def test_filter_gabella(self):
filename = util.get_wradlib_data_file('misc/polar_dBZ_fbg.gz')
data = np.loadtxt(filename)
clutter.filter_gabella(data,
wsize=5,
thrsnorain=0.,
tr1=6.,
n_p=8,
tr2=1.3)
def ex_clutter_gabella():
# load the example data
import numpy as np
# Todo: link right data set
testdata = np.loadtxt(os.path.dirname(__file__) + '/' + 'data/polar_dBZ_fbg.gz')
# calculate the clutter map
clmap = clutter.filter_gabella(testdata,
wsize=5,
thrsnorain=0.,
tr1=6.,
n_p=8,
tr2=1.3)
# visualize the result
ax, pm = vis.plot_ppi(clmap)
ax.set_title('cluttermap')
pl.show()
def ex_clutter_gabella():
# load the example data
import numpy as np
# Todo: link right data set
testdata = np.loadtxt(
os.path.dirname(__file__) + '/' + 'data/polar_dBZ_fbg.gz')
# calculate the clutter map
clmap = clutter.filter_gabella(testdata,
wsize=5,
thrsnorain=0.,
tr1=6.,
n_p=8,
tr2=1.3)
# visualize the result
ax, pm = vis.plot_ppi(clmap)
ax.set_title('cluttermap')
pl.show()
def test_filter_gabella(self):
filename = util.get_wradlib_data_file('misc/polar_dBZ_fbg.gz')
data = np.loadtxt(filename)
clutter.filter_gabella(data, wsize=5, thrsnorain=0., tr1=6.,
n_p=8, tr2=1.3)
def main(radar, tilt_list, ref_fieldname):
"""
Apply a clutter removal workflow to a single radar volume
===============
(1) Gabella texture filter
(2) Despeckle
Parameters:
===============
radar: pyart radar object
tilt_list: list
tilt list to process. Empty triggers processing for all tilts
Returns:
===============
radar: pyart radar object
"""
#config
rain_cut_dbz = 10.
#min_dbz = 0.
#snr_threshold = 10
#define the indices for the required sweep
sweep_startidx = radar.sweep_start_ray_index['data'][:]
sweep_endidx = radar.sweep_end_ray_index['data'][:]
refl_data = radar.fields[ref_fieldname]['data'].copy()
clutter_mask = np.zeros_like(refl_data) #clutter flag = 1, no clutter = 0
#build list of tilts to process for gabella filter
if not tilt_list:
tilt_list = np.arange(len(sweep_startidx))
else:
tilt_list = np.array(tilt_list)
#loop through sweeps
for k in tilt_list:
#extract ppi
ppi = refl_data[sweep_startidx[k]:sweep_endidx[k] + 1]
#generate clutter mask for ppi
clmap = clutter.filter_gabella(ppi,
wsize=5,
thrsnorain=rain_cut_dbz,
tr1=10.,
n_p=8,
tr2=1.3)
#insert clutter mask for ppi into volume mask
clutter_mask[sweep_startidx[k]:sweep_endidx[k] + 1] = clmap
#add clutter mask to radar object
clutter_field = {
'data': clutter_mask,
'units': 'mask',
'long_name': 'Gabella clutter mask',
'standard_name': 'CLUTTER',
'comments': 'wradlib implementation of Gabella et al., 2002'
}
radar.add_field('reflectivity_clutter',
clutter_field,
replace_existing=True)
#apply clutter filter to gatefiler
gatefilter = pyart.correct.GateFilter(radar)
gatefilter.exclude_equal('reflectivity_clutter', 1)
#generate depseckle filter
gate_range = radar.range['meters_between_gates']
#apply limits
if gate_range < 250:
print('gate range too small in clutter.py, setting to 250m')
gate_range = 250
if gate_range > 1000:
print('gate range too large in clutter.py, setting to 1000m')
gate_range = 1000
#rescale and calculate sz
despeckle_factor = 1000 / gate_range #scale despeckle according to gate size
despeckle_sz = 15 * despeckle_factor #1000m = 15, 500m = 30, 250m = 60
#apply despeckle to gatefilter
gatefilter = pyart.correct.despeckle.despeckle_field(radar,
ref_fieldname,
gatefilter=gatefilter,
size=despeckle_sz)
#apply filter to mask
cor_refl_data = np.ma.masked_where(gatefilter.gate_excluded,
refl_data.copy())
#return radar object
return cor_refl_data, gatefilter