News:

Due to SPAM attacks, new members must be approved before posting.  Please email jclough@warrenpinnacle.com when registering and your account will be approved.

Main Menu

Time step and time zero

Started by cpapiez, November 12, 2010, 01:04:26 PM

Previous topic - Next topic

cpapiez

Hi Jonathan,

What is the best time step to run for NWI dates that are primarily in the 1980's?  We are running SLAMM by county so should we adjust the timestep based on the time zero/NWI date?  When we compared a 25 year and 10 year time steps for Dorchester County (NWI 1991) we saw significantly different results for tidal flats (15532 vs. 4603), Regularly flooded marsh (13013 vs. 3906), Transitional Salt Marsh (7738 vs. 2748), and Estuarine Open Water (106751 vs. 131923).

We are primarily interested at 2050 and 2100 results.  Any advice would be really appreciated.  As of right now, we have run all Maryland coastal counties with a 25 year time step but I'm concerned about the potential difference if we would have run it with a 10-year time step.  What are most people using for large scale projects like this?

Thank you!

Chelsie

Jonathan S. Clough

Hi Chelsie:

SLAMM should produce pretty-much the same results for smaller and larger time-steps.  One difference may be that a category only changes classification to its next successive category once per time step.  Therefore if you have dry land with open water elevations that dry land will need to change through transitional marsh, salt marsh, tidal flats, and finally open water so it will only become open water after four time steps.  Therefore there may be some differences and running with more time steps will be somewhat more precise, especially in making the model adjust to the initial condition.

In my last test of the model, I saw very little differences between 5 year and 25 year time steps.  Sounds like your experience is different.  I would probably think this is a result of marshes converting to tidal flats in one time step and open water in the next time step.  Also, erosion for tidal flats won't start until the tidal flats are created so if it takes 25 years for that creation to take place you might be missing out on some of the erosion.

Overall, I'd generally think that models run with more time steps are somewhat more accurate.  It's worth testing on your site, though, and understanding the reason for any differences is also important.

-- Jonathan