The NOAA NCEI Passive Acoustic Data Archive captures duty cycle information for all our data providers, which span the gamut of continuous recordings to having multiple sampling regimens during the same deployment. We'd like to move away from duration and interval values as they can be ambiguous and easy to misinterpret. Instead, we'd like to use percentage of time sound was recorded. The question for the group is what time period should we ask that percentage reflects: hour? day? duration of deployment? Is there an existing standard for this information?
$\begingroup$
$\endgroup$
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
1
I'm not sure if better to add a comment or an answer, but I really appreciate this question and hopeful for the discussion outcome. My thoughts went longer than a comment so here they are...
My initial thought is minimum report percent as duration of deployment. This allows the most flexibility for variable or potentially unpredictable sampling regimens. For example, some PAM gliders sample continuously but then are off for brief surfacings or when the pump runs at the bottom of a dive. Hourly or daily percent could vary, but the full deployment could be averaged out.
If a survey used a very standard approach, then in that case reporting hourly or daily could be done in addition to the total survey duration.
-
$\begingroup$ Thanks! Deployment sounds like a very practical and useful duration to use. $\endgroup$ Commented Oct 19, 2023 at 14:22
$\begingroup$
$\endgroup$
1
Maybe I misunderstand the question, but I would disagree to only report time percentage.
For example, 50% duty cycle, could mean 1 min on - 1 min off for the whole day. If, say, a group of dolphins is passing in the morning, then they would always picked up (at least if they are not swimming too fast or are silent).
On the other side, if the 50% duty cycle mean 6 hours on - 6 hours off, starting at midnight, then the change you are missing a passing group of dolphins is much higher (zero if dolphins pass in the morning)
So, I consider the 'on-time' as important information. In fact it is the effort of the survey, which should not be ignored, and duty cycling of PAM recordings is equivalent of duty cyling of effort.
I understand that the use of duty cycle may be confusing, so I would simply stick with 'on time' and 'off-time', equivalently with 'on-effort' and 'off-effort'.
If the question addresses not the sampling but the occurance of sound (e.g. 10% of my recordings have dolphin clicks) then similar arguments hold. It is important to differentiate between occurance of sound every 10th on-time period, or for 10 consecutive recordings, and the nothing. In the first case there are easily different encounters, while in the second case, most likely only one encounter was recorded.
-
1$\begingroup$ Thanks!. Agreed the details of the duty cycle are very important. % doesn't provide that. Ideally, we capture both. We have run into an issue in the interpretation of 'duration' and 'interval'. Recording 2 min / hr, can get reported to us as a duration of 2 minutes and an interval of 60 minutes or a duration of 2 minutes and an interval of 58 minutes. Clearer guidance would help here - as you note, state explicitly "off time/off-effort" instead of interval. Perhaps % of time recorded would be a helpful addition as a high level metric but not be the only info capture $\endgroup$ Commented Oct 19, 2023 at 14:16