ASGS NCFS Sandy Advisory 27 not working

[rsignell-usgs]

The ASGS NCFS Sandy Advisory 27 is not working http://cmgdata.usgsportals.net/#module-metadata/c9f450b6-ce87-3ac0-b8b7-8ae7f27f9051/a37fc631-e4d1-426d-a2c7-de79d5710818

I'd like to check the WMS endpoint to see where the problem is, but where do I find the endpoint?

[kwilcox]

@rsignell-usgs Yes the WMS server isn't working, but only because the DAP endpoint you are advertising in your CSW isn't working: http://mrtee.europa.renci.org:8080/thredds/dodsC/ASGS/sandy/27/nc6b/blueridge.renci.org/ncfs/nhcConsensus/00_dir.ncml

[rsignell-usgs]

Okay, thanks. Is there a way I could have figured this out on my own?
(like access to error logs, or ability to click on a dataset and see what endpoints are being accessed?)

[kwilcox]

The services are available using the "Download" dropdown (which was added as part of this project)

I have since taken it off of the Axiom sci-wms server, it will be readded automatically when the CSW finds a good DAP endpoint to work with.

[rsignell-usgs]

Excellent. Thanks. @brianoblanton or @cbcunc, do you know where these Adcirc datasets moved to? I can't even get to: http://mrtee.europa.renci.org:8080/thredds/catalog.html

[rsignell-usgs]

Ah, I should have googled first! I just tried searching on "thredds" and "renci" and this came up: http://opendap.renci.org:1935/thredds/catalog/ASGS/sandy/17/nc6b/blueridge.renci.org/ncfs/nhcConsensus/catalog.html?dataset=ASGSTCResults2/sandy/17/nc6b/blueridge.renci.org/ncfs/nhcConsensus/00_dir.ncml

[rsignell-usgs]

I propose that it's time for renci to try sci-wms, now that we've got it dockerized.

[BrianOBlanton]

?yes, the mrtee server is just a dev machine for thredds. It's my desktop.

Sorry about that. if there are other datasets pointing to mrtee.europa.renci.org, let me know and I'll make sure they are on the production server (opendap.renci.org).

[BrianOBlanton]

Sounds good. It'll be the first of the year, though, bec/ @cbcunc? is out 'til then, and @brianoblanton is out with flu+pneumonia, and wouldn't know what to do in any event.

[rsignell-usgs]

In the meantime, I added the production opendap endpoint to my sci-wms, and it's now in our CSW service, so it should get picked up automagically.

[rsignell-usgs]

@cbcunc, do you have your head above water yet?

@brianOBlanton, @cbcunc, sci-wms is now easily installable using Docker, thanks to the work of @kwilcox and @daf: https://hub.docker.com/r/axiom/sci-wms/

[rsignell-usgs]

The WMS request was going to Axiom's WMS sci-wms, which I think pointed to the old opendap site. I've upgraded the metadata in our CSW so that it points to the USGS sci-wms endpoint, which has been updated to point to the correct RENCI opendap site. So as soon as our CSW gets harvested by Axiom , the ASGS stuff should work again, and we should see good water levels here:

http://cmgdata.usgsportals.net/#module-metadata/c9f450b6-ce87-3ac0-b8b7-8ae7f27f9051/a37fc631-e4d1-426d-a2c7-de79d5710818

oops, actually no. Talking with @kwilcox ,he reminded me that we also need to set display=true for every parameter we want to show up. So that means that we need to have @BrianOBlanton add these attributes to his .ncml files. :disappointed:

[rsignell-usgs]

@BrianOBlanton has updated his ncml, setting display="true" for zeta and swan_HS (free surface and significant wave height). Axiom harvests every 3 hours, so if all goes well, this should be working at 6pm ET (in 1.5 hours or so).

[rsignell-usgs]

Okay, the good news is that significant wave height showed up! Woo hoo!

Something a bit funny in that the legend is missing and I can't change the color limits. So I need a new version of sci-wms or something?

[rsignell-usgs]

@kwilcox , this also exposes a tiling problem with certain wms requests, eh? It looks like the offshore triangles along certain tile seams are bigger than sci-wms expected, resulting in gaps along the offshore tile seams. Should we raise this issue on sci-wms?

[kwilcox]

This looks be a sci-wms regression bug from when we moved to using pyugrid. sci-wms pads the triangles that it draws to try to avoid this, but it seems to not be padding enough and losing the outside triangles when they are large.

[kwilcox]

@rsignell-usgs There is no scale settings because the client can't figure out what units the model are in.

Example:

long_name: "water surface elevation above geoid"
standard_name: "sea_surface_height_above_geoid"
coordinates: "time y x"
location: "node"
mesh: "adcirc_mesh"
units: "metric"
_FillValue: -99999.0

All of the variables need to be updated with appropriate units

[kwilcox]

Issue about the missing grid cells here: https://github.com/sci-wms/sci-wms/issues/113

[rsignell-usgs]

@BrianOBlanton , Can you please modify the NcML for the dataset http://opendap.renci.org:1935/thredds/dodsC/ASGS/sandy/27/nc6b/blueridge.renci.org/ncfs/nhcConsensus/00_dir.ncml.html again and add the units attribute units="m" to the variables zeta, Swan_hs, etc?

[BrianOBlanton]

Sure, as soon as I get in, probably around 8:20

Cheers,

[cid:d5779455-8492-4dab-ba24-39760e2b2bc6]

Sent from Ninehttp://www.9folders.com/

From: Rich Signell notifications@github.com Sent: Feb 2, 2016 5:39 AM To: USGS-CMG/usgs-cmg-portal Cc: Brian Blanton Subject: Re: [usgs-cmg-portal] ASGS NCFS Sandy Advisory 27 not working (#128)

@BrianOBlantonhttps://github.com/BrianOBlanton , can you please modify the NcML for the dataset http://opendap.renci.org:1935/thredds/dodsC/ASGS/sandy/27/nc6b/blueridge.renci.org/ncfs/nhcConsensus/00_dir.ncml.html once again and add the attribute units="m" to the variable zeta?

— Reply to this email directly or view it on GitHubhttps://github.com/USGS-CMG/usgs-cmg-portal/issues/128#issuecomment-178502511.

[BrianOBlanton]

OK. Done…

Cheers,

[cid:67825134-05E6-46B5-B50E-6C520AE6F1AE]

Brian O. Blanton, Ph.D. Director of Environmental Initiatives Oceanographer Renaissance Computing Institute University of North Carolina at Chapel Hill 100 Europa Drive Suite 540 Chapel Hill, NC, 27517

Brian_Blanton@Renci.Orgmailto:Brian_Blanton@Renci.Org 919-445-9620 (O)

From: Rich Signell notifications@github.com<mailto:notifications@github.com> Reply-To: USGS-CMG/usgs-cmg-portal reply@reply.github.com<mailto:reply@reply.github.com> Date: Tuesday, February 2, 2016 at 5:39 AM To: USGS-CMG/usgs-cmg-portal usgs-cmg-portal@noreply.github.com<mailto:usgs-cmg-portal@noreply.github.com> Cc: Brian Blanton bblanton@renci.org<mailto:bblanton@renci.org> Subject: Re: [usgs-cmg-portal] ASGS NCFS Sandy Advisory 27 not working (#128)

@BrianOBlantonhttps://github.com/BrianOBlanton , can you please modify the NcML for the dataset http://opendap.renci.org:1935/thredds/dodsC/ASGS/sandy/27/nc6b/blueridge.renci.org/ncfs/nhcConsensus/00_dir.ncml.html once again and add the attribute units="m" to the variable zeta?

— Reply to this email directly or view it on GitHubhttps://github.com/USGS-CMG/usgs-cmg-portal/issues/128#issuecomment-178502511.

[rsignell-usgs]

@kwilcox, I just tried displaying the significant wave height during Sandy from this dataset and encountered a number of problems. http://cmgdata.usgsportals.net/#module-metadata/c9f450b6-ce87-3ac0-b8b7-8ae7f27f9051/b5bca240-0bb6-4a22-ba78-2da62873f0a0

1. When I move the time slider back to the peak of the storm (oct 30), there are wave heights above the color scale
2. When I try to change the color scale, I can't move the sliders to increase the max wave height
3. I can type in a higher wave height to increase the color scale maximum, but then the entire image turns white
4. I can't click and get a time series.
5. Color scale is oriented the wrong direction

[rsignell-usgs]

@kwilcox, where do we stand on this issue?

[rsignell-usgs]

@BrianOBlanton, now that ncWMS2 supports UGRID, would you be willing to try out installing ncWMS2, perhaps on the same Tomcat that is running the THREDDS Data Server?

Grab the ncWMS2 war from here (as the latest improvements haven't made it to a new release yet): https://github.com/Reading-eScience-Centre/edal-java/issues/47#issuecomment-199256262

Documentation (but it's simple!) here: http://reading-escience-centre.github.io/edal-java/ncWMS_user_guide.html

In the future when ncWMS2 integrated with the TDS, it could be then removed...

[BrianOBlanton]

@rsignell-usgs,

Yes. We'll do it. Will this work with tomcat5? that's what the production server is still running.

But we have a dev server with tomcat7.

[rsignell-usgs]

@BrianOBlanton , I think it requires tomcat7, so try the dev server. Maybe best to try it there anyway -- this is still experimental. When I access the data remotely, it works, but it is SLOW.

[BrianOBlanton]

So does ncWMS2 turn the triangles into hexagons?

[rsignell-usgs]

I'm including @guygriffiths to be sure, but I think the answer is that for triangles with node-based data, the default-sc layer option creates flat-shaded facets that are the "dual" of the triangles, so yes, hexes.

[guygriffiths]

To confirm, ncWMS2 uses a nearest neighbour colour scheme for data where the faces are not explicitly specified. So for a triangular node based grid, this appears as hexagons.

[rsignell-usgs]

@BrianOBlanton , did you guys ever get around to installing ncWMS2? If not (or even if you did), you can now install ncWMS2 using Docker, thanks to @kwilcox: https://hub.docker.com/r/axiom/docker-ncwms/

[rsignell-usgs]

Moving this to milestone v2.0 also.