Minimal amount of code to regrid UKV data to OSGB grid, keeping native grid size of original UKV data. Based on Gist by @DPeterK https://gist.github.com/DPeterK/c5061f336a91a3ce9790c206a5459b4a

iris_regrid_ukv_to_osgb.py

# (C) Crown Copyright, Met Office. All rights reserved. 2020
#
# Python code written for demonstrative purposes only.
# Authored by Kevin Donkers (@kaedonkers) and Peter Killick (@DPeterK)
# See https://gist.github.com/DPeterK/c5061f336a91a3ce9790c206a5459b4a

import iris
import numpy as np
import cartopy.crs as ccrs

# --- UKV data ---
fname_ukv = './ukv_daily_t1o5m_mean_20200101.nc'
cube_ukv = iris.load_cube(fname_ukv)
crs_ukv = cube_ukv.coord_system().as_cartopy_crs()

# --- Subset UKV to extent of OSGB ---
osgb = ccrs.OSGB()
uk_extent_bl = crs_ukv.transform_point(osgb.x_limits[0], osgb.y_limits[0], osgb)
uk_extent_tr = crs_ukv.transform_point(osgb.x_limits[1], osgb.y_limits[1], osgb)
cube_ukv_uk = cube_ukv.intersection(grid_latitude=[uk_extent_bl[1], uk_extent_tr[1]],
                                    grid_longitude=[uk_extent_bl[0], uk_extent_tr[0]])

# --- OSGB grid cube ---

# Here is where you can change the grid spacing to your liking
ny, nx = cube_ukv_uk.shape
y_points = np.linspace(osgb.y_limits[0], osgb.y_limits[1], ny)
x_points = np.linspace(osgb.x_limits[0], osgb.x_limits[1], nx)

# Construct dimension coordinates
x_coord = iris.coords.DimCoord(x_points,
                               standard_name="projection_x_coordinate",
                               units='m',
                               coord_system=iris.coord_systems.OSGB())
y_coord = iris.coords.DimCoord(y_points,
                               standard_name="projection_y_coordinate",
                               units='m',
                               coord_system=iris.coord_systems.OSGB())

# Mapping of coordinates to data dims.
dcad = [(y_coord, 0), (x_coord, 1)]

# Make cube with data array of zeros
data = np.zeros(cube_ukv_uk.shape)
grid_cube = iris.cube.Cube(data,
                           units="1",
                           long_name="grid cube",
                           dim_coords_and_dims=dcad)
grid_cube.coord_system = iris.coord_systems.OSGB()
grid_cube

# --- Regrid UKV to OSGB ---
regrid_scheme = iris.analysis.Linear(extrapolation_mode='mask')
ukv_osgb = cube_ukv_uk.regrid(grid_cube, regrid_scheme)
# Save regridded cube to another NetCDF file
iris.save(ukv_osgb, 'ukv_osgb_outputfile.nc')