projectTracks: Project tracking data

Description

projectTracks is a convenience function to project tracking data to a an equal-area projection for use in kernel density analysis.

Usage

projectTracks(dataGroup, projType, custom)

Value

Returns a SpatialPointsDataFrame, which can be used for the following functions: findScale, estSpaceUse,

indEffectTest, repAssess

Arguments

dataGroup

data.frame or SpatialPointsDataFrame. Tracking data, with fields as named by formatFields. Must contain 'Latitude' and 'Longitude' columns.

projType

character. Select type of equal-area projection to use. Two options are are available: 'cylin' projects to a World Cylindrical Equal Area projection, and 'azim' projects to a Lambert Azimuthal EA.

custom

logical (TRUE/FALSE). Choose whether projection will use default centering parameters or whether to set projection center on centroid of latitude and longitude in dataGroup.

Input data can be tracks split into trips (i.e. output of tripSplit)

Details

Data are transformed to either a World Cylindrical Equal Area, or a Lambert equal-area projection. Cylindrical projections generally appear better for data that are distributed more along one axis, while azimuthal appear better for data that is distributed evenly within a radius. The most important thing is that the data are in an equal-area projection for Kernel Density Analysis (e.g. estSpaceUse).

If custom=TRUE, the projection will be centered on the data. This is particularly preferable for data that cross the international dateline, or near the poles. However, it is important to recognize that this projection is specific to inpute dataset (i.e. dataGroup) so if projectTracks is run again with even slightly different data, the projections will differ, which may cause issues down the line if merging spatial datasets again.

NOTE that these projections may not be the most appropriate for your data and it is almost certainly better to manually identify a projection appropriate for your study region. Custom projections are centered on the centroid of the tracking locations, which is biased for locations close to the poles. In this case it would be better identify an appropriate polar projection for your study are instead of relying on projectTracks. So it is not strictly necessary for projectTracks to be used in track2KBA analysis, what is important is that an equal-area projection of some kind is used when constructing utilization distributions.

Examples

Run this code


dataGroup <- data.frame(Longitude = c(1, 2, 3, 1), 
                 Latitude =  c(1, 1, 2, 1),
                 ID = rep("A", 4),
                 DateTime = format(
                   lubridate::ymd_hms("2021-01-01 00:00:00") + 
                   lubridate::hours(0:3)
                    )
                 )

## project tracks 
tracks_prj <- projectTracks(
dataGroup, 
projType = "azim", 
custom = "TRUE"
)