search

International-Soil-Radiocarbon-Database / ISRaD

Repository for the development and release of ISRaD data and tools
https://international-soil-radiocarbon-database.github.io/ISRaD/
24 stars 15 forks source link

Overlapping/duplicate depths within a profile #266

Open jb388 opened 2 years ago

jb388 commented 2 years ago

Layers with overlapping or duplicate depths within the same profile create issues when aggregating data. Based on earlier discussions of what constitutes a profile, I suggest that going forward, QA/QC should flag these profiles.

I don't think there are many of these profiles in the current version of ISRaD, but I recently ran across one (Gentsch_2018), so there may be more. In this case the authors had sampled multiple layers from a large pit face in order to better characterize the heterogeneity of the soil profile. I reassigned the original profile IDs to the plot level, and assigned the overlapping or duplicate depths to separate profiles. I added a note at the profile level to clarify this change.

Any thoughts about disallowing these sorts of overlapping profiles going forward? Any objections to implementing an official check in QA/QC?

@aahoyt @CaitlinPries @coreylawrence @ShaneStoner @SophievF

SophievF commented 2 years ago

I think it is a good idea to implement a warning in the QAQC. I recently used a package that flagged overlaying depth layers and I went back to these templates and it turned out that most of them a error in the depth column (numbers mixed, 5 instead of 6 etc.) However, I think there are also a few studies that took more than one sample from the same depth - I don't know how to treat these ones. One option could be to have a profile for each of them or to average them, but I think this will be a case by case thing - as you did it for Gentsch_2018

ktoddbrown commented 2 years ago

Another option might be to break these profiles at their layer boundaries and duplicate the layer values as needed. This will preserve any weighting for profile upscaling.