KernelHR 4.27
The program can be run either in interactive or command line
mode. The program is highly flexible and powerful, allowing
the user much control over the estimation method. Default
values simplify operation in command line mode, and provide
the methods that have been found to give the most accurate
results (Seaman and Powell 1996). When used interactively,
the program prompts the user to enter file names, and to make
choices about what options to run. When run in command line
mode, these choices are made with command line parameters.
Output includes area estimates from both the fixed and
adaptive kernels (the fixed kernel estimate is highly preferred),
as well as many indicators of internal conditions and diagnostics
for the estimates. Optional output includes files of density
estimates for numerical or graphical analysis, and contour
levels for specified volume percentages of the estimated
utilization distribution.
HIGHLIGHTS OF KERNELHR PROGRAM FEATURES
1. Many ranges can be run, either in a single file or in
multiple files, without constant interaction from the keyboard.
It is not necessary to make individual files for each home
range, nor to start the program manually for each one.
2. There is no fixed limit to the number of observations
per range. The program can run as many locations as the
memory of your computer will hold. Rough calculations suggest
that with 8 megs of RAM, the program should be able to handle
about 10,000 observations per home range (providing that
nothing else is using memory, i.e. you are not running within
Windows, etc.).
3. The user has great control over the smoothing parameter.
The major automated methods for choosing the level of smoothing
are available (Least Squares Cross Validation, the Reference
Method), as are manual methods. LSCV is performed on normalized
data, making it suitable even for data measured in different
units in the two dimensions (i.e. non-spatial data).
4. Output is available for almost any values calculated
by the program. In addition to the area of the estimate,
you can output densities or utilization distribution at
grid points, or densities at observed locations. Furthermore,
the program saves much information about each run: grid
size, number of grid cells, amount of smoothing, method
of selecting the smoothing parameter, and the volume of
the estimate.
5. You can select the units for output of the area estimates:
square kilometers, hectares, and square meters are available.
6. Contour levels for up to 11 levels of utilization distribution
may be output. This is true "volume percentage" data, rather
than "point percentage". Volume percentages give the exact
percentage of the UD, point percentages only give estimates.
7. The level of smoothing can differ between the two dimensions.
8. The user has optional complete control over the grid
spacing, there is no maximum grid size.
9. On the down-side... There are no pop-up menus or slick
output graphics. All output is in ASCII format, and must
be imported into a graphics or GIS program for presentation.
But this gives you complete control and flexibility!
D. Erran Seaman
National Biological Service
Forest and Rangeland Ecosystem Science Center
Olympic Field Station
600 E. Park Ave.
Port Angeles WA 98362-6798
phone:(360) 452-0303
FAX:(360) 452-0348
e-mail:Erran_Seaman@NPS.GOV
|
