This document lists data processing steps for the four methods used to
process GAP CTD data (Table 1). The four methods (Typical, Typical CTM,
Temperature-Salinity Area, and Minimal Salinity Gradient) use different
parameters for temperature alignment (Step #7) and conductivity cell
thermal mass correction (Step #8), which are described in Table 2 and
Table 3. Steps 2-13 are performed inside of the run_gapctd() function.
Table 1. Data processing steps for Typical, Typical CTM, Temperature-Salinity Area (TSA), and Minimal Salinity Gradient (MSG) methods for processing data using the gapctd package.
| Step | Description |
|---|---|
| 1 | Convert CTD data from .hex to .cnv, extracting Pressure, Conductivity, Time Elapsed, and Data Flags. This is performed using the gapctd::setup_gapctd_directory() function, which uses system commands to run SBE Data Processing with .hex files and .xmlcon files from G:/RACE_CTD/. The batch script (.bat) and program setup file (.psa) needed to execute the system command are in the /inst/extdata/gapctd/ directory of the gapctd repo (i.e., /extdata/gapctd/ of the package directory when the package is installed). |
| 2 | Load the converted .cnv files into a ctd object using oce::read.oce(). |
| 3 | Assign haul metadata to the ctd object using gapctd::assign_metadata(x = [ctd]). |
| 4 | Split data in ctd object into downcast, bottom, and upcast sections based on CTD data time stamps and haul metadata using gapctd::section_oce(x = [ctd], by = 'datetime') |
| 5 | Median filter temperature and conductivity channels with five (5) scan windows for each channel using gapctd::median_filter(x = [ctd], variables = c("temperature", "conductivity"), window = c(5,5)) |
| 6 | Low pass filter temperature, conductivity and pressure using 0.5, 0.5, and 1.0 second time constants, respectively, using gapctd::lowpass_filter(x = [ctd], variables = c("temperature", "conductivity", "pressure"), time_constant = c(0.5, 0.5, 1), precision = c(4, 6, 3), freq_n = 0.25) |
| 7 | Align temperature. Apply optimal alignment from step #5 to temperature channel using gapctd::align_var(x = [ctd], variables = "temperature", interp_method = "linear", offset = [offset]) |
| 8 | Conductivity cell thermal mass correction (Cell TM). Apply optimal conductivity cell thermal mass correction to upcast and downcast data using gapctd::conductivity_correction(x = [ctd], alpha_C = [alpha], beta_C = [beta]). Parameters vary depending on the method (Typical, Typical CTM, TSA, MSG) |
| 9 | Slowdown. Flag data where the mean ascent/descent speed of the CTD for a five scan window around an observation was < 0.1 meters per second, except near-bottom. Uses: gapctd::slowdown(x = [ctd], min_speed = 0.1, window = 5) |
| 10 | Derive depth, salinity, density, absolute salinity, sound speed, and squared buoyancy frequency for downcast, bottom, and upcast data, using gapctd::derive_eos(x = [ctd]) |
| 11 | Bin average. Calculate averages by 1-m depth bin for all variable channels (excluding scans flagged ‘bad’). gapctd:::bin_average(downcast, by = “depth”, bin_width = 1). Uses gapctd::bin_average(x = [ctd], by = "depth", bin_width = 1) |
| 12 | Density inversion check. Check for unreasonable density inversions based on the squared buoyancy frequency, N2, where inversions are considered unrealistic when N2 < -1e-4 using gapctd::check_density_inversion(threshold = -1e-4, threshold_method = "bv", correct_inversion = TRUE) |
| 13 | Completeness check. Check for casts with an unreasonable amount of missing data using gapctd::qc_check(prop_max_flag = 0.1, prop_min_bin = 0.9). |
| 14 | Visual error inspection. Visually inspect downcast and upcast profile data for errors. Remove and interpolate the errors. Uses gapctd:::wrapper_flag_interpolate() as a wrapper for gapctd::qc_flag_interpolate(x = [ctd], review = c("density", "salinity")) |
| 15 | Review profiles. Review downcast and upcast profile data for quality and select profiles to include in the final data product. Uses gapctd::review_profiles() |
Table 2. Alignment offsets parameters by data processing method.
| Method | Channel | Offset (s) | Description |
|---|---|---|---|
| Typical | Temperature | -0.5 | Typical setting recommended by the manufacturer. |
| Typical CTM | Temperature | Estimated | Use grid search to find the offset parameter that maximizes the coefficient of determination (r2) between first derivatives of temperature and conductivity with respect to time for individual upcasts and downcasts, to the nearest 0.01 s (Ullman and Hebert, 2014). |
| TSA | Temperature | Estimated | Same as Typical CTM |
| MSG | Temperature | Estimated | Same as Typical CTM |
Table 3. Conductivity cell thermal mass correction parameters by data processing method.
| Method | α | β | Description |
|---|---|---|---|
| Typical | 0.04 | 1/8 | Typical setting recommended by the manufacturer. |
| Typical CTM | 0.04 | 1/8 | Typical setting recommended by the manufacturer. |
| TSA | Estimated | Estimated | Estimate optimal conductivity cell thermal mass correction parameters (alpha and beta) by using numerical optimization (Broyden-Fletcher-Goldfarb-Shanno with box constraints) to find parameter values that minimize the area between temperature-salinity curves from upcasts and downcasts (Garau et al., 2011). Starting parameters based on initial grid search across values of α (0.001, 0.01, 0.02, 0.04, 0.08, 0.12), β (1, 1/2, 1/4, 1/8, 1/12, 1/24). Box constraints: 0 ≤ α ≤ 1, 0.1 ≤ β-1 ≤ 45. Typical values (alpha = 0.04, beta = 0.125) are used if optimization does not converge. |
| MSG | Estimated | Estimated | Same as TSA but optimization minimizes the path distance of the salinity curve for individual casts instead of the area between downcast and upcast T-S curves. |
Garau, B., Ruiz, S., Zhang, W. G., Pascual, A., Heslop, E., Kerfoot, J., & Tintoré, J. (2011). Thermal lag correction on slocum CTD glider data. Journal of Atmospheric and Oceanic Technology, 28(9), 1065–1071. https://doi.org/10.1175/JTECH-D-10-05030.1
Ullman, D. S., & Hebert, D. (2014). Processing of underway CTD data. Journal of Atmospheric and Oceanic Technology, 31(4), 984–998. https://doi.org/10.1175/JTECH-D-13-00200.1