Plot Tide Forecasts notebook returns unexpected output for TPXO9-altas-v4?
Closed this issue · 5 comments
Hi Tyler,
Thanks for making this tool for Python. I am new to both this tool and python, so I might be doing something wrong, but when I compare the output of the "Plot Tide Forecasts" notebook at the La Jolla Tide Gauge (default location) using the TPXO9-altas-v4 model (Svetlana Erofeeva gave me access to the model files via the box app), I get output that seems incorrect. I have tried this for a variety of dates and compared to two other versions of La Jolla Tide data:
This GUI: https://tidesandcurrents.noaa.gov/noaatidepredictions.html?id=9410230&units=metric&bdate=20210629&edate=20210706&timezone=GMT&clock=24hour&datum=MSL&interval=hilo&action=dailychart
and this netCDF dataset water_levels.nc : https://datadryad.org/stash/dataset/doi:10.5061/dryad.n5qb383
Any chance you also see an inconsistency / have ideas what could be going on?
Thanks for your help!
-Bonnie
Hi Bonnie,
Great question! There was something weird going on. The SIO tide gauge was right near the boundary of the tide model so it was interpolating weird. I turned on the EXTRAPOLATION
option for each of the tide model readers which should hopefully fix the problem you're seeing. Plots with the new extrapolated estimates (and the tide chart from NOAA) are below:
I'll put in a PR to fix this in a second. Thanks for pointing it out!
Cheers,
Tyler
Oh wow, thanks so much Tyler! I've been banging my head on this for a while and really glad I posted here. Really appreciate it!
Hi @bludka, are you still having trouble? Got an email last night but can't find your messages here.
Hi Tyler,
Apologies, I think that was user error. Looks good now!
-Bonnie
Hi Bonnie,
Glad things worked out 😀
Tyler