Depth-integrate one CTD cast's profile

The provider's definition for the CTD (Rasmus Swalethorp, 2026-09-23, questions.csv Q02, adopted 2026-10-01): "For the CTD data here we should just sum up all the 1m bins within the top 200m or as deep as the station was on shallower stations. For the bottle data we will need to integrate between points. I believe trapezoidal integration was used in the past" (CalCOFI/workflows#102). So method = "sum" (the default) sums the bin_m bins from the surface to z_max (default 200 m), or to the deepest good sample if shallower, which is recorded; method = "trapezoid" is the trapezoidal integral between points, for bottle data (and the CTD rule of calcofi4db 4.16.0-4.17.x). For chlorophyll-a in mg m^-3 the result is mg m^-2.

Usage

ctd_integrate(
  depth,
  value,
  qual = NA,
  z_max = 200,
  z_top_max = 5,
  method = c("sum", "trapezoid"),
  bin_m = 1
)

Arguments

depth

depth (m), one cast.

value

the quantity per unit volume.

qual

quality codes of value (NA = good).

z_max

integration floor (m), default 200.

z_top_max

deepest acceptable first sample (m), default 5.

method

"sum" (the provider's CTD rule: the sum of the bin_m bins) or "trapezoid" (between points, for bottle data).

bin_m

the bin thickness "sum" adds up (m), default 1.

Value

one-row tibble: integrated, depth_reached_m, status (ok, shallow, no_surface, no_data), z_max, method.

Details

The sum runs over the bins centred at bin_m, 2 * bin_m, ..., bottom (200 bins of 1 m in the top 200 m). The answer is silent on missing bins, so the 4.16.0 behaviour is kept for both methods: the shallowest good sample is carried up to the surface when it is no deeper than z_top_max (default 5 m), a cast that starts deeper is NA (status no_surface, so a missing surface is never silently skipped), and a bin missing inside the profile (a flagged or NaN bin) is filled by linear interpolation between its neighbours rather than counted as zero. Flagged (8/9), missing and non-finite samples are dropped first.

Examples

ctd_integrate(0:300, rep(1, 301))                       # 200 over 0-200 m
#> # A tibble: 1 × 5
#>   integrated depth_reached_m status z_max method
#>        <dbl>           <dbl> <chr>  <dbl> <chr> 
#> 1        200             200 ok       200 sum   
ctd_integrate(c(0, 100, 300), c(1, 1, 3), method = "trapezoid")
#> # A tibble: 1 × 5
#>   integrated depth_reached_m status z_max method   
#>        <dbl>           <dbl> <chr>  <dbl> <chr>    
#> 1        250             200 ok       200 trapezoid