GIS Applications for Complex Human Disasters

1
GIS Applications for Complex Human Disasters

• Case Study of Kosovo
• By Sandy A. Johnson
• For APPD 799 - Dr. Coles

2
Posing a Question to Investigate(with an eye
towards policy and implementation)

• What towns are going to require the most
  resources for the recovery period?
• What type of assistance will be needed to rebuild
  each town?
• Are there hot zones where potential for disease
  or famine is higher?
• Are refugees really going to want to move back
  (i.e. security, motivation)

3
The Questions Evolve

• What towns are going to require the most
  resources for the recovery period?
• Which towns were attacked?
• What was the nature of the attack?
• What was the damage?

4
and Grow

• What type of assistance will be needed to rebuild
  each town?
• What did the town contain (transportation hub,
  telecommunications, hospitals, refineries, mills,
  etc.)?

5
and Grow...

• Are there hot zones where potential for disease
  or famine is higher?
• How might the lack of these resources impact
  town/regional recovery?

6
and Grow.

• Are refugees really going to want to move back
  (i.e. security, motivation)
• What was the pre-war population vs. the current
  population?
• Where did residents move to?
• Can they physically return (are there roads,
  trains, etc.)
• Can they afford to return (assistance, incentives
  and current living conditions)

7
The Search for Data Begins

• Geographic data
• Administrative boundaries
• Census data
• Town locations
• Roads, railways
• Elevation
• Had to adjust map to reflect boundaries after
  1991-92 breakup

Source CIA World Factbook www.cia.gov
8
Record of the Destruction

• New York Times daily coverage of the war
• Serbian attacks
• Refugee movement
• NATO airstrikes (number and location)
• Casualties
• Descriptions of type of destruction
• Accessible information
• Dont double count attacks

Source New York Times Photo Associated Press
9
Record of the Destruction

• Transform narrative into a database
• Maintain important details

10
Impact on Agriculture

• Examine municipal level productivity
• Did not use for class project due to insufficient
  data (lacked comparison over time)
• FAO crop reports

11
Merge the Databases

• Matching town names proved difficult (pre- and
  post-Yugoslav Republic names change).
• Towns from Maproom do not have names
• Use more detailed maps to name towns
• Still, about 1/3 of the conflict points are lost.

Source National Geographic, February 2000
12
Refine Fields in Database

• Code location of attack
• Code perpetrator

13
Metadata

• Country/Region in Former Yugoslav Republic
• 1 Kosovo
• 2 Montenegro
• 3 Bosnia
• 4 Serbia

• Country
• 1 Yugoslavia
• 2 Albania
• 3 Macedonia
• 4 Bulgaria

14
Define Class of Attacks

• 1 Intimidation
• 2 Injury/Fighting
• 3 Death
• 4 Property Destruction
• 5 Death and Property Destruction
• 6 Rape and/or Torture
• 7 Rape and Death
• 8 Death, Property Destruction, Rape/Torture

15
Define Attack Intensity

• 39 different towns/cities remained in database
  after merged.
• Tally number of incidents. Problems with this
  measure.
• Define look of intensity in ARCVIEW.

16
Establish Timeline

• Concentration of information on period from March
  21 to Mary 27, 1999.
• Have scattered incidents prior to that but
  difficult to establish actual time difficult to
  obtain verifiable reports.
• Ten week division
• At some point, would like to tie in weather
  patterns in the region.

17
Final Database
18
The Kosovo Conflict A Spatial View
ofPerpetrators, Targets and Intensity

• By Sandy Anne Johnson
• for APPD 799-03
• Prof. Coles
• April 1, 2000

19
Conflict in Kosovo
In the summer of 1998, the first rumors of
Serbian action against ethnic Albanians in the
Kosovo province of Serbia infiltrated Western
media. The Serbian campaign of intimidation and
violence escalated throughout the first quarter
of 1999. The rumors became fact with the
discovery in January 1999 of a mass grave in
Racak. Further campaigns of slaughter of Kosovo
civilians and increasing streams of refugees put
the Serbian military forces in the publics eye,
and pressure on the global community. Members of
the North Atlantic Treaty Organization (NATO)
took unilateral action against the Serbian army
in March. A bombing campaign was initiated with
the express goals of stopping Serbian aggression
and making Kosovo safe for the fleeing Albanians.
20
The bombing campaign was greeted initially with
increased violence by the Serbian military
against the ethnic Albanians in Kosovo. The
trickle of refugees into neighboring Macedonia,
Albania, and Montenegro grew to an estimated
800,000 by the end of April. NATO intensified its
bombing campaign. In May, a cease-fire was
tentatively set. Serbian troops pulled out of
Kosovo in June, 1999 and the province fell under
the bailiwick of the United Nations. The arduous
process of rebuilding Kosovo and resettling the
refugee began. This paper uses Geographic
Information Systems (GIS) analysis to examine the
dispersion and intensity of Serbian and NATO
military actions leading up to and during the
Kosovo conflict.
21
Objectives

• To examine where Serbian and NATO forces
  concentrated their attacks
• To chart the extent/types of violence in Kosovo,
  and show which parties committed these acts of
  violence
• To develop a timeline showing escalating or
  declining violence
• To examine strategic NATO targets and better
  understand success or failure of NATOs goals

22
Methodology
Based on coverage of the NATO campaign in The New
York Times from March 23, 1999 through May 26,
1999 I created a database of military actions. I
included major Serbian actions during the
previous six months. Each case in this database
is a separate incident of violence coded by date
and location. Where available, specific details
(i.e. deaths, injuries, incidence of rape, etc.)
are included. For the figures included in this
report, I merged my database with geo-referenced
information from the University of Pennsylvania's
website (http//www.maproom.psu.edu/dcw).
Metadata from this website included information
on railways and bridges which I used to create
Figure 5.
23
Based on information from The New York Times, I
developed a coding system for intensity of
violence. This, as well as my full database and
metadata, is available upon request. Data from
the University of Pennsylvania had many cities or
towns located on the map of Yugoslavia which had
no names. Where possible, I added names by
cross-references the U of P map to other maps.
Despite this, I was unable to match approximately
one third of the towns/villages mentioned in The
New York Times and so they are omitted from this
report. Most of the locations lost were sites of
Serbian campaigns. This points to an area for
further research developing a fuller map of
Serbian attacks to better understand both the
extent and pattern of Serbian attacks.
24
Figure 1 Classification of Military Actions
This figure shows the intensity of military
actions during the Kosovo conflict. The majority
of NATOs campaign involved property destruction.
Primary NATO targets were in Serbia. Of the
Serbian actions which could be tied to
villages/towns, the campaign of violence included
a higher degree murder and, unlike the NATO
campaign, torture and rape. What this figure
fails to show is the actual death rate associated
with each incident. The highest individual
casualty report from Serbian action was 100 each
in Kacenic, Celina, and Merdare. Unconfirmed
reports from other villages totaled 487. Total
reported moralities are 787. The highest
unconfirmed casualty report was 200 killed during
the NATOs numerous air raids of the major
population center of Beograd. Other casualties
totaled approximately 297. Total reported
moralities are 497.
25
Figure 1
26
Figure 2 Timeline of Serbian Attacks
This figure shows the pattern of Serbian attacks
concentrated largely within Kosovo. Most
reported incidents reflected a single attack -
the idea being Serbian forces then occupied the
small villages and refugees fled. Several cities,
such as Pristina, were subject to more than one
attack reflecting higher population densities and
attendant need for more than once campaign for
Serbian forces to capture or successfully clear
the town.
27
Figure 2
28
Figure 3 Bombing Intensity of NATO Air Raids
NATO air raids were directed against major
Serbian industrial and populations centers.
While this figures shows many of the targets were
scattered throughout Serbia, a number were
located in Kosovo. Pristina, in particular, was
subject to intense bombing campaigns designed to
destroy infrastructure, and fuel processing
plants. Given the devastation wrecked on the
populace by Serbian forces, and ensuing
evacuation, this target present a certain irony
in the NATO strategy in order to force the
Serbian military out of Kosovo, NATO destroyed
(at least in part) the very city refugees hoped
to return to.
29
Figure 3
30
Figure 4 Timeline of NATO Air Strikes
Figure 4 shows the major target of NATOs air
raids. Weeks four and nine proved particularly
heavy periods due largely to favorable weather
conditions. Major targets included Beograd, the
capital of Serbia, Nis and Novi Sad - both major
population centers and industrial areas. NATO
also targeted Pristina, a strategic Serbian
foothold in Kosovo, and Podgorica, a Serbian
supply terminal in Montenegro.
31
Figure 4
32
Figure 5 Strategic NATO Targets
As demonstrated in this figure, Beograd, Nis,
Novi Sad and Pristina served as strategic targets
for NATO bombing as they were crossroads for
telecommunication and transportation nexuses.
33
Figure 5