I found several questions in regard of reprojecting data into a CRS with unit of meters and what CRS to use. Usually the counter question is, "for what region, on what scale" and there the discussions stops. I have tried finding a good CRS myself, but as I am no expert on these things, this whole CRS business is extremely confusing.
So, I am looking for a good CRS with units in meters for: Central Germany, a town called Paderborn. The Region of interest is quite small, ca. 15km around Paderborn.
I create noise simulations for wind turbine parks. So, I have point data (currently in EPSG:3857) for wind turbines. Then, in a python script, I create a equidistant grid around those points, calculate the noise levels on those grid points, use contour-methods to create "isophones" (lines with identical noise level, e.g. 45dBA) and save this data as line or polygon data.
Generally, this all works very nice and gives beautiful results.
But the really dirty part is this: In my script I take the EPSG:3857 data and transform it into EPSG:4326. Then I use a self coded haversine transform to use the already transformed data into "my grid" with an origin I choose. I run the simulation, do the analysis, use again a self coded inverse haversine transform to EPSG:4326 and the transform again back to EPSG:3857.
That probably is not a good idea... Any recommendations?