willvieira / stmanaged Goto Github PK

View Code? Open in Web Editor NEW

1.0 3.0 1.0 43.75 MB

:package: R package for the State and Transition Model with forest management

Home Page: https://willvieira.github.io/STManaged/

License: MIT License

R 100.00%

stm sdm celular-automata forest forest-management transition-model

stmanaged's Introduction

STManaged: State and transition model for the eastern North American forest

The {STManaged} R package runs the State and transition model for the eastern North American forest, with integrated forest management practices. This package allows you to model the spatially explicit dynamics of four forest states (Boreal, Temperate, Mixed and Regeneration) over space and time. You will be able to set the intensity of four management practices (plantation, harvest, thinning and enrichment) that aim to increase the northward range shift of forest.

Installation

Install the STManaged package with the devtools (or remotes) package:

devtools::install_github("willvieira/STManaged")

On Linux you need to install the ImageMagick++ library (see here), or you can install a version of this package without the animation() function.

Quick start

library(STManaged)

# Create the initial landscape defining the range of annual mean temperature and the cell size:
initLand <- create_virtual_landscape(climRange = c(-2.61, 5.07), cellSize = 2)

# Print the initial landscape
plot_landscape(initLand, Title = 'initial_landscape')

# Run the model for 100 years with temperature increase of 1.8 degrees
lands <- run_model(steps = 20,
                   initLand,
                   managInt = c(0, 0, 0, 0),
                   RCP = 4.5)

# Some functions are already built in to check the model output
## Forest state occupancy for first and last year
par(mfrow = c(2, 1))
plot_occupancy(lands, step = 0, spar = 0.4)
plot_occupancy(lands, step = 20, spar = 0.4)

## Range limit shift of Boreal and Temperate states over time
plot_rangeShift(lands, rangeLimitOccup = 0.7)

## animated gif of the dynamics
animate(lands, fps = 5, gifName = 'RCP4.5', rangeLimitOccup = 0.7)

Further description

For a detailed vignette about the functionatilies of this package, look here.

stmanaged's People

Contributors

Stargazers

Watchers

Forkers

yangxhcaf

stmanaged's Issues

Change temperature (to unscaled) and steps (to years) arguments

Sensitivity to management intensity

A first exploration on the effect of management intensity (managInt argument of run_model() function). This argument defines the management practice and the intensity of the practice in percentage per year. I.e. managInt = c(0.1, 0, 0, 0) means that we will apply 10% of plantation of temperate species on regeneration stands present at the boreal region every time step.

Variables:

management practices: notManaged, plantation, harvest, thinning, enrichment
management intensity: 0.05, 0.1, 0.2, 0.4, 0.8
RCP = 4.5
cell size = 0.5
replications = 15

Sensitivity to range limit occupancy

A quick test to define the best value of occup of the range_limit() function. This argument determines the threshold to define the range limit of a forest state. It determines the minimum occupancy a row of the landscape must be occupied by a specific forest state to be considered part of the state distribution.

Variables:
Occupancy: 0.7, 0.75, 0.8, 0.85, 0.9, 0.95 %
Cell size: 0.5, 0.8, 1, 2.5 Km

We will have 6 x 4 = 24 simulations x 15 replications with a different initial_landscape.

Error when running the model in parallel

The problem is related with object parCell that is not found.

Sensitivity to cell size

Here I test how different cell sizes can impact the migration rate of forest states. And how the variation in cell size can interact with forest management practices.

Variables:

Cell size: 0.1, 0.3, 0.5, 0.8, 1, 2.5, 5 km²
Forest management: notManaged, plantation, harvest, thinning, enrichment
Management intensity: 0.15
RCP: 4.5
Climate increase: linear growth
Repetition: 30

Interacting cell size and forest management, we have a total of 7 x 5 = 35 simulations. For each simulation, we will have 30 repetition with a different initial_landscape.

Here is one of the 30 initial landscapes for each cell size:

Sensitivity to RCP scenario

A sensitivity analysis to explore the effect of RCP scenarios. We are using four climate change scenarios according to IPCC and we will interact these RCP scenarios with four cell sizes of the landscape.

Variables:
cell size: 0.3, 0.5, 0.8, 1
RCP: 2.6, 4.5, 6, 8.5
management: 0
replications = 15