I'm trying to compare the spatial distribution between two groups of fishing locations (for each of several different years) in the Bering Sea (lat's ~53 to 65N, long's ~ -178 to -158W. My preliminary explorations have used kernel density estimation (with an unconstrained plug-in bandwidth matrix) and then examined different overlap indices of their home ranges (quantified via utilization distribution overlap index and Bhattacharyya's Affinity) to compare similarities/differences (all performed in R).

However, I'm concerned about the interpretation of these data when the data may span more than one UTM zone. Most of my data are in UTM zone 3 with some data points seeping into zone 4. In more extreme cases, my data may span from zone 1 to 4.

Am I being really ignorant in thinking that I can compare across zone boundaries?

Since I am interested only in relative differences between calculations with the same distortion, will it not matter?

Should I simply make sure that I have my UTM origin set at the southwesterly most point of the data and I'm fine from there?

I've been reading Stack Exchange and other sites all morning and have seen several discussions about measuring across UTM zones, but I haven't been able to logically extend such responses to a method such as this one.

closed as too broad by PolyGeo Jul 26 '17 at 21:22

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • You can probably safely reproject from the edge of zone 4 into zone 3 with negligible distortion, but spanning four UTM zones requires reprojecting all the data into an appropriate projection (e.g., a custom Albers Equal Area) covering the study area. – Vince Jan 9 '15 at 18:51
  • Thanks, folks! I guess I wasn't sure if I could "simply" reproject the data to minimize the distortion of if the nature of UTM zones would preclude kernel density analysis across such an area (despite reprojection). Also, if distortion increases with distance then, would it be better to set my origin in the middle of the data, or as I suggested in the OP, to set the origin at the bottom left. – Jordan Jan 9 '15 at 19:05
  • Different projections have different parameters. The origin of the {X,Y} space doesn't matter as much as the latitude and standard parallels (if conic). The projection center should be in the center of the study area; the reference lat & lon are only used to control negative coordinate values. – Vince Jan 9 '15 at 19:19

I suggest (as Vince) to put the center of your custom projection in the middle of the study area at 168 W 59 N.

The following projections might give best results:

+proj=laea +lat_0=59 +lon_0=-168 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs
+proj=aea +lat_1=53 +lat_2=65 +lat_0=59 +lon_0=-168 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs
+proj=tmerc +lat_0=59 +lon_0=-168 +k=1 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs

They look pretty much the same (with the UTM zones as reference):

enter image description here

Not the answer you're looking for? Browse other questions tagged or ask your own question.