Hello, I am a new user of InVEST coastal vulnerability model and I have a couple of questions regarding setting up a regional model in the Mediterranean Sea.
First, the user made WWIII file includes the WavP columns: could you be a little more specific about the values in these columns and the way they are calculated (manually or not)?
Secondly, I intend to include the sea level rise option. Is there a preferred number of points in the shape file? Also, the ranking values 1-5 are automatically calculated by the model as I understand, is that correct?
The only info I have to offer comes straight from the User’s Guide:
16 columns named WavP_X , where X has the same values as listed above. These variables are used to estimate wave exposure for sites that are directly exposed to the open ocean. They were computed from WW3 data by first estimating the wave power for all waves in the record, then splitting these wave power values into the 16 fetch sectors defined earlier. For each sector, we then computed WavP by taking the average of the top 10% values
I guess the missing information is how to estimate wave power for all waves to begin with? I’m not an oceanographer, but I believe ocean wave power is a function of the wave height and period.
No. Ideally, the points represent the location of some empirical data. I would just use whatever you have.
Yes, as with many of the other variables, the SLR values are interpolated onto the shore points based on proximity, and then these values are converted to ranks 1 - 5 representing 5 percentiles of the data. See table here: Coastal Vulnerability Model — InVEST® documentation
I need to be very confident on the interpretation of some points, so please let me confirm the following:
So, actually the wave power is calculated -manually- with the equation 20 in the “wave exposure section” in “the model” chapter of CV user guide, which I believe agrees with your comment above. Is that correct?
In another post in Wavewatch III data generation: Wave power and Data information it think you mentioned that the model does the calculation. Maybe that did not strictly referred to the automated calculation by the model, and the manual calculation as described above is the case?
Also, regarding the use of “proportion” as referred in the user guide for the REI_PCTX and WavPPCTX values, it is defined as the division of the number of the 10% highest values of each sector / number of 10% highest values of all sectors (i.e. the total number of the values which are the 10% highest from all the whole timeseries)?
Finally, Height, Period and Wind speed in the guide, refer to Significant wave height, peak wave period and horizontal wind speed at 10m?
Yes, but it’s slightly more complicated than that. To determine overall “wave exposure” the model calculates the power of local wind-driven waves (using a height & period function), but relies on the WW3 data’s “WavP_X” for the power of ocean-driven waves. And then some logic is used to determine if a given shore point is exposed to the ocean waves or the local wind-driven waves. (Coastal Vulnerability Model — InVEST® documentation)
I think it’s like this:
Of the measurements (across all sectors) that fall into the top 10% of wave power, what proportion of them were measured in sector 0? And in sector 22, 45, and so on.
Of the measurements (across all sectors) that fall into the top 10% of wind speeds, what proportion of them originated from sector 0, 22, 45, and so on?
I can’t really speak to this myself. But you can see how the model applies the data in the wind & wave exposure equations from the User’s Guide. Do you think these particular metrics make sense to use in those equations?
Thank you for your answer, I believe understand the description form the user guide on how the model uses input data for determining overall wave exposure, with seaprate treatment of the ocean-driven and wind-driven waves etc.
Regarding WavP_X, my intention was to assure how to calculate the necessary values in the WavP_X columnn of the shape file’s attribute table, i.e. to secure that the related model input given by the user is correctly calculated, that’s why I initially put the question. I hope I understood right that the way is through equation 20 and no other implementation in needed.
As far as I am aware, these metrics are commonly used and since intense events’ effects have to be considered, it makes sense. However, other wave parameter definitions may be appropriate for the specific model, e.g. mean wave period, so a confirmation on what is expected by InVEST was the objective here.
Unfortunately we no longer have a coastal engineer on staff to share details into precisely what equations underlie the WavP values. I cannot say whether equation 20 was used. But I also don’t think it matters if you use precisely the same equations we used. There are no hidden assumptions in the model’s logic that would break down if some other equally valid formula for wave power was used instead.
The important thing is that WavP_X is calculated as the average of the top 10% of wave power measurements from each sector. However you see fit to calculate those wave powers is okay. And it could even be okay if you decided that the average of the top 5%, or top 15%, was a more meaningful metric than the top 10%, with respect to capturing exposure to intense wave events.
Thanks for sharing your opinions on that. I agree the wave power measurements should reflect intense events.
If/when we update the WaveWatch3 data, we will be sure to keep better documentation on the methods and make them easier to reproduce.
I think I get your point… Since the final outcome is a ranking (1-5), the important thing is not the formula itself but to use the same formula for all the caclulations. That is to say the values of wave power as such do not affect in any other level but the input for the ranking system caclulation (percentile count in ranking table 1).