Visualize results of simulation run on original NC9 and NC9rev with water levels offset by 0.7 ft (0.21 m)

[ththelen]

@caseydietrich I am spinning off the visualization and analysis of the raised WL ADCIRC run into a new issue. Once the simulation finishes I will visualize water levels over both our NOV 1-15 time frame of interest and a part of late October as recommended by you to check water level offsets over a longer time frame.

CD comments: I will start a simulation with an initial condition to raise the water levels by 0.7 ft (0.21 m). I will use the original NC9 mesh, just so we can compare the effects of these two changes (higher initial water levels, deeper channels) independently. I should have these results in a few weeks.

Originally posted by @caseydietrich in https://github.com/NCSU-CHAZ/Sunny-Day-Flooding/issues/5#issuecomment-1180853342

[ththelen]

Good morning @caseydietrich, do you have an estimate for when the NC9 run with water levels raised by 0.7 ft (0.21 m) will be complete? We had previously discussed visualizing water levels from this run over both our NOV 1-15 time frame of interest and a part of late October. I am wondering when I should block off time to do that work.

[caseydietrich]

@ththelen Sorry for the dropping the ball on this. The simulation is finished, and I posted the full set of results on the cluster at the location below. Please copy these files into your space before working with them.

/share/jcdietri/For/Thomas/220711-NC9+0.21m-Nov2021/3-Nov1-Nov15.tgz

Regarding the re-analysis for the water level offset, I have been thinking about this, and I was incorrect to suggest that we should compute this from the model results. Instead, we should be using the observations at the gauges, and looking at the difference between the predicted and actual time series. This will give us a true measure of the effects of the long-term effects on the water levels. How hard will it be for you to download the observations from Oct and repeat your offset calculations?

I will start a simulation with both the 0.21-m offset and your revised NC9 mesh.

[ththelen]

Hi @caseydietrich, thanks for passing along the model results. It should only take a few hours to download observations from a couple NOAA stations (Wilmington and Wrightsville), then run the offset calculations for October 2021. As I understand your comment, this would just involve finding the average difference between Predicted and Verified water levels (as shown in blue and green respectively in the snip below) from the stations over the month of October. I do not understand why this is the preferred method to obtain an offset instead of comparing with model results, but perhaps this is something to discuss at our next meeting.

I can also work on pulling time series from the +0.21m run. I am hopeful I can visualize these results prior to our meeting next week.

[caseydietrich]

@ththelen It is a good question. The 'Predicted' water levels at the NOAA stations are not actually observed -- they are from a harmonic analysis of the tides without any other forcings -- what the tides would be, without any long-term atmospheric and oceanographic effects. The 'Actual' water levels are what was actually observed, so they do have those long-term effects. Their difference should give an indication of the magnitude of these effects.

By using only the observations, we avoid any possible biases in the model.

[ththelen]

Comparisons of NOAA predicted vs verified (observed) water levels are done. For the month of October 2021, NOAA verified water levels were an average of 0.51 ft higher than predicted values.

I was also curious how the NOAA predicted versus verified stacked up against the ADCIRC modeled vs NOAA verified over the Nov 9-14 timeframe that we used to come up with the 0.7 ft=0.21 m WL offset. That comparison is shown in the tables below.

I still have questions about why we would use the predicted NOAA water levels to determine the ADCIRC offset instead of the modeled results. I understand that the NOAA 'Actual' water levels are influenced long term effects that both ADCIRC and the NOAA 'Predicted' water levels do not capture. However, 'Predicted' water levels do not capture what I will call medium-term effects, like wind setup for example, that would be captured by modeled water levels. To me, it makes sense to parse out the long-term effects that ADCIRC cannot simulate by comparing it to model results that include medium-term effects, not just precited tidal harmonics. It seems to me like we would want to address biases in the model through the WL offset by comparing modeled results to actual results.

[ththelen]

@caseydietrich I think that our +0.21 m WL offset may have been lost at some point during the ADCIRC simulation and/or file jockeying process. When I visualized the time series results for the ADCIRC run you left in the /share/jcdietri/For/Thomas/220711-NC9+0.21m-Nov2021/ repository, I am seeing that the results from the +0.21m run are almost identical to the original NC9 runs. Plots for Wilmington and CB water levels with both the water levels overlaid are provided below. The NC9+0.21m WLs are almost exactly on top of the NC9 WLs at both locations. I calculated the average difference between water levels at both locations, and it is less than 0.005 feet. This leads me to suspect that the +0.21 m offset we thought we were running the model with got lost somehow. Does that seem like a reasonable conclusion to you based on this data, or is there something I am missing here?

[caseydietrich]

@ththelen Argh, okay, there must be an error in the simulations. Looking at it now, I don't see anything obvious, but let me look farther and re-do them. Sorry about this.

[caseydietrich]

@ththelen The results for the simulation of NC9rev+0.21m are available on the cluster here:

/share/jcdietri/For/Thomas/220808-NC9rev+0.21m-Nov2021/

The simulation of NC9+0.21m will take another 1-2 days.

[caseydietrich]

@ththelen The results for the re-run simulation of NC9+0.21m are available on the cluster here:

/share/jcdietri/For/Thomas/220809-NC9+0.21m-Nov2021

Hopefully the offset was carried through this re-run simulation.

[ththelen]

@caseydietrich I checked the two +0.21m simulations. The WL offset was definitely carried through the most recent simulation results you provided. Per our discussion last week, I will not spend any more time visualizing these results. I will wait for the +0.5 ft = +0.15 m offset NC9 and NC9rev simulation packages you are running to be complete as this is the offset we settled on in last week's meeting. Please let me know in issue #13 when these runs are complete and ready to share with me.