Perpendicular distances for sightings made while circling

A group spotted from the census track is often circled for photographs, identification, and a proper count, and the sighting is logged during that circle rather than at the moment of detection. This ties each such sighting back to the point on the track-line where the aircraft broke off, and measures the distance from there.

Usage

circling_distance(
  dat,
  by = NULL,
  units = c("m", "km", "nmi"),
  position = c("exact", "logged")
)

Arguments

dat: A data frame of NARWC survey data in survey order, with LATITUDE, LONGITUDE, LEGTYPE, and ideally CIRCLE from flag_circling() and S_LAT/S_LONG. Without CIRCLE, LEGTYPE == 4 is used.
by: Grouping columns identifying one occupation of a survey line, as in track_bearing().
units: "m" (default), "km", or "nmi".
position: Which position to treat as the animal's. "exact" (default) uses S_LAT/S_LONG only. "logged" falls back to the record's own LATITUDE/LONGITUDE where no exact position exists — but that is the aircraft orbiting the animal, offset by the radius of the circle, which at a few hundred metres is the same size as the distances being measured. Use it knowing that, and read position_source to see which applied.

Value

A tibble with one row per row of dat, populated only for circling sightings:

distance: Perpendicular distance from the census line.
along: Signed along-track distance from the break-off point to where the animal was abeam; negative means back down the line.
side: "left", "right", "on-track", or NA.
radial: Straight-line distance from the break-off point.
bearing: Inbound track bearing at the anchor.
anchor_event: EVENTNO of the anchor record, or its row number when EVENTNO is absent.
position_source: "exact" or "logged".

Why these detections matter

A circling sighting is usually a genuine on-effort detection: the animal was seen from the track-line, which is why the aircraft left it. Giving it no distance drops it from the detection function altogether, and the dropped detections are not a random sample — they are disproportionately the close, conspicuous, or high-priority groups that were worth breaking off for. Losing those thins the near-zero end of the distance distribution, exactly where a detection function is most sensitive.

The anchor

The reference point is the last census record before the circling began — the LEGSTAGE == 3 break-off record where one exists (handbook 8.A.20), and otherwise the last record still on the line. anchor_event reports which record was used so the choice can be checked.

The bearing is the inbound heading: from the previous distinct census position to the anchor, which is the direction the aircraft was flying when the animal was detected. It is deliberately not the centred difference track_bearing() uses, because the record after a break-off is the resume record, and a resume point offset from the break-off would swing the bearing by an amount that has nothing to do with the track being flown.

Break-off point, or the line through it

The aircraft usually flies past a group before turning, so the break-off point is beyond the animal rather than abeam of it. radial is the straight-line distance from the break-off point to the animal; distance is that position projected perpendicularly onto the census line, and along is how far back along the line the animal was abeam. distance is the one a detection function needs — radial will generally be larger, by exactly the margin along records.

What this cannot fix

The position is recorded some minutes after detection, so the animal has moved, and the distance is an estimate of where it was when detected rather than a measurement of it. That error is not quantified here. Treat circling distances as a distinct and less reliable source — which is why distance_source marks them, and why they should be excluded from a detection function unless their inclusion is a deliberate, stated choice.

References

Kenney, R.D. (2023) The North Atlantic Right Whale Consortium Database: A Guide for Users and Contributors, Version 8, sections 4.2 (event 12), 8.A.20, and 8.A.21. NARWC Reference Document 2023-01.

Examples

path <- system.file("extdata", "narwc-example.csv", package = "distsamp")
dat <- flag_circling(make_leg_id(read_narwc(path)))
#> `read_narwc()` renamed 2 columns:
#>   LAT_DD  -> LATITUDE
#>   LONG_DD -> LONGITUDE
#> All matched an exact entry in the alias table; `narwc_column_mapping()` returns this, and `quiet = TRUE` silences it.
d <- circling_distance(dat)
d[!is.na(d$distance), ]
#> # A tibble: 1 × 7
#>   distance along side  radial bearing anchor_event position_source
#>      <dbl> <dbl> <chr>  <dbl>   <dbl>        <dbl> <chr>          
#> 1     250. -400. right   472.       0           39 exact