Файл: [D._Kim_Rossmo]_Geographic_Profiling(BookFi.org).pdf

© 2000 by CRC Press LLC

1994). The first such profile was prepared in 1990. To date, requests have
come from a variety of federal, provincial, state, and local law enforcement
agencies, including the RCMP, the FBI, and Scotland Yard. The cases involve
crimes of serial murder, serial rape and sexual assault, serial arson, serial
robbery, serial exposers, sexual homicide, and kidnapping.

The implementation of geographic profiling within the policing com-

munity is a direct result of the SFU study on target patterns of serial mur-
derers. This is an example of criminological research, done in an academic
university setting, having practical policy and procedural implications for
the criminal justice system. While the original research focused on serial
killers, the hunting patterns of serial rapists, arsonists, and other repetitive
violent offenders are similar and can also be geographically profiled. This
resemblance is not surprising considering what the relevant theory tells us.
Murder and violent sex crimes comprise the two single largest categories in
geographic profiling casework, followed by arson and robbery.

10.4.1 Profiling Considerations

A geographic profile is only one part of what is usually a large and complex
police investigation. Understanding its role in the overall picture and its
relationship to other behavioural science methods is helpful. The following
sequence outlines how geographic profiling typically fits into a criminal
investigation:

1. Occurrence of a crime series;
2. Traditional investigative techniques;
3. Linkage analysis;
4. Criminal profile;
5. Geographic profile; and
6. New investigative strategies.

The ability to identify the existence of a serial offender is the starting

point of the process. The extent and accuracy of the linkage analysis is
important as the more crimes, the more accurate the geographic profile. But
providing there is no spatial bias, and not too many locations are missed,
unlinked crimes are not a critical problem. The CGT algorithm is also quite
robust and its results are not significantly affected by the mistaken inclusion
of an unconnected crime; generally, at least 90% of the information should
be accurate.

If traditional investigative methods are successful, then there is unlikely

to be a need for profiling. The development of new investigative strategies
based on the profile is the topic of the following chapter. That leaves the
relationship between geographic and psychological profiling.

© 2000 by CRC Press LLC

While primarily empirical, geographic profiling has both quantitative

(objective) and qualitative (subjective) components. The objective compo-
nent uses a series of geostatistical techniques and quantitative measures, such
as the CGT program, to analyze and interpret the point pattern formed by
target sites. Because the validity of these measures depends upon number of
locations, they are inappropriate for smaller crime series. The subjective
component of geographic profiling is based upon the reconstruction and
interpretation of the offender’s mental map (Homant & Kennedy, 1998). A
criminal profile is not a necessary precursor for a geographic profile, but the
insights it provides to offender personality, behaviour, and lifestyle are useful,
particularly in cases involving only a few locations. A geographic profile, in
turn, helps refine a psychological profile, focus its application, and increase
its utility. The two types of profiles optimize each other, and act in tandem
to help investigators develop a “picture” of the person responsible for the
crimes in question.

For example, in 1996 the RCMP investigated a series of 14 arsons of

house carports in Burnaby, British Columbia, and requested psychological
and geographic profiles. When a suspect who closely matched both profiles
came to the attention of the investigator, he carefully focused on him, initi-
ating surveillance and then interviewing strategies. The offender confessed.
He closely matched the criminal profile, and literally lived across the street
from the highest part of the geoprofile, in 0.6% of the hunting area (0.02 mi

2

).

Many different crime factors and environmental elements are considered

in the construction and interpretation of a geographic profile. The most
relevant ones include:

1. Crime locations — Offence locations and times are the most impor-

tant data in a geographic profile. Also significant are the number and
types of crime sites, their parsing, and the crime location set.

2. Offender type — The type and number of offender(s) affect crime

geography. If multiple criminals living apart are involved, the geopro-
file will focus on the dominant one’s residence. Large, amorphous
gangs may not be suitable for geographic profiling because of changing
group composition. Psychological profiling assists in interpreting
offender behaviour by providing information on personality, back-
ground, and level of organization.

3. Hunting style — Criminal hunting methods influence both encounter

and body dump site target patterns. Hunting style must be considered
when preparing a geographic profile.

© 2000 by CRC Press LLC

4. Target backcloth — Constrained or patchy target backcloths limit the

degree of offender choice, affecting the importance of certain crime
site types for the profile.

5. Arterial roads and highways — People, including criminals, do not

travel as the crow flies. Not only must they follow street layouts, but
they are most likely to travel along major arterial routes, freeways, or
highways.

6. Bus stops and rapid transit stations — Offenders without vehicles may

use public transit or travel along bicycle and jogging paths. The loca-
tions and routes of these should be taken into consideration.

7. Physical and psychological boundaries — People are constrained by

physical boundaries such as rivers, ocean, lakes, ravines, and highways.
Psychological boundaries also influence movement. For example, a
criminal of low socioeconomic status may avoid an upper class area,
or a black offender might not wish to go into a white neighbourhood.

8. Zoning and land use — Zoning (e.g., residential, commercial, indus-

trial) and land use (e.g., stores, bars, businesses, transportation cen-
tres, major facilities, government buildings, military institutions)
provide keys as to why someone may be in a particular area. Police in
Britain conduct site surveys and location inventories in the area sur-
rounding a crime to help identify what may have brought an offender
to a particular location. Similarly, information about the peak area in
a geoprofile provides insight to a criminal’s anchor point, and zoning
classification helps determine if this is a residence or workplace. For
example, the geographic profile for a series of bank robberies occur-
ring during the early afternoon pointed towards a commercially zoned
area. Time and location factors correctly suggested an offender who
was committing the crimes during his lunch break from work.

9. Neighborhood demographics — Some sex offenders prefer victims of

a certain racial or ethnic group. These groups may be more common
in certain neighbourhoods than in others, affecting spatial crime pat-
terns.

10. Victim routine activities — The pattern of routine victim movements

provides insight to how the offender is searching for targets.

11. Singularities — Single offences that do not appear to fit the overall

pattern of the crime series may provide important clues and should
be carefully reviewed.

12. Displacement — Media coverage or uniformed police presence can

cause spatial displacement, affecting the locations of subsequent crime
sites.

© 2000 by CRC Press LLC

Crime locations are the basis of a geographic profile, and a given murder

can involve separate encounter, attack, murder, and body dump sites. But
other locations that are not crime sites per se may also be connected to an
offence. Examples of such locations include credit or bank card use, mailings,
telephone calls, vehicle rentals or drops, witness sightings, and found prop-
erty or evidence sites. In these cases, it may be possible to geographically
profile a single crime, depending upon the number and types of locations.

In October 1995 two teenage girls in the municipality of Abbotsford,

British Columbia, were attacked on the street at night by a man with a
baseball bat. One victim was murdered and dumped in the Vedder Canal,
some 20 miles away; the other was left for dead, but she somehow managed
to revive and make her way to a nearby hospital. A few days later, the
Abbotsford Killer began a series of bizarre actions starting with several taunt-
ing 911 telephone calls. He then stole and defaced the murder victim’s grave-
stone and dumped it in the parking lot of a local radio station. Finally, he
threw a note wrapped around a wrench through a house window; in the
note, he admitted to other sexual assaults. These actions provided 13 different
sites for the geographic profile. He was eventually caught through a local-
based strategy initiated by the Abbotsford Police Department. His residence
was in the top 7.7% of the geoprofile (0.6 mi

2

).

A geoprofile may result in two peak areas, an indication the offender

has more than one anchor point. Manley Eng, responsible for a series of
arsons in Saanich and Victoria, British Columbia, left a crime pattern that
resulted in dual peaks — one which contained his residence, and the other,
his probation office. Information regarding land use, zoning, and area
characteristics help interpret such outcomes. Examples of multiple anchor
points include:

• Residence and work sites;
• Residence and social or family sites;
• Present and previous residences; and
• Two or more offenders living apart.

Between 1994 and 1998 the Mardi Gra Bomber was responsible for a

total of 36 explosive devices, most in the Greater London area. These were
mailed or delivered to locations near bank machines, supermarkets, pay-
phones, businesses, and residences. Even though the targets and delivery
methods varied, the underlying spatial pattern of the crimes remained con-
sistent. Scotland Yard requested a geographic profile which produced two
high probability regions; a primary area around Chiswick in west London,
and a secondary peak in southeast London. When police detectives arrested
two elderly brothers, it turned out they lived in Chiswick, and their family

© 2000 by CRC Press LLC

resided in southeast London. The geoprofile identified the convicted
offender’s home in the top 3.4% (9.1 mi

2

) of the hunting area.

Investigators may benefit from a geographic perspective on their crimes,

independent of a formal profile. It is not just what offenders do, but also
what they do not do, that is of interest. Some specific questions worth con-
sidering include:

1. Locations — What are the locations connected to this crime or crime

series? Where are they? What are the distances and travel times
between them?

2. Time — When did the crimes occur (i.e., time, weekday, date)? What

was the weather on those dates? What are the time lags between
crimes?

3. Site selection — How were the crime locations accessed? What else is

in their general area? How might the offender have known of these
locations? What criminal purpose or function did they serve?

4. Target backcloth — What is the geographic arrangement and avail-

ability of the target group? What degree of control did the offender
have over the choice of crime locations? Has displacement (spatial or
temporal) occurred?

5. Hunting — What hunting method did the offender use? Why were

these sites chosen, and not other possible locations? What was the
offender’s probable mode of transportation?

10.4.2 Operational Procedures

10.4.2.1 Information Requirements

Certain operational procedures are followed in the request for and preparation
of a geographic profile. Assessing the appropriateness of the crimes, collection
of necessary information, and coordination between the profiler and the inves-
tigator are necessary. The following is the content of a standard letter sent by
the Vancouver Police Department to investigators requesting a geographic
profile. It outlines basic information requirements and procedural details:

Geographic profiling is an investigative support technique for serial violent
crime investigations. The process analyzes the locations connected to a series
of crimes to determine the most probable area of offender residence. It
should be regarded as an information management system designed to help
focus an investigation, prioritize tips and suspects, and suggest new strate-
gies to complement traditional methods.

A preliminary assessment is necessary to determine if a case is appropriate
for geographic profiling. This is best done by directly contacting the Van-