Hi,
I am been using coastal vulnerability model and have some question about the inclusion of sea level rise (slr) trends in the model.
I already have done some runs, and would like to include some sea level rise scenarios within my analysis. I am been using the RCP trends for sea level rise in my area of interest (e.g. 0.2m of project slr for the year 2065, equates to 10 mm/year to the 20 year period studied).
The problem I have is that running with the slr information, the vulnerability index is lower when comparing with the run without slr information, without changing anything else!
As my understanding the vulnerability index would be higher, given that the exposure equation is a sum of the different ranks, but I didn’t find any place where to rank slr, only to include the projected trend (mm/year) within the shapefile, as referred in the manual.
Where I could making some mistake? I already have a shapefile with trend column, but nowhere to rank this input.
Hi Jacinto, thanks for posting. Dealing with SLR can be tricky in this model, as it tends to be somewhat different than the other variables. The interesting variation of SLR is usually the variation over time (across scenarios) as opposed to variation across space (within one scenario). Most of the other variables, and the model as a whole, are designed to show relative differences across space.
The model handles the ranking of the SLR values in a similar way that it ranks other continuous variables, such as relief or surge potential (see Table 4.1 of the User’s Guide and How it Works: Sea level change). That may not be useful at all if your entire area of interest is subject to the same SLR trend. If the trend varies across space in your AOI, then it might be useful.
So, if you wish to compare SLR scenarios where the SLR trend for the whole region changes from one scenario to another, you might do this calculation manually after running the model. For example, you could open the coastal_exposure.csv output and create a new R_slr column with rank values of 3 for a baseline run, and then a column with ranks of 4 or 5 for a scenario. And you could use all the columns prefixed with R_, to re-calculate the exposure index using equation #1 in the user’s guide.
Finally, the exposure equation is a a geometric mean, rather than a sum, so adding another variable to the equation will not necessarily yield a larger exposure value.
Thank you very much for your input, Dave. It was really helpful.
It is true that for my study region I just have overall trends for each given scenarios and not differences within scenarios between different area. I was thinking of doing that if nothing else worked, and yeah, I forgot about the mean in the equation
Hi all,
I found this thread to be very helpful, as I had exactly the same problem and now I have a way forward, thank you.
I was wondering, tho, @dave, when you said to use rank 3 for baseline runs and ranks 4 or 5 for a climate change scenario, did you actually recommend to rank baseline as 3 because it is the medium value or was just a hypothetical?
I want to run the model for current and future climates under 3 SSPs scenarios (say for 2050 and 2100 for example), so I thought to calculate the rankings same as the other variables: divided in percentiles. So I would get current sea-level rise values and all three SSPs sea-level rise predictions and use these values to obtain the percentiled rankings (for each of time frames). By doing so the current sea-level rise would become a rank of 1 because it is the lowest.
However, should we maybe consider 0 mm/years of sea-level as the first value on the rank? Assuming the lowest exposure would be absence of sea-level rise. I know in the model the percentiles are calculated only within the range of appeared values, however since sea-level is a post-model estressor there is this possibility. If so, current sea-levels wouldn’t be ranked as 1 anymore, but would be somewhere in the middle closer to future scenarios.
Does anyone have any input on how to go about that?
Thank you,
Stella
I suggested it because it was the middle value, but that was a very simple and arbitrary suggestion. I think your other ideas are great improvements on this.
This makes sense and sounds very defensible. Then your scenario results would framed as “risk from future increased rates of sea-level-rise”.
Technically there are places in the world where relative sea-level is dropping as land is uplifting, so in that sense 0 mm of SLR would not represent the lowest possible exposure. That might not be relevant for your study area, but I think it makes the point that what matters more are the rates of SLR actually represented in your study area and in your scenarios.
In that context your first idea of calculating ranks based on percentiles of the observed values makes the most sense, I think. If you wanted to, you could also include scenario where there is 0 mm of SLR. It wouldn’t be a realistic future scenario, but it could be useful to illustrate the point that the current rate of SLR is already creating some amount of vulnerability.
