Skip to contents

Add simple swept-area estimates to a datras_raw object

Source: R/swept_area.R
add_swept_area_simple.Rd

Compute haul-level swept-area estimates from towing distance and trawl width using a simple rule-based approach.

Usage

add_swept_area_simple(
  x,
  min_speed = 1,
  min_dist = 0,
  max_dist_dev = 0.2,
  impute_missing = FALSE,
  width = "WingSpread",
  verbose = TRUE,
  full_report = FALSE
)

Arguments

x

A datras_raw object.

min_speed

Minimum plausible ground speed. Values below this threshold are treated as missing. Defaults to 1.

min_dist

Minimum plausible towing distance. Values below this threshold are treated as missing. Defaults to 0.

max_dist_dev

Maximum relative deviation allowed between recorded distance and distance reconstructed from HaulDur and GroundSpeed. Distances with larger deviations are treated as missing. Defaults to 0.2.

impute_missing

Logical. Currently included for compatibility, but not used explicitly in the present implementation.

width

Character string specifying which spread column to use as trawl width. Must be either "WingSpread" (default) or "DoorSpread". Use "DoorSpread" for surveys that record door spread but not wing spread. Note that the fixed beam-trawl widths and the gear-specific plausibility bounds are only applied when width = "WingSpread".

verbose

Logical. If TRUE (default), print a summary of swept-area missingness and imputation by survey and gear.

full_report

Logical. If TRUE, the summary also includes per-column missingness counts (na_DoorSpread, na_WingSpread, na_Distance, na_HaulDur, na_GroundSpeed) for the original input columns. Defaults to FALSE.

Value

A datras_raw object with the following fields added to HH:

  • SweptArea: swept area in \(km^2\) based on haul-specific spread,

  • SweptArea_median: swept area in \(km^2\) based on gear-specific median spread,

  • SweptArea_imputed: logical flag, TRUE when the spread and/or towing distance used for that haul was imputed.

A survey-by-gear summary is attached as attr(x, "swept_area_summary"), with columns survey, gear, n_records, n_imputed, prop_imputed, and n_NA / prop_NA (hauls whose SweptArea is still NA after imputation, e.g. gears that cannot be imputed). When full_report = TRUE, per-column missingness counts (na_DoorSpread, na_WingSpread, na_Distance, na_HaulDur, na_GroundSpeed) of the original input columns are appended.

Details

The function uses recorded WingSpread where available, applies a series of basic plausibility filters and gear-specific corrections, imputes missing values from gear-level medians, reconstructs towing distance when needed from HaulDur and GroundSpeed, and returns two swept-area estimates.

Swept area is calculated in \(km^2\) as: $$ SweptArea = Spread \times Distance \times 10^{-6} $$ where \(Spread\) is taken from the column selected by width (in metres) and \(Distance\) is in metres.

A second estimate, SweptArea_median, is also returned. This uses a gear-specific median spread rather than the haul-specific recorded or imputed spread: $$ SweptArea\_median = Spread\_median \times Distance \times 10^{-6} $$

The function applies the following steps:

  • removes implausible values of GroundSpeed, Distance, and the selected spread column,

  • removes distance values that differ too much from reconstructed distance based on speed and haul duration,

  • when width = "WingSpread": applies fixed widths for selected beam trawl gear codes and accounts for double-beam trawls via GearEx == "DB",

  • imputes missing spread values using gear-specific medians,

  • imputes missing ground speed using gear-specific medians,

  • reconstructs missing towing distance from haul duration and ground speed.

See also

add_swept_area(), add_swept_area_fishglob()