README.md

In maudqueroue/intercali: Intercalibration Des Différents Types De Suivis Aeriens

Intercalibration Des Différents Types De Suivis Aeriens

This package contains many functions to simulate aerial monitoring data. From these data simulations, it is possible to obtain abundance and distribution estimates of species

Installation

You can install the development version of intercali from GitHub with:

# install.packages("devtools")
devtools::install_github("maudqueroue/intercali")

Example

Extract map with desired densities

This function allows, from a density map density_obj (density object of the dsims packages), to provide a map of class sf (data.frame). It allows to recalculate the correct density ratios of the provided map according to the desired number of individuals in the area N.

An example of this function use a dataset of density dataset_density created thanks to the package dsims. From this density object, the aim is to create a map with the correct densities. Here a density corresponding to 500 individuals in the studied area. Then the function plot_map allows to plot the map.

# Load the package 
library(intercali)

# Use dataset from the package
data(dataset_density)

# extract map
map <- extract_map(density_obj = dataset_density,
                   N = 500,
                   crs = 2154)

# Plot map
plot_map(map_obj = map)

Simulate individuals with a inhomogenous Poisson point process

From an sf class density map map_obj (data.frame), an inohomogene Poisson point process is used to simulate the presence of individuals in the studied area. The probability of presence of an individual is dependent on the density given by the map. The function return a dataframe, here ind, containing the differents individuals simulated and their geographic coordinates. Then the function plot_obs allows to plot the simulated individuals ind on the map with density information.

# Simulate individuals
ind <- simulate_ind(map_obj = dataset_map,
                    crs = 2154)
#> Registered S3 method overwritten by 'spatstat.geom':
#>   method     from
#>   print.boxx cli

# Plot indviduals
plot_obs(obs_obj = ind,
         map_obj = map)

Create transect design

The create_transect function allows to create transects in the same way that the make.design from the dssd package. The create_transect function allows, among other things, to: * choose different types of survey design such as zigzag or parallel transects * choose the desired transect length (approximately) * choose the angle of the transects * choose the spacing between transects

Here we use as the studied area, a shapefile coming from the dssd package and use the make.region of the dsims package. Then the function plot_transect allows to plot the created transects transect_obj on the study area map_obj. From this region object, zigzag transects (design = ezigzag) with a approximative length of 400000m line_length are created in the study area.

library(dssd)
library(dsims)

# Use of the St Andrews bay map from the dssd package
shapefile.name <- system.file("extdata", "StAndrew.shp", package = "dssd")

# Creation of the object with the make.region function of the dsims package
region <- make.region(region.name = "St Andrews bay",
                      shape = shapefile.name,
                      units = "m")

# Creation of the transects
transects <- create_transect(region_obj = region,
                             crs = 2154,
                             design = "eszigzag",
                             line.length = 400000,
                             design.angle = 30,
                             truncation = 400)

# Plot transects
plot_transects(transect_obj = transects, 
               map_obj = region, 
               crs = 2154)

The grid created with extract_map does not perfectly match the total area of the region because it only keeps squares of the same size, so the edges of the study region are slightly cropped. With the create_transect function, the transects are created on the basis of the total region region_obj and not on the basis of the grid map_obj. Hence there is a need to resize the transects with the fonction crop_transect in order to perfectly fit the density map use to simulate individuals map_obj.

# Crop transects
cropped_transects <- crop_transect(transect_obj = transects,
                                   map_obj = map)

To do distance sampling analyses, it is important to segmentize the transects in order to (potentially) use covariates on the detection fonction. Then the function segmentize_transect allows to segment the transects. From the created transect object transect_obj and by choosing the desired length in m length_m for the segments, the function returns a new table with the segments cut and named according to the transect and the segment, for example, Sample.Label: 1-2 indicating the segment 2 of transect 1. The function is not mine it was found here.

Thanks to the ifsegs argument, the function plot_transect allows to highlight with different colors the different segments in transects.

# Segmentize transects
segs <- segmentize_transect(transect_obj = cropped_transects,
                            length_m = 2000,
                            to = "LINESTRING")

# Vizualize segments
plot_transects(transect_obj = segs, 
               map_obj = map, 
               crs = 2154,
               ifsegs = TRUE)

The get_monitored_area function is used to calculate the area of the study area that is monitored by the transects. The maximum distance monitored truncation_m on the transects is used to calculate the area of the study area map_obj truly followed by the transects transect_obj. In this example, the maximum distance at which an individual can be observed truncation_m is chosen at 400m.

get_monitored_area(transect_obj = segs,
                   map_obj = map,
                   truncation_m = 400)
#> 260672990 [m^2]

Obtain individuals detected according to the transect design and a detection function

To calculate the probability that an individual could be observe thanks to the transects simulated, the first thing to do is to calculate distance between transects/segments and the observation of individuals. The function calculate_distance calculates the nearest transect/segment transect_obj for each simulated individual obs_obj. It returns an array with the name of the closest transect/segment for each individual and the distance in m and in km between them.

dist <- calculate_distance(obs_obj = ind, 
                           transect_obj = segs, 
                           crs = 2154)

Then, the detection_function is use to simulate the probability that a individual could be observed according to the sample design.

• The detection function could be an uniform detection function with a probability of detection g_zero on the whole strip band until the maximum distance of observation (in m) truncation_m.

• The detection function could also be a half normal detection function for which we can choose the effective strip width (in km) esw_km i.e. the distance at which there are as much non detected individuals before this distance than detected individuals after this distance. For the half normal detection function it is also possible to choose the proability of detection at 0 meter g_zero and the maximum distance of observation (in m) truncation_m.

library(ggplot2)
library(cowplot)

detected <- detection(dist_obj = dataset_dist,
                   key = "unif",
                   g_zero = 0.8,
                   truncation_m = 250) 

a <- ggplot(detected, aes(x=distance_m, y=proba)) +
  geom_point(color = "indianred4") +
  xlim(0,500)
  #title("uniform detection")

detected <- detection(dist_obj = dataset_dist,
                   key = "hn",
                   esw_km = 0.16,
                   g_zero = 1,
                   truncation_m = 400)

b <- ggplot(detected, aes(x=distance_m, y=proba)) +
  geom_point(color = "indianred4") +
  xlim(0,500)
  #title("Half normal detection")

plot_grid(a,b)

maudqueroue/intercali documentation built on Oct. 8, 2022, 2:09 p.m.