel into Z-score,
we produce for each pixel a Moisture Stress Index
(MSI)—expressing the pixel’s distinctive moisture stress
at specific time within the complete time series.
• The Z score represents the distance in standard
deviations of a sample from its population mean
Z = [(Xi - XMEAN) / XSD]
• Then, MSI =
- [(NDVIi,j,t - NDVIMEAN) / NDVISD]
So the MSI is a measure at a specific time of the distance in
standard deviations of a pixel’s moisture stress from its
mean (average) moisture stress across that pixel’s
complete time series.
(The negative sign inverts the values, so pixels with low
scores get mapped as bright, moisture-stressed pixels.)
Late Summer MSI: Monsoonal
Rains Promote Fungal Growth
Arid Foresummer MSI: The
Southwest is Dry, promoting
endosporulation
Sample Moisture Stress Map
Tucson length of moisture stress
The Coccidioidomycosis Model
• Dispersion-related conditions are important
predictors of coccidioidomycosis incidence
during fall, winter and the arid foresummer.
• Comrie (2005)* reported precipitation during the
normally arid foresummer 1.5-2 years prior to
the season of exposure is the dominant
predictor of the disease in all seasons,
accounting for half of the overall variance.
* Comrie, A.C., 2005. Climate factors influencing coccidioidomycosis
seasonality and outbreaks. Environmental Health Perspectives,
doi:10.1289/ehp.7786.
We deploy spaceborne sensors,
such as this Advanced Very High
Resolution Radiometer (AVHRR),
which produces 1km pixels we use
to map surface moisture dynamics
What can the spectrum of
vegetation tell us about surface
moisture?
A Spectral Index of Moisture Stress
• Dry leaves show an increase in the red
(Red) wavelengths and a decrease in the
near-infrared (NIR) wavelengths
• We can represent this relationship as a
Normalized Difference Vegetation Index
(NDVI), which we can compute from
spaceborne satellite data using this simple
equation:
NDVI = (NIR – Red / NIR + Red)
But how can you use an NDVI time series to
measure moisture stress in highly diverse settings?
Technology may be the answer, but
what was the question?
• Will human societies on our planet
promote actively the alliances between the
natural and social sciences required to
manage infectious disease effectively?
• Remote sensing empowers new and novel
views of a world in which natural and
human dimensions must co-exist.
• The multi-scale requirements of
epidemiology and mapping technology
can come together: To perceive unity in
diversity, to focus on conflict resolution
and consensus building—to move the
process of disease hazard management
forward.