Transcript RISKandi

Geo-Information in animal
disease outbreak management
30th May 2008, Bilthoven
Henk Janssen, Bas Vanmeulebrouk, Igor Staritsky (Alterra,
Wageningen-UR)
Paul Leijs (Food and Consumer Product Safety Authority, Ministry of
LNV)
Presentation setup



Introduction
Knowledge and modelling system: Riskandi
Dutch GIS-pilots in operational animal disease
crisis management
Presentation setup



Introduction
Knowledge and modelling system: Riskandi
Dutch GIS-pilots in operational animal disease
crisis management
What is GIS?
Geographic Information System
GIS is a System of computer software,
hardware and data, methods and people
to help manipulate, analyze and present
information that is tied to a spatial location




Spatial location – a geographic location
Information – visualization of analysis of data
System – linking software, hardware, data
People– a thinking explorer who is the key to the power of
GIS
EARLY APPLICATION OF GIS FOR CHOLERA
CONTROL IN 19TH CENTURY LONDON
Dr. John Snow
(1813 -1858)
Concentration of cholera cases around pump sites
Modern mapping of John Snow’s work on cholera
using GIS.
RISK ANimal DIseases:
Knowledge system and electronic atlas
RISKANDI is a Decision
Support System to assess
ex ante the effectiveness
of spatial strategies to
diminish the risks of spread
of animal diseases
RISKandi: Example
Risk for transmission in the neighbourhood of an
infected farm (CSF):


different intervention strategies
different infectious periods
Different intervention strategies
Edo Gies en Hans Baveco (Alterra), Gert-Jan Boender en Gonnie Nodelijk (ASG)

Culling only the infected herds

Culling within a radius around the infected herds
(500 m, 1 km, 1.5 km)

Vaccination within a radius around the infected
herds
(500 m, 1 km, 1.5 km, 2 km)
Different infectious periods
BOP5: Edo Gies en Hans Baveco (Alterra), Gert-Jan Boender en Gonnie Nodelijk (ASG)
Infectious period
Infection
Start infectious period
Detection
Start culling
End culling
Control-delay
Latent
Control-duration
Three infectious periods: 3, 7 and 9 weeks
Culling of infected herd after 3 weeks
infection
Culling of infected herd after 7
weeks infection
7
Culling of infected herd after 9
weeks infection
When the infection
period is 9 weeks
culling within a 1-km
radius is not enough to
prevent a big outbreak
Conclusion:




RISKANDI is usefull to answer ‘what if’ questions
RISKANDI can identify locations having highest risk for
spread of CSF
RISKANDI can qualify effects of intervention strategies on
the risk of spread of CSF
RISKANDI can quantify average costs of culling (without or
after vaccination), vaccination costs or costs due to
standstill?
(per grid cell of 250-m2, and within a radius)
Operational research questions of VWA
(Food and Consumer Product Safety Authority in NL)




VWA needs to maintain quarantine areas
VWA needs timely, correct and complete farm and area
information in case of contagious animal disease outbreak
VWA desires online (internet-based) GIS-applications as
integrated part of their workflow
Our question: can we prepare adequate datasets and
software tools?
Workflow during animal crisis situations
Farm -> suspicious
symptoms
-Veterinarian or other
source
Letter of suspicion
Specialists team (Animal Health
Service and VWA)
within 3 h
- Clinical inspection
- Sampling
- Filing
- Declaration of suspicion
Positive testing
INFORMATION DEMAND
ACTION !
- location of farms
- Surveillance zone (10km)
- location and number of animals
- Protection zone (3 km)
- geographic relations
- Culling of infected holding + 1
km zone
Pilot web-GIS-applications for use in contagious
animal disease crises for the Dutch Ministry

PIKET (English: Picket


but also Picket Officer )
Application for geographical determination and textual description of
quarantine areas (Surveillance zone + Protection zone)
GEMMA (Geographic Event Management and Monitoring
Application)



Data-selection tool, based on information from Dutch LPIS (‘BRP’), I&R,
GIAB (farm location (x,y on shed level)
Internet based
Time line
Approach



Data: testing of quality of available datasets
Process analysis: embedding GIS in workflow
Application: The proof of the pudding…
Coupling of GIAB and crop parcels
34 %
10 % farm ended
GIAB-2005:
AHS-2005:
44 %
• unique farm
locator (ubn)
• NAW
• Cattle
• Sheep/Goats
• Pigs
• Name Address
• x-y
• brs-nr
• 180.639 loc.
• Name Address
Ubn-Brs
92 %
• ubn-nr
• brs-nr
64 %
12 %
Annual Census 2005:
77 %
• Cattle
•Sheep/Goats
LPIS
• Pigs
• brs-nr (unique)
• Poultry
• Name Adress
• Horses
• Crop
• Crops
• x-y
Coupling of parcels to farm locations
Combination
% parcels
% grassland
UBN - BRS
64
79
Postal code house nr
32
18
LPIS – Annual Census
3
2
No coupling
1
1
Visualize
data
quality
Light green =
arable crops
Conclusion:



VWA/LNV have all relevant data
Data quality is good
GIS helps to show ‘what we don’t know’
Workflow: first step-> describe quarantine
area
Suspicion site (roads, rail, water way)
Suspicion site (only roads)
Suspicion site (roads + junctions)
Suspicion site (barricaded junctions)
Suspicion site (‘stops’ around zone)
Suspicion site (route around zone)
Suspicion site (description of route)
Suspicion site (with aerial photo)
Conclusion:

GIS can be used to automate the process of
describing quarantine areas
Workflow: Data selection tool
Active layers
selectable by placing
a ‘tick’, the ‘i’ makes
the data of the geoobjects ‘identifiable’
Identify to show
details
Map layers are
served using ArcIMS
(ESRI)
Creation of a new suspicion
New suspicion by
selecting a
agricultural holding
Pan and zoom in the
map (standard
functionality)
New suspicion by
selecting a location
on the map
Circles with
adjustable radius
Selection and presentation of farms
Foot and Mouthsuspicion
Initial suspician
Different sub-selects
(based on defined
circles)
Farm with parcels
within the circles
Interactive drawing of boundaries
Digitizing tools
Time line tool
Digitized area with a
name
Second suspicion
Presentation of the timeline of suspicions
Tool to release a
certain suspicion
Deleting a suspicion
Store suspicions
Retrieve suspicions
from the database
‘Free’ holdings in
green
Timeline selector
‘Free’ holding
restricted by second
suspicion
Details of a holding
Inspection tool (for
details of the farm)
Data from I&R and
Annual Census
Parcels of the
selected holding
Holding 221449 in
Google Maps
Holding 221449
Reports
A printable map
(PDF)
All holdings in
Excell
Holding 221449
Support of crisis simulation/training
Support of crisis simulation/training
Support of crisis simulation/training
Concluding remarks
Work flow can be supported with GIS
 Challenge: operational integration of data,
GIS-tools and end user process
 Kostunrix metaphore
 It is all about data quality Data needs to be:





Complete,
Timely,
Correct
Available
Thank you for your attention