Sea Level Rise

Hello,

I am running the Coastal Vulnerability model and I have only 1 station for the whole coastline I want to analyse. I checked that we need to provide a shapefile with a TREND field representing the yearly sea level rise in mm. I used the data available in the http://www.psmsl.org/ to create a linear regression to estimate the rate of sea level variation. I come up with a 3.5 mm/year whcih seems to be in line with what I have read.

My question is: how do we make different scenarios of SLR for our model? Is it enough to multiply this rate value by, for instances, the future scenarios of global rise for 2100 predicted by the National Climate Assessment (0.2, 0.5, 1.2, 2 m)?

Thanks in advance


Comments

  • gverutesgverutes Moderator, NatCap Staff
    Hi pcabral,
    For our CV model, users can provide data representing the spatial variation of net sea level rise/decline: 1) within one scenario and/or 2) across multiple future scenarios.  Based on your description it sounds like your are designing scenarios to reflect the latter (alternative future SLR scenarios) but lack information showing spatial variation within each scenario (only one tide gauge/ station).  

    If my understanding is correct, then it may be difficult to answer questions about future SLR impacts with our CV model based on just one station in your entire study area.   You could localize the global scenarios from NCA to the 3.5 mm/year rate but it would likely lead to an obvious SLR variable scoring for your future scenarios of 2,3,4 and 5, respectively.  With each higher global rise scenario leading to a slghtly higher exposure index score (all other variables being equal).

    It would help me if you could share some visual information about your study area and the psmsl.org data you acquired.   Please attach a map or table to your response and I'd be happy to advise.

    -Gregg 

  • Hi Gregg,

    Thanks for your comment. The country I am working on is Mozambique. The only station with recent data is 986. I am putting in attach the excel file from which I have obtained the yearly rate.

    So, if I go for different scenarios with just one station, I guess I just multiply this value of 3.5 by the yearly rate provided by NCA scenarios. Right?


    Thanks in advance for your advice.

    Best regards,
    Pedro

  • gverutesgverutes Moderator, NatCap Staff
    My pleasure, Pedro.  

    Unfortunately, it's not as simple as multiplying 3.5 by the yearly rate provided by the NCA scenarios.  Below I list a couple of options now that I have a better understanding of your intentions...

    The current version of the CV model does not run multiple future SLR scenarios at one time.  Each run represents a "snapshot" in time.  If you were to run the model for each individual future SLR scenario, it would rank the SLR variable a "3" each time (since you only have one station) and you would see no difference across scenarios.   For our work in the USA and Myanmar, we used a post-processing script to account for this difference in SLR across future scenarios.  Since you don't have information showing spatial variation within each scenario for SLR variable, I believe you have just about reached the limit of what our model can tell you in regards to future SLR impacts and coastal exposure.  I would encourage you to consider other data sources to see if you might find down-scaled global climate projections.  For example, The Center for Climate Systems Research (CCSR) at Columbia University's Earth Institute has experience in this area.  

    If you'd like to proceed with the one station, attached is an Excel spreadsheet from our US NCA analysis (Arkema et al. 2013).  It shows how we localized the global NCA scenarios based on rates collected from each NOAA station along the US coastline. It sounds like you are on the right track with your calculations.  If you decide to go this route, please take a look at the formulas in the spreadsheet and let me know if you have any questions.  I can also share my Python script to help you run the model for alternative future SLR scenarios.
    -Gregg



  • Hi Gregg,

    Again thanks for your support and your excel file. You've been really helpfull! I have read your paper carefully.  

    In my case, I have only 1 station with data from 1961 to 2001, with only 12 observations. From here I calculated the trend (3.5). My "current" situation is year 2000, because of the data I have, e.g. habitats, population, etc. Using your  formulas I have calculated the values for each scenario. Instead of using 14 years as you do in your paper, I have used 39 years because this is the time range I have for my 12 SL observations. Now, I will do a model run for each scenario. It takes quite a few time because I am using lots of data. 

    If you find anything weird in my reasoning, please let me know (I am attaching the excel file with the calculations).

    Thanks.
    Have a nice weekend!

    Best regards,
    Pedro

  • Hi Gregg,

    In the meantime I understood exactly what you meant by "reaching the limit of what the CV can tell me in regards to future SLR impacts and coastal exposure if I have only one station". I have run the model several times with different trend values and the results are always the same. 

    I will try to find more data so I can work with climated change scenarios. 


    Thanks for your help!

    Best regards,
    Pedro


  • gverutesgverutes Moderator, NatCap Staff
    Pedro,
    Sounds like a good plan.  In some of the places we've worked it has been difficult to find data that is spatially varied for SLR.  Please let us know if you find more data.  I'd be curious to know the source.
    You may also consider looking at climate impacts to the distribution and function of natural habitats.  This would be adjusting your map layers and ranks for the "habitat" input variable based on future stress to coastal-marine ecosystems as a result of climate change.
    Regards,
    Gregg

  • Hi Gregg,

    I followed your suggestion  and I  have run 2 scenarios with and without habitats (mangroves, coral reefs and dunes).The results are very interesting as they allow to understand the role of the habitats in costal vulnerability. For a exposure > 3.5, there is an increase in impacted population of about 24% when habitats are removed. The (Arkema, 2013) paper was also very useful for using the same distances and rankings.

    Yes, I would love to use SLR but I guess I wll not be able to find suitable data. However, if I find it I will definitly share it here with you.

    Thanks for your guidance.

    Best regards,

    Pedro

  • gverutesgverutes Moderator, NatCap Staff
    Happy to help.  Feel free to post your preliminary results.  We are curious to see what you've produced.
    Best wishes,
    Gregg
This discussion has been closed.