Standardize kernel density data?

360
1
12-13-2010 08:27 AM
ElizabethPhillips
New Contributor
Hi there,

I have transect survey data with animal counts that I successfully calculated kernel density for.  However, there are >20 transects total, and in some areas (not all), multiple transects were conducted.  Thus, the densities for these areas are inflated.  I would like to standardize my data for number of transects conducted within a certain cell. 

I have pondered this and tried a few things.  First, I tried to extract the kernel density raster and join to the table with transect data.  However, this doesn't work as my original table has >18,000 points, and the raster extraction then tries to match to each point.  Is there a way to extract kernel densities and match to the metadata (e.g., transect name)?  I don't think so, given how kernels work...

My next thought was to calculate kernel densities for each individual transect (>20), and then try to calculate mean density for all transects within cells.  However, I got stuck with this, as the cell statistics only seem to allow input of two different rasters at a time. 

I suppose I could calculate mean # animals/transect in advance, but each point is an individual location, and I would have to first bin the data into cells (say, 3-km), which I'm trying to avoid (unless anyone has advice on an easy way to do this!)...

Does anyone have any other ideas?

Thanks in advance...
0 Kudos
1 Reply
LynGarrah
New Contributor
Hi Elizabeth,

It has been a while since you posted your question.  I am in a similar situation, with road mortality data on one road over 4 years but with different survey efforts in each of the four years of my data set.  Did you figure out how to standardize kernel density layers?  I would like to add my layers together to find the worst road mortality locations over the four years, but I do not want to bias those locations toward the years with more survey efforts.
0 Kudos