Skip to main content


Stochastic Dynamic Programming (SDP)

Using the resource
Potential benefits from using the resource
Able to incorporate complex, nonlinear, dynamic relationships between management actions and outcomes. 
Can incorporate effects of stochastic events.
Allows for optimal approach to differ depending on the state of the system.
Potential limitations from using the resource
Requires highly detailed knowledge of cause-and-effect pathways in order to incorporate complexity.
The capacity to capture greater realism in SDP is attractive, but computational overheads, the curse of dimensionality and the requirement for sophisticated causal understanding mean that most applications are substantially simplified
Practical information
UN languages in which the resource is available:
Development stage:
Full, working product
Contact details
IPBES Secretariat

Stochastic dynamic programming is a single-objective optimisation approach which employs algorithms designed to optimise an objective function under specified constraints. Optimisation approaches can be viewed as providing the analytical machinery to assist in the generation and analysis of ‘target-seeking’ or ‘backcasting’ scenarios. With a detailed understanding of cause-and-effect, stochastic dynamic programming can accommodate non-linear, dynamic outcomes associated with stochastic risk (e.g. risks associated with wildfires) superimposed on the deterministic influence of management actions (e.g. fuel reduction burning in high fire risk places). SDP recognises that what might be considered a desirable action depends on the state of the system.


Subregions covered