Build a concave-hull prediction region from survey positions

Creates an sf polygon describing the spatial footprint of a set of sampling positions, using a concave hull (concaveman::concaveman()). This is used to restrict a prediction grid to the area a survey actually covers, rather than the full model domain.

Usage

make_survey_prediction_region(
  data,
  crs,
  xy_cols = c("X", "Y"),
  coord_multiplier = 1000,
  concavity = 5,
  buffer = 1
)

Arguments

data

A data frame with coordinate columns (see xy_cols) in the same units as the model (typically km).

crs

Coordinate reference system of the projected coordinates.

xy_cols

Character vector of length two naming the coordinate columns. Default c("X", "Y").

coord_multiplier

Numeric factor to convert xy_cols to the CRS units (e.g. 1000 to convert km to m). Default 1000.

concavity

Numeric. Concavity parameter passed to concaveman::concaveman(). Larger values give looser (more convex) hulls. Default 5.

buffer

Numeric. Buffer (in CRS units) applied to the hull to avoid clipping the outermost stations. Default 1.

Value

An sf polygon (single feature) in the given crs.

See also

make_prediction_grid(), compare_surveys()

Author

Mikko Vihtakari

Examples

# \donttest{
data(sebastes)
eggan <- subset(sebastes, cruiseseries == "EggaN")
region <- make_survey_prediction_region(eggan, crs = 32633, concavity = 5)
ggOceanMaps::qmap(region, fill = I(NA))

# }