Approaching a Case
A persistent problem in many criminal investigations is "linkage blindness." That is, investigators fail to recognize when certain crimes have important links to other crimes because they're focusing too much on dissimilarities. For example, in the Zodiac crimes, the cab driver was not immediately linked in the series because previous attacks had been aimed at couples. In other words, investigators had developed a tentative psychological map that did not account for the Zodiac's apparent reasoning. The need to determine links is one of the motivations for developing the FBI's VICAP (Violent Criminal Apprehension Program), but programs are only as good as the data entered into them.
Geographic profiling (sometimes known as geoforensics, but to be distinguished from geoforensics that involves geological profiling) helps to organize an abundance of information via geographical links in order to accelerate the apprehension process.
"Geographic profiling enables you to focus the investigation in a small area of the community," says Inspector Glenn Woods of the Royal Canadian Mounted Police, "rather than on the whole metropolitan area." That means it cuts down on the amount of time and resources required for what can shape up to be a major investigation.
For example, in the case of Rafael Resendez- Ramirez, police agencies in three states noted that a series of murders had taken place near railroads. First in 1997 a couple was attacked near tracks in Lexington, Kentucky, and the boy was murdered. The following year, Dr. Claudia Benton was viciously killed in her Houston-area home, which was also near railroad tracks. Less than five months later, still in Texas, a minister and his wife were bludgeoned to death in the church parsonage. In part because of the track connection, an evidence match was made between this double murder and the killing of the doctor, and a suspect was identified who had been repeatedly deported back to Mexico. The search became focused on Resendez-Ramirez.
Then in only two days, two more women were found murdered in their Texas homes, but the killer had already moved on to Illinois. There he tied up and shot an 80 year-old man in Illinois and killed his daughter with the butt of the shotgun. Both were left inside their trailer, which stood about one hundred feet from a track.
The Railroad Killer had left enough fingerprints at the various crime scenes to be clearly identified. He was eventually apprehended, tried, and convicted for the rape-murder of the physician, although he was clearly linked to the seven other deaths as well. That he stayed near trains had indicated to investigators his on-the move, drifter-type mentality, so it was not difficult to narrow down the suspect list and find this Mexican immigrant who had been in trouble with the law for years. Although he was not officially analyzed with a computerized profiling system, the geographical similarities helped to rapidly link him to successive crimes.
It was Vancouver detective inspector Kim Rossmo who in 1995 wrote a doctoral dissertation at Simon Frasier University's School of Criminology on a method of geographical profiling that has become an oft-cited source. The first such profile was done out of his police department in 1990 and Rossmo's seminal work has been utilized extensively in Canada and the U.K., with more limited application in the U.S.
Preferring to work with a psychological profile as part of the data, he feels that a geographical analysis highlights the crime location, any physical boundaries that were present (that might not otherwise be noticed), and the types of roads and highways that come into both the abduction and body dump sites. He also notes the importance of the routine activity of the victims, because people tend to stick with familiar territory. That means that an analysis of all the crime scenes could provide clues about where an offender lives.
Like psychological profilers, those who concentrate on geographical analysis are also trying to determine how sophisticated and organized an offender is, whether the crime was planned or opportune, and whether the offender approached a high or low risk victim. However, they are also trying to take it a step further to use objective measurements to pinpoint as precisely as possible the locus of criminal activity.
The construction of a geographical profile involves:
- Complete familiarity with the case file
- Examination of the crime scenes
- Interviews with investigators and witnesses
- Study of area maps
- Analysis of neighborhood demographics for both the abduction site and body dump site
- Computerized analysis
To assist in the scientific angle, Rossmo developed a computer program, Criminal Geographic Targeting (CGT), which assesses the spatial characteristics of a crime. (Environmental Criminology Research, Inc. developed a prototype called Rigel.) Using specific measurements, the program makes numerous calculations and produces a topographic map based on the locations of a series of similar crimes. This map shows via color arrangements and graphs the "jeopardy surface," or likelihood that some area is the location of the killer's home or base of operation. It is then superimposed on a street map on which the crimes are pinpointed, which are thought of as "fingerprints" of the offender's cognitive map. The program's predictive power is related to the number of crime sites, and the more the better.
To provide these results, CGT takes into account known movement patterns, comfort zones, and hunting patterns. Right-handed criminals who are trying to escape in a hurry will flee to the left, and will discard weapons to the right. When lost, males go downhill while females go up.
Initially Rossmo tried out his program on cases that had been solved to see how accurate it was. For example, Robert Clifford Olson was arrested in 1981 in Vancouver, B.C., for picking up two hitchhikers. He subsequently confessed to eleven murders, mostly young girls and boys. Rossmo generated a map of his crimes and pinpointed within a four square block are where Olson had actually lived.
Rossmo prefers to have at least five crimes in a clearly linked series to analyze, or at least five crime activity sites. From that, and from records such as suspect lists, police reports, and motor vehicle information, he enters information into his program and builds from there. It can be added to or modified whenever new data comes in. In the case of a serial rapist, for example, Rossmo used 79 of the more than 100 crime scenes and came up with a red dot on his computerized map that turned out to be the very spot to where the man lured his victimshis basement.