Feedback Loop Example: Wildland Fire Spread
Do you have questions or comments about this model? Ask them here! (You'll first need to log in.)
WHAT IS IT?
This model is a replication of that described by Peterson (2002) and illustrates the 'spread' feedback loop type described in Millington (2013). In spread feedback, spatial patterns influence spread processes which influence later spatial patterns and hence later spread processes. In ecology, this concept of processes being shaped by their history is known as the memory of the process (Peterson 2002). For wildland fire, memory _ and therefore feedbacks between spread and spatial pattern _ has been conceptualized as being contained in vegetation flammability as a function of time since last the fire. In this model version red patches indicate areas burned in a given time step, green patches the length of time since the last fire at that location (lighter, fire more recently).
HOW IT WORKS
The simulated environment contains vegetation, the flammability (probability of spread, pfs) of which is governed by time since last fire (tsf). In each timestep a given number of fires are ignited at randomly selected locations in the simulated environment. Fires spread to neighbouring locations determined by the vegetation flammability. The number of fires ignited in each timestep is set by the user, as is the relationship between pfs and tsf).
HOW TO USE IT
First, set the number of ignitions per timestep by moving the ignition-rate slider (larger number will result in more fires per timestep). Next, determine the relationship between tsf and pfs by moving the log-alpha and tp-max sliders. The relationship between tsf and pfs for the selected values can be observed in the TSF Curve plot by clicking setup (the number of fires per timestep will also be displayed in the ignitions monitor). See Peterson (2002, p.332) for details of how log-alpha (via alpha, the memory parameter) and tp-max influence the shape of the TSF Curve.
Click Go to run the simulation. If the model runs slowly on your machine you can try reducing the total area of the environment (right-click environment, click edit, then reduce the world-width and world-height values).
THINGS TO NOTICE
Notice how recently burned patches are less likely to burn in a given time step, and so the pattern of previous burns influences contemporary burns (with fires more likely to burn dark areas in the previous time step). The proportion of the landscape with different tsf and pfs values are shown by the histogram plots.
THINGS TO TRY
Try changing the log-alpha value (producing different values for alpha, the memory parameter) to change the TSF Curve and the ignition-rate value to change the number of ignitions per timestep. Look at how these changes influence the size of fires and how restricted they are by patterns of vegetation.
CREDITS AND REFERENCES
Millington, J.D.A. (2013) Three types of spatial feedback loop in coupled human and natural systems. Ecology and Society [URL HERE]
Peterson, G. D. (2002). Contagious disturbance, ecological memory, and the emergence of landscape pattern. Ecosystems, 5(4), 329-338.
Code licenced by James D.A. Millington (http://www.landscapemodelling.net) under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License (see http://creativecommons.org/licenses/by-nc-sa/3.0/)
Model and documentation available from http://openabm.org
Comments and Questions
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;; ;; ;; Feedback Loop Example: Wildland Fire Spread December 2012 ;; ;; ;; ;; Code licenced by James D.A. Millington (http://www.landscapemodelling.net) ;; ;; under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 ;; ;; Unported License (see http://creativecommons.org/licenses/by-nc-sa/3.0/) ;; ;; ;; ;; Model and documentation available from http://openabm.org ;; ;; ;; ;; Model used in: ;; ;; Millington, J.D.A. (2013) Three types of spatial feedback loop in coupled ;; ;; human and natural systems. Ecology and Society [URL HERE] ;; ;; ;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; globals [ alpha p-max ignitions fire-front max-tsf ] patches-own [ tsf pfs ] to setup ca reset-ticks set alpha 10 ^ log-alpha set p-max 0.24 set ignitions ignition-rate * (world-height / 100) ;this is more feasible ask patches [ set tsf 1 set pfs set-pfs tsf set pcolor scale-color lime tsf tp-max 0 ] update-plots end to-report set-pfs [ age ] ifelse(age > tp-max) [ report p-max ] [ report p-max * (age / tp-max) ^ alpha ] end to go ask patches [ set pcolor scale-color lime tsf tp-max 0 ] repeat ignitions [ ignite ] ask patches [ set tsf tsf + 1 set pfs set-pfs tsf ] set max-tsf max [tsf] of patches tick end ; Code for 'ignite' procedure by George L.W. Perry, University of Auckland (http://web.env.auckland.ac.nz/people_profiles/perry_g/) to ignite ask one-of patches [ set tsf 0 set fire-front patch-set self ;; a new fire-front spread ] end ; Code for 'spread' procedure modified from code by George L.W. Perry, University of Auckland (http://web.env.auckland.ac.nz/people_profiles/perry_g/) to spread while [ any? fire-front ] ;; stop when we run out of fire-front [ let new-fire-front patch-set nobody ;; empty set of patches for the next 'round' of the fire ask fire-front [ set pcolor red let N neighbors with [ tsf > 0] ask N [ if random-float 1 <= pfs [ set tsf 0 set new-fire-front (patch-set new-fire-front self) ;; extend the next round front ] ] set fire-front new-fire-front ] ] end
There is only one version of this model, created about 12 years ago by James Millington.
Attached files
File | Type | Description | Last updated | |
---|---|---|---|---|
Feedback Loop Example: Wildland Fire Spread.png | preview | Preview for 'Feedback Loop Example: Wildland Fire Spread' | over 11 years ago, by James Millington | Download |
This model does not have any ancestors.
This model does not have any descendants.