I'm having a bit of meltdown about projections and coordinate reference systems, and was hoping for a few pointers.
I have 3 GIS projects, at 3 sites in Europe (the largest of which measures about 10x30km) all using similar data.
My data layers include point features (from GPS, in lat lon), derived buffers around these points, manually created polygon features (representing landuse) which I want to convert to raster for analysis and some externally derived rasters reprenting imagery, elevation, etc. I will also have some GPS points representing animal locations.
I'm currently using ArcMAP, but also trying to migrate to R and/or QuantumGIS (in preparation for losing my Arc license). I guess my problem isn't software-specific anyway.
I want to measure distance accurately (e.g. 1000m buffer around point features; nearest-neighbour distance) as well as area (point density; percent cover of different landuse within a buffer). Currently I'm working in WGS84 as this is the system my GPS points come in. I'm wondering whether this is the best projection (though as I understand it, it is not a true projection) to be using for these small-scale analyses. I've read that it might be better to use a site-specific transverse Mecator centred on each site. Will it make much difference? (I also want to make some nice figures for presentation, so it would be good if e.g. 1km buffers were actually round rather than eliptical - maybe this is a different point though)
I realise this sort of question has been asked a million times before. I'm not being lazy, but I currently have massively restricted web access so can't afford to figure out the answer myself.
أكثر...
I have 3 GIS projects, at 3 sites in Europe (the largest of which measures about 10x30km) all using similar data.
My data layers include point features (from GPS, in lat lon), derived buffers around these points, manually created polygon features (representing landuse) which I want to convert to raster for analysis and some externally derived rasters reprenting imagery, elevation, etc. I will also have some GPS points representing animal locations.
I'm currently using ArcMAP, but also trying to migrate to R and/or QuantumGIS (in preparation for losing my Arc license). I guess my problem isn't software-specific anyway.
I want to measure distance accurately (e.g. 1000m buffer around point features; nearest-neighbour distance) as well as area (point density; percent cover of different landuse within a buffer). Currently I'm working in WGS84 as this is the system my GPS points come in. I'm wondering whether this is the best projection (though as I understand it, it is not a true projection) to be using for these small-scale analyses. I've read that it might be better to use a site-specific transverse Mecator centred on each site. Will it make much difference? (I also want to make some nice figures for presentation, so it would be good if e.g. 1km buffers were actually round rather than eliptical - maybe this is a different point though)
I realise this sort of question has been asked a million times before. I'm not being lazy, but I currently have massively restricted web access so can't afford to figure out the answer myself.
أكثر...