I ran the elevation analysis in Slamm 6 and have the statistical output. I am unclear what it all means though. ???
The output table shows minimum elevations from -1 (estuarine beach, tidal flat, ocean beach and inland shore) to 1 (should I have negative values? How are the stats helpful? I am not clear what I should be comparing or looking at in those columns.
My initial condition (3 meter dem 2005, nwi 2006) and 2006 files have similar output except 2006 has alot more Estuarine Beach and an inland open water pond converts partly to esturine open water (that can't be accurate).
Also some of the code 20 (irregular flooded marsh) is converted to code 8 (regular flooded marsh), when I look at the lidar it follows the code 20 that was converted, so maybe that is more accurate?
Code 6 is reduced (I only have a small area in this category) in the 2006 output vs the initial condition. I read somewhere in the manuals that that code usually needs to be adjusted. How do I adjust it? or any of the others?
When I ran Slamm with a 10 meter DEM, only the beach was reduced significantly between the initial condition and the first year (2005 in that case).
I am also wondering how the elevation graphics tools help. I have looked at the profile and 3d view- should they be smooth? When I ran my 10 meter dem in slamm the tool showed smooth elevations with the preprocessor set to false. When I set it to true the elevations were up and down where roads etc were located in the marsh. Which is correct? I don't have any dikes and I 'burned' the impervious layer (1 and 2 slamm codes) into the nwi slamm layer as mentioned in one of the manuals.
one other question: I set my run using lidar data to 0 for frequency of large storms since I read that pertains to barrier islands and I have no barrier islands nearby. But is that accurate? There are big storms in the winter here.
I hope this makes sense. Thanks for your help.
I assume that you have looked at the users guide or used the "help button" from the elevation analysis screen...
According to NWI classifications, beaches extend down to MLLW. As the SLAMM vertical datum is MTL, this means that the beach will be one half-tide unit below MTL. Or -1 HTU (half-tide units). One HTU is MTL - MLLW, (or alternatively MHHW-MTL) So elevations from-1 HTU to 1+ HTU are to be expected.
Dry land below the salt-boundary will convert to estuarine beach if it is bordered by beach. I wouldn't stress the inland open water conversion, though that can occur if the water elevation is located below the salt boundary. Generally, though I just assume open water is open water...
The adjustment is to be made in the elevation analysis screen. Again, look at the help file it tells which columns are and are not editable.
The elevation graphic tools provide user feedback as to the interaction between the land cover and the land elevation with a 2-D or 3-D basis.
Do not use the pre-procesor if you have LiDAR. It is only for low vertical-resolution elevation data.
Leaving the frequency of large storms at zero should be fine if you are not modeling barrier islands.
I now understand the negative values and I read through the elevation analysis section (and user guide). I guess I just need to play around a little with the elevations and see what I get.
Interesting I tried setting the frequency of large storms to 25 years and I get streaking inland of several categories after 2050 (25 yr increments). I noticed I get alot of streaking in my NC runs as well. Is that ok, or do I need to adjust a particular setting? thanks.
The streaking in model predictions is often a function of overwash. What is your cell-size? The overwash was originally conceptualized with 100 meter cells and we have adjusted it to work with 30 meter cells but we have found things not to work as well with smaller cell sizes.
In one recent model application with 10 meter cell sizes we had to concede:
QuoteThe SLAMM overwash model was designed using data sets of a coarser resolution (i.e. larger cell sizes). For this simulation, overwash predictions resulted in a scramble of land-cover types rather than a coherent prediction of island migration. For this reason, the effects of large storms were ignored in this simulation.
I am using a 3 meter cell size for CA and 6 meters for NC. I will turn off the overwash parameter in NC and see if that fixes the problem. Thanks.