I start with the following Raster:
class : RasterLayer
dimensions : 900, 900, 810000 (nrow, ncol, ncell)
resolution : 1, 1 (x, y)
extent : -523.4622, 376.5378, -4658.645, -3758.645 (xmin, xmax, ymin, ymax)
crs : +proj=stere +lat_0=90 +lon_0=10 +k=0.93301270189 +x_0=0 +y_0=0 +ellps=WGS84 +units=km +no_defs
source : memory
names : layer
values : 0, 2500 (min, max)
If i reproject this raster to latlon with:
raster_projection <- projectRaster(rb, crs = "+proj=longlat +datum=WGS84 +no_defs")
it creates a reprojection with significant shift in data:
class : RasterLayer
dimensions : 944, 1014, 957216 (nrow, ncol, ncell)
resolution : 0.0136, 0.00857 (x, y)
extent : 2.00348, 15.79388, 46.99609, 55.08617 (xmin, xmax, ymin, ymax)
crs : +proj=longlat +datum=WGS84 +no_defs
source : memory
names : layer
values : -0.6159579, 3583.708 (min, max)
I think the data points are so to say: "too far up".
Is this to be expected?
How can i create a reprojection that is more true to original ?
Thank you!
The new extreme values are likely at the edges; and so your problem may be caused by a few cells. A more appropriate comparison could be to look at a boxplot or quantiles. You could also use terra::project
instead. I think that method is less prone to this.