Dear list,
I am an entomologist and have basic statistics knowledge, but spatial aspect is
new to me. Currently reading heavily to better understand spatial statistics.
Let me explain the issues I am experiencing with my insect counts. I have a
dataset of an insect count for two consequent years, for about 80x40km region.
Data was sampled on agricultural fields basis (100 samples from each field) and
averaged then this average value was attached to related point, in GIS. This
repeated for the following year, but this time, some of fields are not sampled,
instead, some samples were taken from nearby fields.
I want to create hotspot maps for each year. My problem is these different
points. The good thing is, over 70% of points are same fields for both years. I
am trying to find a way whether the difference among these close points are
statistically important or not. If not, I can include them too, for hotspot
tests.
When I create hotspot maps for each individual year with all points belonging
to this year (point numbers are different, say, 100 points for first year, 120
points for the next), hotspots are still appearing in same or similar locations
(with same bandwidth for each year), do not know how to indicate these
datasets/points are statistically and/or spatially not different.
To summarize, I do not want to waste samplings for points that not appear in
both years, so I need to find a way, whether two or more datasets randomly
sampled from same region are statistically different or not, by means of
geographical locations and collected data.
To obtain a uniformity, a solution may be creating a grid, (eg, as quarters of
utm rectangles, 5x5km) and averaging points of corresponding year falling into
each cell, then running hotspot analysis on this polygons, for each year. But
this time, finding out the most appropriate grid cell size/bandwidth becomes a
problem..
I appreciate if you may direct me on possible solutions.
I did my best to simplify the problem above, but may further explain if needed.
If I can handle this point selection stage, want to apply GWR to the data.
Regards,
Alper
_________________________________________________________________
Hotmail: Free, trusted and rich email service.
[[alternative HTML version deleted]]
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
