CA-Markov Land-Use Change Model
Also known as: CA-Markov model, cellular automata Markov, land-use change simulation, CA-Markov arazi kullanımı modeli
CA-Markov is a hybrid spatio-temporal model that projects land-use and land-cover change by combining a Markov chain — which predicts how much of each class will change — with cellular automata, which decide where that change happens. Widely used for urban-growth and land-cover forecasting, it answers both the quantity and the location of change, something neither component does well alone.
Key highlights
- Predicts both the quantity (Markov) and location (CA) of land change.
- Calibrated directly from observed dated maps and driver layers.
- Supports scenario analysis (policy vs business-as-usual).
- Widely implemented and accepted in land-change science.
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 CA-Markov to project land-use/land-cover change and simulate urban growth or deforestation scenarios when you have at least two classified maps to calibrate transitions and driver layers to define suitability. It is a standard tool in land-change science, urban planning, and environmental management for exploring 'business-as-usual' versus policy scenarios. It assumes the calibration-period transition rates and drivers remain valid into the future (a strong stationarity assumption), is sensitive to the suitability layers and CA neighbourhood/iteration settings, and should be validated (e.g., by predicting a held-out date and comparing with metrics like Kappa or the figure-of-merit). For agent-driven or behaviourally rich dynamics, agent-based models are an alternative.
Strengths & limitations
- Predicts both the quantity (Markov) and location (CA) of land change.
- Calibrated directly from observed dated maps and driver layers.
- Supports scenario analysis (policy vs business-as-usual).
- Widely implemented and accepted in land-change science.
- Assumes calibration-period transition rates persist (stationarity).
- Sensitive to suitability layers and CA neighbourhood/iteration choices.
- Markov chain alone is non-spatial; quality hinges on the CA allocation.
- Requires careful validation against held-out data to be trustworthy.
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
Why combine cellular automata with a Markov chain?
A Markov chain predicts how much of each land class changes but has no spatial dimension — it cannot say where. Cellular automata allocate change spatially using neighbourhood rules and suitability but need a target amount. CA-Markov uses the Markov chain for quantity and the CA for location, covering both aspects of land-change forecasting.
What does the model assume about the future?
That the transition rates and suitability relationships estimated over the calibration period remain valid going forward (stationarity). This is a strong assumption; under policy shifts or new drivers it can fail, which is why scenario variants and validation against held-out data are important.
How is a CA-Markov projection validated?
Typically by calibrating on earlier dates, predicting a later date for which the true map is known, and comparing predicted versus observed change using metrics such as the Kappa family or, preferably, the figure-of-merit that accounts for correctly predicted change rather than persistence.
Sources
- 1.Clarke, K. C., Hoppen, S., & Gaydos, L. (1997). A self-modifying cellular automaton model of historical urbanization in the San Francisco Bay area. Environment and Planning B, 24(2), 247–261.
- 2.Muller, M. R., & Middleton, J. (1994). A Markov model of land-use change dynamics in the Niagara Region, Ontario, Canada. Landscape Ecology, 9(2), 151–157.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 2). CA-Markov. ScholarGate. https://scholargate.app/spatial-analysis/ca-markov