Risk Terrain Modeling (Criminology)
Also known as: RTM, Risk Terrain Analysis, Environmental Risk Factor Modeling, Spatial Risk Factor Modeling
Risk Terrain Modeling (RTM) represents crime risk as a function of the environment: it identifies the features of a landscape — bars, bus stops, vacant lots, pawn shops, schools — that attract or generate crime, maps each one's spatial influence as a separate risk layer, and combines those layers onto a raster of place to produce a relative risk surface. Introduced by Joel Caplan and Leslie Kennedy around 2011, RTM 'brokers' environmental criminology theory and GIS methods so that crime forecasting rests on the qualities of places rather than on the history of crime alone.
Key highlights
- Forecasts risk from durable environmental features, so it can flag high-risk places that have little prior crime, enabling prevention rather than reaction.
- Names the specific, often modifiable risk factors driving each location, directly informing place-based and problem-oriented interventions.
- Grounds crime mapping in environmental and routine-activity theory rather than in unexplained clusters of past incidents.
- Combines heterogeneous data sources (businesses, land use, infrastructure) into a single interpretable relative risk surface.
- Supports empirical validation, since factor weights are estimated and predictive accuracy can be tested on held-out crime.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use RTM when you want to explain and forecast crime in terms of the environmental conditions of places — and especially when you want to support prevention by acting on modifiable features rather than chasing past incidents. It suits settings with good georeferenced data on candidate risk factors (land use, businesses, infrastructure) and enough crime events to estimate factor effects. It is well matched to problem-oriented policing and place-based interventions because it names the specific features driving risk. It is less appropriate when relevant environmental data are missing or coarse, when crime is too sparse to estimate factor weights, or when the question is purely about short-term temporal contagion (where near-repeat methods fit better). RTM identifies risky places, not offenders.
Strengths & limitations
- Forecasts risk from durable environmental features, so it can flag high-risk places that have little prior crime, enabling prevention rather than reaction.
- Names the specific, often modifiable risk factors driving each location, directly informing place-based and problem-oriented interventions.
- Grounds crime mapping in environmental and routine-activity theory rather than in unexplained clusters of past incidents.
- Combines heterogeneous data sources (businesses, land use, infrastructure) into a single interpretable relative risk surface.
- Supports empirical validation, since factor weights are estimated and predictive accuracy can be tested on held-out crime.
- Output quality depends heavily on the availability, accuracy, and completeness of georeferenced risk-factor data.
- Results are sensitive to analyst choices — which factors to include, their spatial operationalization, and the raster cell size.
- It identifies risky places, not offenders, and does not by itself explain the causal pathway from feature to crime.
- Risk factors are often correlated with one another, complicating attribution of risk to any single feature.
- A static risk terrain can become outdated as land use, businesses, and the built environment change over time.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How does RTM differ from hot-spot or kernel density mapping?
Hot-spot and kernel density methods map where crime has clustered in the past and project that pattern forward. RTM instead models why places are risky by combining environmental features into a risk surface, so it can highlight conducive locations even before crime concentrates there. In practice the two are complementary: hot-spot maps describe the recent past, while RTM explains and forecasts risk from the qualities of place.
How are the risk factors and their weights chosen?
Candidate factors are drawn from environmental criminology theory and local knowledge, each operationalized as a proximity or density layer with a spatial reach justified a priori. RTM then enters the layers into a regression of observed crime counts per cell and keeps the factors whose estimated coefficients are statistically meaningful, using those coefficients as weights. This empirical selection guards against arbitrary hand-tuning and lets the surface be validated against actual crime.
Does a high-risk cell mean crime will definitely occur there?
No. The relative risk score ranks how conducive a place's environment is to crime, not a guarantee. High-risk cells are where prevention is most likely to pay off, but crime is probabilistic and depends on offenders, opportunities, and timing that the static terrain does not capture. RTM is best read as a prioritization tool whose forecasts should be evaluated against held-out crime.
Sources
- 1.Caplan, J. M., Kennedy, L. W., & Miller, J. (2011). Risk terrain modeling: Brokering criminological theory and GIS methods for crime forecasting. Justice Quarterly, 28(2), 360–381.
- 2.Caplan, J. M., & Kennedy, L. W. (2016). Risk Terrain Modeling: Crime Prediction and Risk Reduction. University of California Press.ISBN 9780520282933
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Risk Terrain Modeling (Criminology). ScholarGate. https://scholargate.app/criminology/risk-terrain-modeling-criminology