Learning Objectives
By the end of this knowledge development session, student divers will be able to:
- State the purpose of the Station Report and explain why its format is standardized rather than left to the author.
- List the required sections of a Station Report.
- Explain why every reported value carries an uncertainty statement, and state what is done when a value falls below it.
- Distinguish the measured, derived, and interpreted parts of the report and explain why they are kept separate.
- Explain why the metadata — who, where, when, with what — is part of the measurement rather than an administrative heading.
- State how many figures to report for an absolute value and for a difference.
- State what should be recorded when something goes wrong on a dive.
Presentation Notes
a. The Station Report is the product of this course. Not the dives, and not the numbers. A student who has occupied a station, computed coefficients, and drawn a conclusion has done the work; a student who has written it down in a form another person can read, check and compare has finished it. Certification depends on the report.
b. Its format is standardized, and that is the point. A report written in whatever form its author preferred can be read but not compared. The value of many stations is that they can be set beside one another — this reef against that reef, this morning against this afternoon, this year against next. That only works if every report contains the same fields, computed the same way, with uncertainty expressed the same way. Standardization is what turns a collection of reports into a dataset.
c. The report has seven sections, holding five kinds of information. The distinction matters, because the sections are a filing system and the five kinds are what must never be mixed up.
The five kinds: metadata, which says who measured what, where, when and with which instrument; calibration, which says how well; measurements, which are things a sensor or an eye actually registered; derived quantities, which are things computed from measurements; and interpretation, which is what the author thinks it means.
The seven sections of the form: Metadata, Calibration, Measurements, Derived Quantities, Horizontal Survey, Interpretation, and What Went Wrong. The survey section holds measurements and the last holds a plain account of failures — both are separated out because they are filled in once for the course rather than once per occupation.
d. Metadata is part of the measurement. A temperature without a depth, a time, a position and an instrument identity is not a measurement — it is a number. The position is recorded to the precision available, referenced to the same point used for the Reef Cartographer survey of the site so that the specialties overlay. The time is recorded in a stated time zone, because a diurnal comparison is meaningless without one. The instrument serial numbers matter, because an offset belongs to a particular logger and not to loggers in general.
e. Calibration is reported, not merely performed. The report states the cross-calibration offsets before the dive, the offsets after it, and the residual spread that remained once the offsets were removed. That spread gives u(T) for the temperature array, and it was measured on that day with that equipment, in shaded seawater from the site. The same applies to the optical work, and it produces two different spreads that must not be confused. The difference between the two Teams' independent extinction distances is the between-observer spread; half of it is the observer term in u(y). The inward cycles each observer took to close from vanish to reappearance is the within-observer bracket, and that is metadata about how sharp the judgement was. Both are recorded, and they are not interchangeable.
f. Every reported value carries an uncertainty statement. A sensed quantity carries its measured uncertainty or method limit; a derived quantity carries a propagated uncertainty; and every comparison names the threshold that governs it. Not a manufacturer's specification — the number the team measured. A reader who sees a temperature difference of 0.3 °C needs to know whether this array could resolve 0.04 °C or 0.4 °C, and those two reports say entirely different things while looking identical. A value without its appropriate uncertainty statement is incomplete and the report is not finished.
g. Write down what the instrument showed, then round when you report. The raw record holds the full displayed precision, because rounding at the moment of transcription is irreversible and destroys the small differences the course exists to detect. The reported value is rounded to the uncertainty of the quantity being reported — and as Topic 3 established, that is not the same for every field. An absolute temperature is limited by instrument accuracy and is reported to whole degrees or one decimal. A difference between two readings from a calibrated set is limited by the applicable threshold and may carry two. A difference between uncalibrated instruments is not reportable at all. Precision follows the quantity, not the instrument.
h. Below the limit, say so plainly. A difference smaller than the applicable threshold is recorded as unresolved. A quantity the instrument cannot see is recorded as below detection limit, with the limit stated. A distance beyond which the Teams did not swim is recorded as greater than that distance. None of these are failures, and none of them should be converted into a number to make the report look complete. A report full of honest non-results is better science than one full of false precision.
i. Do not summarise a profile with a single number. As Topic 4 established, an average is only meaningful when the values are repeated samples of one quantity. Readings from different depths in a stratified column are measurements of different quantities, and their mean describes water that exists at no depth. Where the column tests as optically and thermally uniform, a station mean is defensible and should be reported with its spread. Where it does not, report the profile and no summary statistic.
j. Measured and derived are kept apart, and labelled. The measured section contains depth, temperature, illuminance, extinction distance and surge period — things a sensor or an eye actually registered. The derived section contains the attenuation coefficients, the thermal gradient, the optical product, and anything else arrived at by arithmetic. The attenuation coefficient from these loggers is labelled Kd(photopic), not Kd, because the sensor weights light the way an eye does and not the way a plant does. Labelling derived quantities as derived is not modesty; it tells the reader how many independent observations stand behind the report.
k. Interpretation comes last and stays in its own section. The prediction, recorded and dated before the dive that tested it. What the comparison showed. Which differences exceeded their applicable threshold and which did not. Which resolved differences could be attributed to a mechanism and which could not. And what the author thinks it means, clearly marked as what the author thinks rather than what the instruments said.
l. Record what went wrong. A logger that flooded, a disc measurement abandoned because a boat passed through the sightline, a diver who lost depth in surge, a cloud bank that arrived halfway through the light record — all of it goes in. A report that describes only the parts that worked is not a record of what happened. It also robs the next team of the most useful thing you learned, which is how the method fails.
m. Write for someone who was not there. The test of a Station Report is whether a diver at another reef, next year, can read it and know exactly what was done, how well, and with what confidence — and then go and do the same thing well enough that the two reports can be compared. Everything in the required format exists to serve that test.
n. The report is where the course's argument comes to rest. Every topic in this section has said a version of the same thing: a measurement without a stated uncertainty is not a measurement, and the discipline of saying how well you know something is what separates data from numbers. A digital display will hand you more digits than the sensor can justify; the Station Report is where you decide which of them you are entitled to keep. It is the reason a diver's inexpensive instruments can produce something a scientist is able to use.
Appendix reference. The required Station Report format, field by field, is given in the Station Report Recording Forms. Worked examples of the calculations that populate it are in the Technical Supplement, Appendix B.