NTR: 4x P01 codes for SPI Sediment Analysis [BODCNVS-2309]

danibodc commented 2 months ago

4 new terms requested for parameters produced from Sediment Profiling Imagery (SPI) analysis.

Details of the Sediment Profiling Imager: https://www.oceanimagingsystems.com/docs/REMOTS-3731D-24MP.pdf https://www.oceanimagingsystems.com/products.php#SPI

the prism penetrates the top few centimetres of the seafloor
obtains a vertical cross section of the sediment-water interface by imaging the sediment in profile via a water-filled prism.

At each SPI station the camera was ‘hopped’ along the sea bed at 10m intervals, providing 5 images for analysis. Images were analysed using the software ImageJ for penetration depth, apparent redox penetration depth (aRPD) and surface roughness, following the method as described in Teal et. al (2010).

Originator's Parameter names:

Total aRPD area (cm^2)
Average aRPD depth (cm)
Prism penetration (cm)
Surface roughness (cm)

These are described on p.180: https://www.researchgate.net/profile/Jonathan-White-32/publication/269630850_Review_of_standards_and_protocols_for_seabed_habitat_mapping/links/55e06b7608ae2fac471b6de3/Review-of-standards-and-protocols-for-seabed-habitat-mapping.pdf

prism penetration depth: gives an indication of relative sediment compaction
sediment boundary roughness: indicates the degree of physical disturbance or biotic activity at the sediment water boundary
sediment apparent redox potential discontinuity depth (aRPD): assesses the depth of oxygenated sediment on the bottom

Proposed P01 preferred labels: <html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

Originator's label | Proposed P01 preferred label -- | -- Total aRPD area | Area of apparent redox potential discontinuity layer {aRPD} in the sediment by sediment profiling imager and computation using ImageJ software algorithm Average aRPD depth | Depth (mean) of apparent redox potential discontinuity {aRPD} in the sediment by sediment profiling imager and computation using ImageJ software algorithm Prism penetration | Penetration depth of prism in the sediment by sediment profiling imager and computation using ImageJ software algorithm Surface roughness | Surface roughness of the sediment-water interface by sediment profiling imager and computation using ImageJ software algorithm

R.e. 'Surface roughness of the sediment-water interface by sediment profiling imager and computation using ImageJ software algorithm'

Am not sure if not specifying a matrix is ok here. The sediment-water interface is technically the floor/bed.

danibodc commented 2 months ago

Would require the following new elements:

S06 - Surface roughness I have seen several different terms used to describe this: sediment boundary roughness, surface boundary roughness, surface roughness. 'Surface roughness' is present in the data files we have received. All appear to be synonymous.
S06 - Penetration depth Similar to existing term https://vocab.nerc.ac.uk/collection/P01/current/CONEMXDP/
S29 - apparent redox potential discontinuity (with and/or without 'layer')
S29/S26 - sediment-water interface Alternatively we could use 'seafloor' which is proposed as a new S18 here: https://github.com/nvs-vocabs/P01/issues/213, however we might need to be broader e.g. 'bed'. However the literature related to SPI favors 'sediment-water interface'.
S04 - sediment profiling imager
S05 - computation using ImageJ software algorithm

gwemon commented 2 months ago

Thanks @danibodc. A few comments and questions in relation to this ticket

"Penetration depth" is a useful concept to have in S06. However as a stand alone term, it needs to be defined carefully because it can be applied to multiple domains and take slightly different meaning when associated with an object of interest and matrix in the P01 combination. QUDT does not have a broad "Penetration depth" in its collection of quantity kinds but it has a more specific term "https://qudt.org/vocab/quantitykind/LondonPenetrationDepth"

Regarding the associated P01, I would prefer e.g.: "Penetration depth of the instrument prism in the sediment by etc" - but do we even need to specify "prism"? also do we need to specify by sediment profiling imager and computation using ImageJ software algorithm? A more generic alternative could be: "Penetration depth of the instrument in the sediment"?

I note though that you quoted that "prism penetration depth: gives an indication of relative sediment compaction"

gwemon commented 2 months ago

For: Area of apparent redox potential discontinuity layer {aRPD} in the sediment by sediment profiling imager and computation using ImageJ software algorithm and Depth (mean) of apparent redox potential discontinuity {aRPD} in the sediment by sediment profiling imager and computation using ImageJ software algorithm

I think this works well. It could be that we need to use the S06 for Depth (spatial coordinate) here? Because it is the depth at which something is rather than the depth of something? Not sure...

gwemon commented 2 months ago

For: Surface roughness of the sediment-water interface by sediment profiling imager and computation using ImageJ software algorithm

Yes, I think that works well. Surface roughness would be a useful re-usable concept to have in S06 although it might be tricky to define? I see that its expressed as a length but could not find a definition that relates it to a length in my (very quick) search.

for the S29 yes I do like sediment-water interface. Again this is a useful re-usable concept to have. I agree that it could also be seafloor or sediment (bed seems strange in tis context without a "the" in front of it. But sediment-water interface would get my vote.

alicecbodc commented 2 months ago

Thanks everyone for taking a look at these codes! They look great. I've got a couple notes for context:

Surface roughness calculation - from the research I have done into ImageJ, what I believe has happened is that the originator has used a plugin to effectively turn the interface into a line graph in cross-section (often called rugosity or Ra in geomorphology). However, I found the following link to a study of rugosity by SPI analysis, which calculated surface roughness (here, "Boundary roughness - maximum minus minimum penetration depth for each replicate image") https://apps.ecology.wa.gov/publications/documents/0703025.pdf, page 28. This isn't quite as accurate as rugosity as a calculation, so it could very depending on the time the originator spent analysing the images and modelling that sediment surface as a line. Ra would be the gold-standard of roughness analysis, whereas the penetration depth analysis outlined above would be a lot less accurate.
This link here goes into a little more detail on Ra (surface roughness): https://get-it-made.co.uk/resources/surface-roughness-explained#:~:text=It%20is%20calculated%20as%20the,a%20surface's%20peaks%20and%20valleys. - this isn't scientific, however it draws a lot of parallels to some available ImageJ plugins for surface roughness in the way that it is calculated as a length. Without going back to the originator, I'm not sure we would know exactly what algorithm is being used here
Sediment-water interface vs. other terms - with my geology hat on, I would agree that sediment-water interface encompasses this more comprehensively than "bed", "seafloor" or "sediment" because of the transitory nature of the interface

I hope this helps - do let me know if you need any more info on the above!

nvs-vocabs / P01

NTR: 4x P01 codes for SPI Sediment Analysis [BODCNVS-2309] #266