Mitigating gridding artifacts question

[rwall2]

Hi there. I'm running into an issue where when I grid data I get these spikes at the farthest range gates. I doubt this is a bug, but instead an expected result of gridding data. Are there any options I can play with to remove these spikes?

This is currently what I'm doing:

    grid = pyart.map.grid_from_radars(radars,
                                      fields=["Z"],
                                      weighting_function='NEAREST',
                                      nb=1, 
                                      bsp=1, 
                                      min_radius=500,
                                      grid_shape=(200,
                                                  400,
                                                  400),
                                      grid_limits=((0,17500),
                                                   (-250000, 250000),
                                                   (-250000, 250000))
                                    )

Here "radars" is a collection of sweeps from one radar volume that I'm gridding into one grid object. I've slightly adjusted the beam width and beam spacing, as well as the minimum radius, but still get the spikes.

[zssherman]

Is radars one radar object with all the sweeps in that one radar object? The other thing that can be occurring is that further out the roi isn't large enough to capture neighboring gates for the nearest neighbor, what does upping the ROI look like?

[rwall2]

If I increase the ROI, the spikes lessen but remain visible. After some research using AI, it sounds like I should use either another weighting_function or mask the farthest range gates before gridding. The former takes too long for a real-time application I'm developing. So I opted for the latter. That solved my problem, albeit I had to remove some data. Thank you for your suggestion. I appreciate the time you took to reply.

[zssherman]

@rwall2 No worries and sounds good!