Report output
All generated HTML files are fully self-contained: figures are embedded as base64 PNG images, JavaScript libraries are bundled inline, and no external requests are made at view time. The entire output directory can be copied to a USB drive or a vessel intranet server and used offline.
Directory layout
<output.dir>/
index.html summary: map and cruise-wide overview panels
casts.html map, sortable table of all casts, acquisition clock
sections.html section cards
timeseries.html time-series group cards
leaflet.html interactive map
sbe_sensors.html sensors by cast and device, and the sensor variables
processing.html every processing step applied, by run of casts
<stem>_inventory.html variables and attributes of each compiled file
(profiles_inventory.html, ladcp_profiles_inventory.html)
corrections_ts.html alignment and cell-thermal-mass fits, when ctdcast fit
has written fits.nc
casts/
cast_001.html
cast_002.html
...
sections/
section_KTout.html
section_FARDWO.html
...
timeseries/
timeseries_<name>.html
...
index.html
The summary page. Its masthead gives the cruise, ship, departure and arrival, duration, number of casts and sections, and maximum depth, and links every other page. Below it:
Map — the cruise track and station positions (a static map; the interactive one is
leaflet.html).Hydrography — cruise-wide overview panels of conservative temperature, absolute salinity and potential density across all downcasts.
Biogeochemistry — the same for oxygen, fluorescence and turbidity.
Velocity — eastward and northward LADCP velocity, when LADCP data are present.
T–S diagram — a CT–SA histogram of all downcast profiles.
casts.html
An interactive Leaflet map (bundled offline) above a sortable table. The map shows one marker per cast and a line per section; hovering a table row highlights its marker and vice versa, and clicking a marker or row opens the station page. A static map prints and shows when JavaScript is off.
The table lists every cast: Cast · File · Start (UTC) · Latitude · Longitude · Depth · QC
flagged · LADCP · Links (QC flagged is the percentage of samples flagged suspect or fail;
LADCP appears when LADCP is configured). An Acquisition clock section below it
compares the acquisition PC’s clock with NMEA time cast by cast and suggests a clock
block when they differ.
sections.html
One shared Section tracks map with every section drawn on it, then a card for each named
section in ctd_groupings.yaml with its name, description and casts, and a link to its
page.
Cast pages — casts/cast_NNN.html
One page per cast. The header gives the cast’s position, time and depth, the Raw input
it was read from (MIXSED2_012.hex + MIXSED2_012.XMLCON or mixsed2_012.cnv) and its
Sampling (24 Hz (native, unbinned), or what a CNV’s # interval line says).
Sections, each shown only when it applies:
Section |
Panels |
|---|---|
Overview |
CT · SA · σ₀ against pressure (with the LADCP U and V profiles beside them when LADCP is present), the station map, and the T–S diagram down against up. |
Hydrography |
CT · SA · σ₀ against pressure, downcast in colour, upcast in grey. |
Biogeochemistry |
O₂ saturation, fluorescence and turbidity against pressure. |
T–S diagram |
Coloured by O₂ saturation, downcast only. |
Stability |
N² and Turner angle, downcast only. Not drawn on a native-rate cast (unbinned 24 Hz scans); an amber note says why. |
Velocity (bottom track) |
The LADCP bottom-track velocity, when LADCP data are present. |
Diagnostics |
Pressure against time; primary minus secondary T and S; downcast minus upcast. |
QC flags |
The distribution of each variable’s values (grey all data, colour the data kept, a dashed line at each gross-range threshold), then the thresholds applied and the flag counts per variable. |
Appendices: (A) Sensors, (B) File provenance (cast, data mode, processing stage and
the tracking_id lineage), (C) netCDF data ranges (minimum, maximum and valid count of
every variable in the file), and (D) Processing provenance (below).
Processing provenance
An appendix on each cast page reports what the Sea-Bird deck unit and Sea-Bird Data
Processing did to the cast before ctdcast read it, and, in a second table, Corrections
applied by ctdcast at stage 3 (and in its products), in the order they ran. Both tables
number their steps (#). ctdcast does not re-derive this;
it recovers it from the raw instrument header, which every stage-1 file carries verbatim
in its raw_metadata attribute and passes forward unchanged through stages 2 and 3. A
cast with no Sea-Bird header (a LADCP cast, or a file converted without its header) has no
provenance appendix.
The Corrections applied before ctdcast table lists each step in file order — the deck conductivity advance first, then each Data Processing module. Its columns are:
Step / Producer / Version — the module and who ran it.
Parameters — the salient settings, close to the header’s own wording.
Variables — the variable(s) the step modified, in ctdcast’s canonical names (the rename map is
CNV_ALIASESinconfig/parameters.py); a name it does not cover keeps its header spelling.Matches reference — a conformance check (see below), shown only for the SBE 9 / 11plus family, whose reference values are documented.
A step whose parameters differ per channel or per sensor — the deck advance (per conductivity cell), cell thermal mass (per cell), the low-pass filter (per time-constant group) — is split into one row each, so a check that passes on one channel and fails on another is not hidden behind a single verdict.
Conformance
The Matches reference column compares each step’s parameters against a documented typical value and shows one of three states, never conflated:
✓ — matches the documented value.
✗ — differs from it.
— — no reference exists to check against (an em dash, never a blank and never a cross).
A ✓ means “matches a documented typical value,” not “correct”; a ✗ means “differs,” not “wrong.” Every reference is configuration-dependent (a cell-thermal-mass α assumes a pump and duct; the deck advance assumes standard plumbing), so each cell names the source it compared against (e.g. SBE manual p.92), and the check is a match indicator, not a verdict. Some modules have no documented reference at all — Wild Edit, for instance, has only example dialog values, so it shows — with those examples offered as suggested starting points, not as a deviation.
A separate Conformance note collects the deviations as hedged, source-citing sentences, kept distinct from the structural Advisories note (which reports what the file is — e.g. already pressure-binned — rather than how its parameters compare). The panel keeps its focus on what was done to the data; per-sensor calibration state lives in the Sensors appendix, and the panel carries a one-line pointer to it.
Sensors appendix
A single table joins each science variable to the physical device that produced it, in one
row: Variable · Role · Ch · Device · Model · Cal date · Slope · Offset. It is read from
the per-cast sensor catalog (the SENSOR_* entries and each variable’s sensor link —
see Data files (netCDF)), not re-parsed from the header. A
device with no stored variable still appears, with an empty Variable
cell; on a .hex cast a device ctdcast does not convert (pH, a user polynomial) is joined to
its raw voltage. For a frequency sensor (temperature, conductivity, pressure) a Slope away from 1
or Offset away from 0 is a datcnv drift/span correction already baked in; that row is
tinted amber and the specific value bolded, so a correction is visible at a glance. A cast
file that predates the catalog falls back to the header-parsed device inventory and
calibration-state tables.
processing.html
Every processing step applied to the casts in profiles.nc, read from each profile’s
processing_record (see Data files (netCDF)), one table per stage: stage 1 (reading the raw
file), the Sea-Bird steps recorded in the raw header, stage 2 (soak and deck), stage 3 and the
products. The columns are # · Casts · Step · Producer · Version · Setting · Variables ·
Matches reference · Reference; # numbers the steps in the order they ran, through the whole
page, and every row of a step carries its number, so sorting by # restores the order after
sorting by another column.
Casts that used the same setting of a step share one row, its Casts column a run such as
001–138; a change of setting starts a new row.A step whose setting changed more than five times is drawn as one varies cast to cast row with the range of each changing number, and a figure of the value against cast number when only one number changes.
A row is amber when the setting differs from the value SBE documents for that step, or when the step did not run as its setting says on some casts: Loop Edit without its soak test (a recording that starts below the surface-soak window), or Bin Average keeping the good values of scans with a flagged value, each get an amber row naming those casts. The caption above the tables states the criterion.
A step that runs more than once on a cast (a filter per variable, a correction per sensor) is matched across casts by the variables it changed.
What a step did on one cast (values marked, scans flagged) is on that cast’s page.
The masthead groups sbe_sensors.html and processing.html in a Provenance row: they
describe how the data were taken and processed, not a data file, so they sit apart from the
Data inventory strip.
Section pages — sections/section_NAME.html
One page per named section, against distance along the section:
Section |
Panels |
|---|---|
Map |
The section track and its stations. |
Hydrography |
CT, SA and σ₀ (with the 27.7 and 27.8 kg m⁻³ isopycnals labelled); triangles along the top mark the profiles. |
Biogeochemistry |
Oxygen ( |
Velocity |
LADCP U (east) and V (north), when LADCP data are present. |
T–S diagrams |
The section’s water masses in temperature–salinity space. |
The fields use discrete colorbars with 20 levels (the T–S diagrams 12; limits overridable with display.vmin /
vmax, see config.yaml). Bathymetry is shown as a filled grey polygon when a GEBCO
file is available.
Time-series pages — timeseries.html and timeseries/timeseries_NAME.html
timeseries.html is the index: a Timeseries locations map and a card for each
time-series group in ctd_groupings.yaml (repeat stations, sorted by time). Each group’s
page has:
Section |
Panels |
|---|---|
Map |
The group’s station positions. |
Hydrography |
CT, SA and σ₀ against time and pressure (downcasts). |
Biogeochemistry |
Oxygen ( |
Velocity |
LADCP U and V, when LADCP data are present. |
T–S diagram |
The group’s profiles in temperature–salinity space. |
Time runs on the horizontal axis, pressure on the vertical axis (increasing downward).