NY Regression 2

[dbing]

I am using the new "Magnitude and Frequency of Floods in New York" (USGS SIR 2006-5112) and have a couple of questions regarding a few of the variables for different regions.

1. What is the "contour-band method" used to compute Average Basin Slope (BS)? I have never heard of it.

2. Where can I get the 30-m DEM maps that is needed for some of the variables?

3. Where can I get the National Land Cover Dataset (NLCD) maps needed for some of the variables?

I know that the Flood-Frequency Tool that comes with the report may do some of these things for me, but the GIS software needed to use the program is expensive and I am trying to determine if I need to purchase it or not.

Thanks

 

[RWF7437]

I know nothing about New York, not even in June. But, it seems to me you are worrying about the wrong things. Regression equations are derived from stream flow records. They are statistical and are subject to variations ( i.e. errors ). Typical error band widths are plus or minus 30% to 50%.

The parameters needed to solve the equations can almost always be derived from the published USGS Quadrangle maps for the drainage basin if interest. Parameters NOT shown on the USGS maps may be found by searching local records, GIS databases or by GOING TO THE AREA AND LOOKING AT IT.

None of this is expensive. None of this requires a computer. None of this requires expensive software.

"Contour Band Method" sounds like a fancy name for a simple idea. In the literature you have, look for the definition of "average basin slope".

good luck

 

[RWF7437]

http://water.usgs.gov/software/NFF/manual/ny/index.html

 

[youngEIT]

I am not familiar with NY regression equations but am familiar with some other states. you probably need to evaluate significant watershed characteristics to plug into the regression equations. you can get 30m or 10m DEM to get the watershed physical characteristics from seamless.usgs.gov. you probably would need land cover in vector format (polygons) to evaluate how much of your watershed has impervious, woods etc. try

http://landcover.usgs.gov/landcoverdata.php#na.

you can also try

http://www.epa.gov/ost/ftp/basins/gis_data/huc/.

it has land cover shapefiles by HUC codes, but i guess they are 1994 or 96 data. to know your HUC code, you can go to

http://nwis.waterdata.usgs.gov/nwis

and find your nearest USGS gaging station. you can get more up to date land cover data from the DEP. you probably would need a GIS software like ArcInfo, WMS or BASINS to evaluate the watershed characteristics.
No idea about the contour-band method.

 

[RWF7437]

dbing,
It would help if you explained what it is you are trying to use the regression equations for. You can use them, as suggested in my previous post, in a very simple, straightforward way. Or, you can invest a lot of time and money by using achhibber's links and GIS software. Unless you are required by your client, or by some regulatory agency, to take the expensive route the simpler would be preferable.

good luck

 

[youngEIT]

RWF7437,
the reason why I mentioned all those links and softwares is because I had to go through all of them when I performed a state-wide flood frequency analysis couple years ago. sometimes, there are 450+ watersheds in your study area for which you need to derive physical characteristics to regress against your peak flow quantiles, which, doing it by hand on a USGS topo would take a lifetime (or maybe more).
also, a lot of states have their equations with independent variables which are not easy to estimate from a USGS quad map (% carbonate area, basin development factor etc), and therefore, warrants use of digital spatial data. apart from that, calculating unique individual areas based on landuse or soiltype (A,B,C or D) for large watersheds is also easier in a GIS program than manual effort. but you are right, it all depends on the level of detail the client wants and the money he wants to spend. if its not a big area of interest, then going down your route is the best option.
Regards.

 

[TerryScan]

I see this is a bit late in relation to the original post. The USGS is developing online models for regression analysis.

http://water.usgs.gov/osw/streamstats/ssonline.html

It shows that NY is undergoing implementation. One may wish to them to check when full implementation is expected.

You may wish to check out a nearby state to see what to expect from the tool, and evaluate whether it will suite your needs.