A few questions about implementing accretion feedback curves:
1. Timestep: I was expecting to run at a much smaller timestep because of the feedback between two variables, but when I ran the same model with 10- and 5-year timesteps I saw only very slight differences. Is SLAMM tracking elevation/accretion between timesteps? Or am I running the model incorrectly? If "Use Model" is set to True, this overrides the constant accretion rate specified above, correct? I'd expect that with a large timestep, a low-lying wetland cell would accrete very rapidly for the long timestep, overshooting and reaching a much higher elevation by the next timestep.
2. Wetland Types: Is there a reason this model is only available for four habitat types (Irreg Flood, Reg FloodTidal Flat, Tidal Fresh)? Are those categories actually generalized and include multiple SLAMM habitats? Seems like you would want to have the option for Transitional Salt Marsh and maybe Swamp/Tidal Swamp/Mangrove?
3. Are there any published examples of this model being implemented? I am selecting a reasonable low and high value and assuming an inverse linear relation between elevation and accretion.
I'm using SLAMM 6.2 beta.
1. We haven't noticed significant relationships between timestep and model results, especially with a difference of 5-years. I expect a spreadsheet model may help to understand the potential effect of overshoot and the compensating reduction of accretion rate in the next timestep. (While the longer timestep may overshoot, the subsequent timestep would compensate in the other direction as the marsh would have moved higher in the tidal frame.)
2. We had the most data for these four categories in the short run. However, in-house we have used this model for mangrove as well in the past, but just haven't yet integrated that into the interface. Transitional salt marsh, an irregularly-flooded category that comes from the NWI scrub shrub (or from dry land that has been converted to a salt marsh) we haven't used it for [note, this was a mistake, the irregularly-flooded marsh parameter or model is actually used as discussed below]. There has been too much uncertainty in our available data, though there would be no problem eventually adding that model to the interface in case others have a different opinion or set of data. Swamp is not tidal so that wouldn't be relevant, I don't think. Tidal swamp, again, there has not been an extensive enough data set or external model to populate the required parameters in SLAMM.
3. This is the best published example: http://www.bioone.org/doi/pdf/10.2112/SI63-0017.1 (Potential Effects of Sea-Level Rise on Coastal Wetlands in Southeastern Louisiana
Author(s): Patty Glick , Jonathan Clough , Amy Polaczyk , Brady Couvillion and Brad Nunley Source: Journal of Coastal Research, 63(sp1):211-233. 2013.)
Best regards! -- Jonathan
That was really helpful - the paper gave me some confidence in how I'm selecting my erosion rates. I also played around with a little spreadsheet model and that definitely showed that the model is not so sensitive to the accretion elevation feedback (or even the accretion rates themselves, within reason). I've attached an example output...
I'm surprised that the transitional marsh doesn't have an accretion rate - it seems like this would creep up the edges of dry land as sea level rises. I'll post a follow up on a recent connectivity thread since it's more relevant there.
Anyways, thanks for your help as always!
Sorry I made a mistake on my previous post -- the transitional marsh does have an accretion rate, it's actually just using the same model as the irregularly flooded marsh.
According to the latest technical documentation for transitional marsh: "The irregularly-flooded marsh accretion model or accretion parameter is utilized."
Oops. -- Jonathan