level II radar data projection


Recently, I plotted level II radar data for the
Northeast United States using gpnexr2, including an
overlay of a 0.1 degree by 0.1 degree latitude
longitude grid. I am looking at small convective
cells, some approximately the size of a 0.1 by 0.1
degree lat/lon box. Using these horizontal grids, I
tried to slice through these storms, creating
crosssections based on the lat/lon of the grid.
However, when creating the crosssections, I noticed
that the location of the cells do not match up exactly
with the lat/lon points. Therefore, a crosssection
(created using a lat/lon taken from my horizontal map)
that I think will cut directly through the storm will
actually just clip it (or miss it all together). Does
this have something to do with the RAD projection? Is
there anyway to correct for this?

Thanks for your time