Recommended way to point out data errors?

[dankovacek]

i.e. HYSETS station 05QA002 reports average discharge ~107 mm/day. I believe this is due to an incorrect drainage area reported in the attributes file based on the corresponding streamflow .csv. I did not check .nc file version. The correct drainage area is 6170 km^2 which gives about 0.8 mm/day.

What is the most helpful way to communicate findings like these?

[kratzert]

Hi @dankovacek,

This is actually a good place to communicate these findings, similarly you could use the Caravan google group.

For this particular gauge: It seems like all the data that we have (gauge location, polygon) seem to agree on ~50km2. I took both of the information from HYSETS, so I assume the error is already there. That being said, I can easily correct that for Caravan. Just one question: If the area is wrong, we should not only update that, but also the polygon, the gauge location, and therefore also the forcings and attributes, as they were derived for the current polygon and gauge location. Do you happen to have the correct gauge location and a corresponding basin polygon?

Best,
Freddy

P.S.: You weren't at EGU last week, or were you? I saw your name in the good modelling practices session but couldn't stay for the talk. I think I spotted Steven Weijs in the audience, so I assumed he would give you talk?!

[dankovacek]

Hi @kratzert, thanks for your response!

The Water Survey of Canada updated several thousand monitoring network polygons since HYSETS was published. They provide polygons, station locations, and conveniently also pour points.

I made an automated documentation generator for summarizing hydrometric dataset changes from source providers. The idea was to automatically process revisions to catchment polygons, find 'breaking changes', and generally organize quality documentation for downstream users, i.e. LSH datasets. It is a pitch to the data owners (WSC) to use an open platform for quality data and knowledge sharing through public/searchable feedback. I processed the WSC component of HYSETS as an example. It generates network summaries, and links to station specific quality information. I called it Camel Farrier, there's a paper in the review pipeline but it's been waiting for first decision for a few months now. But you get the idea.

I would be happy to process these in Caravan format for HYSETS after I submit my thesis draft June 1st -- part of the reason for passing on EGU this time around, sorry to have missed you. :)