def func4():
"""
Test (lat, lt) to (mlat, mlt) conversions.
1, Use a = Apex(date=...); "date" determines which IGRF coefficients are
used in conversions. Uses current date as default.
2, height is needed for better conversion.
champ and grace files use qd coordinates.
3, mlt in qd and apex coordinates are the same.
"""
import champ_grace as cg
from apexpy import Apex as Apex
import matplotlib.pyplot as plt
a = cg.ChampDensity('2005-1-1', '2005-1-2')
b = Apex(date=2005)
mlatt, mlt = b.convert(
lat=a.lat, lon=a.long, source='geo', dest='mlt',
datetime=a.index, height=a.height)
mlat, mlongt = b.convert(
lat=a.lat, lon=a.long, source='geo', dest='qd',
height=a.height)
mlat2 = np.array(a.Mlat)
mlt2 = np.array(a.MLT)
plt.plot(mlat-mlat2)
plt.plot(abs(mlt-mlt2) % 24, '.')
plt.show()
return
def test_convert_invalid_lat():
A = Apex(date=2000, refh=300)
with pytest.raises(ValueError):
A.convert(91, 0, 'geo', 'geo')
with pytest.raises(ValueError):
A.convert(-91, 0, 'geo', 'geo')
A.convert(90, 0, 'geo', 'geo')
A.convert(-90, 0, 'geo', 'geo')
assert_allclose(A.convert(90+1e-5, 0, 'geo', 'apex'), A.convert(90, 0, 'geo', 'apex'), rtol=0, atol=1e-8)
def plot_ion_drift(show=True,save=True):
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
g = g2a
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
for ialt, alt in enumerate([130, 200, 300, 400, 500, 600]):
alt_ind = np.argmin(np.abs(g['Altitude'][0, 0, :]/1000-alt))
alt_str = '%6.2f' % (g['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
3, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g, 'polar', useLT=True),
coastlines=False)
lon0, lat0, ewind, nwind = g3ca.vector_data(g, 'ion', alt=alt)
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection)
hq = ax.quiver(
lon0, lat0, ewind, nwind, scale=1500, scale_units='inches',
regrid_shape=20, headwidth=5)
ax.quiverkey(hq, 0.93, -0.1, 1000, '1000 m/s')
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+tstring, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
if show:
plt.show()
if save:
plt.savefig(spath+'03_ion_drift_%s%s.pdf' % (tstrday,tstring))
return
def plot_temperature():
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
g = g2a
alts = [130, 400]
for ialt, alt in enumerate(alts):
alt_ind = np.argmin(np.abs(g['Altitude'][0, 0, 2:-2]/1000-alt))+2
alt_str = '%6.2f' % (g['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
1, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g, 'polar', useLT=True),
coastlines=False)
lon0, lat0, zdata0 = g3ca.contour_data('Temperature', g, alt=alt)
fp = (lat0[:,0]>slat) & (lat0[:,0]<nlat)
lon0, lat0, zdata0 = lon0[fp, :], lat0[fp,:], zdata0[fp,:]
hc = ax.contourf(lon0, lat0, zdata0, 21,
transform=ccrs.PlateCarree(), cmap='jet',
extend='both')
hc = plt.colorbar(hc, pad=0.17)
hc.set_label('Temperature (K)')
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+titletime, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
plt.show()
return
def satellite_position_lt_lat(self, mag=False, ns='N'):
""" Show the lt and lat positions of the satellite in a polar
coordinate.
Input:
mag: if True, for MLT and Mlat position
ns: N or S for North and South hemispheres, respectively
Output:
hcup, hcdown: scatter handles for the up and down orbits,
respectively.
"""
if self.empty:
return
lt = self['LT']
lat = self['lat']
if mag:
from apexpy import Apex
gm = Apex()
mlat,mlt = gm.convert(
self['lat'], self['long'], 'geo', 'mlt', datetime=self.index)
lt, lat = mlt, mlat
ct = lat>0 if ns is 'N' else lat<0
theta = lt[ct]/12*np.pi
r = 90 - abs(lat[ct])
hc = plt.scatter(theta, r, linewidths=0)
return hc
def plot_vert_vgradrho_rho_diff(show=True, save=True):
rho1 = np.array(g1a['Rho'])
vgradrho1 = \
g1a['V!Dn!N (up)'] \
* cr.calc_rusanov_alts_ausm(g1a['Altitude'],rho1)/g1a['Rho']
rho2 = np.array(g2a['Rho'])
vgradrho2 = \
g2a['V!Dn!N (up)']\
* cr.calc_rusanov_alts_ausm(g2a['Altitude'],rho2)/g2a['Rho']
g1a['vgradrho_diff'] = vgradrho1-vgradrho2
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
for ialt, alt in enumerate([130, 200, 300, 400, 500, 600]):
alt_ind = np.argmin(np.abs(g1a['Altitude'][0, 0, :]/1000-alt))
alt_str = '%6.2f' % (g1a['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
3, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g1a, 'polar', useLT=True),
coastlines=False)
lon0, lat0, zdata0 = g3ca.contour_data('vgradrho_diff', g1a, alt=alt)
fp = (lat0[:,0]>slat) & (lat0[:,0]<nlat)
lon0, lat0, zdata0 = lon0[fp, :], lat0[fp,:], zdata0[fp,:]
hc = ax.contourf(
lon0, lat0, zdata0, np.linspace(-1,1,21)*1e-4,
transform=ccrs.PlateCarree(), cmap='seismic', extend='both')
hcb = plt.colorbar(hc, pad=0.17)
hcb.formatter.set_powerlimits((0,0))
hcb.update_ticks()
hcb.set_label(r'$-\vec{u}\cdot\frac{\nabla\rho}{\rho}$ (up)')
# wind difference
lon1, lat1, ewind1, nwind1 = g3ca.vector_data(g1a, 'neutral', alt=alt)
lon2, lat2, ewind2, nwind2 = g3ca.vector_data(g2a, 'neutral', alt=alt)
lon0, lat0 = lon1, lat1
lon0, lat0, ewind0, nwind0 = g3ca.convert_vector(
lon0, lat0, ewind2-ewind1, nwind2-nwind1, plot_type='polar',
projection=projection)
hq = ax.quiver(
lon0, lat0, ewind0, nwind0, scale=1000, scale_units='inches',
regrid_shape=20, headwidth=5)
ax.quiverkey(hq, 0.93, -0.1, 500, '500 m/s')
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+tstring, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
if show:
plt.show()
if save:
plt.savefig(spath+'06_vert_vgradrho_rho_diff_%s%s.pdf' % (tstrday,tstring))
return
def test_convert_mlt2apex():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(
A.convert(60,
15,
'mlt',
'apex',
height=100,
datetime=datetime,
ssheight=2e5), (60, A.mlt2mlon(15, datetime, ssheight=2e5)))
def test_convert_qd2mlt():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(
A.convert(60,
15,
'qd',
'mlt',
height=100,
datetime=datetime,
ssheight=2e5)[1], A.mlon2mlt(15, datetime, ssheight=2e5))
def plot_temperature_diff(show=True, save=True):
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
for ialt, alt in enumerate([130, 200, 300, 400, 500, 600]):
alt_ind = np.argmin(np.abs(g1a['Altitude'][0, 0, :]/1000-alt))
alt_str = '%6.2f' % (g1a['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
3, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g1a, 'polar', useLT=True),
coastlines=False)
# temperature diff
lon1, lat1, zdata1 = g3ca.contour_data('Temperature', g1a, alt=alt)
lon2, lat2, zdata2 = g3ca.contour_data('Temperature', g2a, alt=alt)
fp = (lat1[:,0]>slat) & (lat1[:,0]<nlat)
lon0, lat0, zdata0 = lon1[fp, :], lat1[fp,:], (zdata2-zdata1)[fp,:]
hc = ax.contourf(
lon0, lat0, zdata0, np.linspace(-80,80,21),
transform=ccrs.PlateCarree(), cmap='seismic', extend='both')
hc = plt.colorbar(hc, pad=0.17)
hc.set_label(r'$T_2-T_1$ (K)')
# geomagnetic pole
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
# wind difference
lon1, lat1, ewind1, nwind1 = g3ca.vector_data(g1a, 'neutral', alt=alt)
lon2, lat2, ewind2, nwind2 = g3ca.vector_data(g2a, 'neutral', alt=alt)
lon0, lat0 = lon1, lat1
lon0, lat0, ewind0, nwind0 = g3ca.convert_vector(
lon0, lat0, ewind2-ewind1, nwind2-nwind1, plot_type='polar',
projection=projection)
hq = ax.quiver(
lon0, lat0, ewind0, nwind0, scale=1000, scale_units='inches',
regrid_shape=20, headwidth=5)
ax.quiverkey(hq, 0.93, -0.1, 500, '500 m/s')
# rho difference
#lon3, lat3, zdata3 = g3ca.contour_data('Rho', g1a, alt=alt)
#lon4, lat4, zdata4 = g3ca.contour_data('Rho', g2a, alt=alt)
#diffrho = 100*(zdata4-zdata3)/zdata3
#hc = ax.contour(
# lon4, lat4, diffrho, [-10], transform=ccrs.PlateCarree(),
# colors='g',linestyles='-')
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+tstring, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
if show:
plt.show()
if save:
plt.savefig(spath+'02_temperature_diff_%s%s.pdf' % (tstrday,tstring))
return
def test_convert_geo2mlt():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(
A.convert(60,
15,
'geo',
'mlt',
height=100,
ssheight=2e5,
datetime=datetime)[1],
A.mlon2mlt(A.geo2apex(60, 15, 100)[1], datetime, ssheight=2e5))
def test_convert_apex2mlt():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
apex_out = Apex(date=2000, refh=300)
assert_allclose(
apex_out.convert(60,
15,
'apex',
'mlt',
height=100,
datetime=datetime,
ssheight=2e5)[1],
apex_out.mlon2mlt(15, datetime, ssheight=2e5))
def satellite_position_lt_lat(self,
ax,
nlat=90,
slat=0,
mag=False,
plot_type='polar',
*args,
**kwargs):
""" Show the (M)lt and (M)lat positions of the satellite in a polar
or rectangular coordinate axis.
Input:
ax: which axis to draw orbit on
nlat: northward latitude limit (default 90)
slat: southward latitude limit (default 0)
mag: if True, for MLT and Mlat position
Return:
hc: handle of the scatter
"""
if self.empty:
return
tmp = self
if mag:
from apexpy import Apex
import datetime as dt
a = Apex(date=self.index.year.mean())
mlat, mlt = a.convert(tmp.lat,
tmp.long,
'geo',
'mlt',
height=tmp.height,
datetime=tmp.index)
tmp['MLT'] = mlt
tmp['Mlat'] = mlat
ltp = 'MLT' if mag else 'LT'
latp = 'Mlat' if mag else 'lat'
ct = (self[latp] > slat) & (self[latp] < nlat)
if 'pol' in plot_type.lower():
if nlat * slat < 0:
print('Error: For polar plot, nlat and slat'
' should have the same signs')
csign = 1 if nlat > 0 else -1
theta = tmp.loc[ct, ltp] / 12 * np.pi
r = 90 - csign * tmp.loc[ct, latp]
hc = ax.scatter(theta, r, linewidths=0, *args, **kwargs)
if 'rec' in plot_type.lower():
hc = ax.scatter(tmp.loc[ct, ltp],
tmp.loc[ct, latp],
linewidths=0,
*args,
**kwargs)
return hc
def pole_unit_vector(t):
"""credit: Eelco Doornbos"""
apexdate = t.year + t.dayofyear / 365 # routine needs data as for example 2015.3
A = Apex(date=apexdate)
glat, glon = A.convert(90, 0, 'apex', 'geo', height=0)
colatrad = radians(90.0 - glat)
lonrad = radians(glon)
return np.array([
sin(colatrad) * cos(lonrad),
sin(colatrad) * sin(lonrad),
cos(colatrad)
])
def plot_den_win(show=True, save=True):
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
g = g2a
for ialt, alt in enumerate([130, 200, 300, 400, 500, 600]):
alt_ind = np.argmin(np.abs(g['Altitude'][0, 0, 2:-2]/1000-alt))+2
alt_str = '%6.2f' % (g['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
3, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g, 'polar', useLT=True),
coastlines=False)
# density
lon0, lat0, zdata0 = g3ca.contour_data('Rho', g, alt=alt)
fp = (lat0[:,0]>slat) & (lat0[:,0]<nlat)
lon0, lat0, zdata0 = lon0[fp, :], lat0[fp,:], zdata0[fp,:]
hc = ax.contourf(lon0, lat0, zdata0, 21,
transform=ccrs.PlateCarree(), cmap='jet',
extend='both')
hc = plt.colorbar(hc, pad=0.17)
hc.set_label(r'$\rho$ (kg/m$^3$)')
# wind
lon0, lat0, ewind, nwind = g3ca.vector_data(g, 'neutral', alt=alt)
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection)
hq = ax.quiver(
lon0, lat0, ewind, nwind, scale=1500, scale_units='inches',
regrid_shape=20)
ax.quiverkey(hq, 0.93, -0.1, 1000, '1000 m/s')
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
# rho difference
# lon3, lat3, zdata3 = g3ca.contour_data('Rho', g1a, alt=alt)
# lon4, lat4, zdata4 = g3ca.contour_data('Rho', g2a, alt=alt)
# diffrho = 100*(zdata4-zdata3)/zdata3
# hc = ax.contour(
# lon4, lat4, diffrho, [-10], transform=ccrs.PlateCarree(),
# colors='g',linestyles='-')
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+tstring, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
if show:
plt.show()
if save:
plt.savefig(spath+'01_den_win_run2_%s%s.pdf' %(tstrday,tstring))
return
def polar_quiver_wind(self, ax, ns='N'):
# Wind vector in lat-long coordinates.
# For different map projections, the arithmetics to calculate xywind
# are different
if self.empty:
return
from mpl_toolkits.basemap import Basemap
from apexpy import Apex
# Creat polar coordinates
projection,fc = ('npstere',1) if ns=='N' else ('spstere',-1)
m = Basemap(projection=projection,boundinglat=fc*40,lon_0=0,resolution='l')
m.drawcoastlines(color='gray',zorder=1)
m.fillcontinents(color='lightgray',zorder=0)
dt = self.index.min() + (self.index.max()-self.index.min())/2
m.nightshade(dt,zorder=2)
#m.drawparallels(np.arange(-80,81,20))
#m.drawmeridians(np.arange(-180,181,60),labels=[1,1,1,1])
# Calculate mlat and mlon
lat_grid = np.arange(-90,91,10)
lon_grid = np.arange(-180,181,10)
lon_grid, lat_grid = np.meshgrid(lon_grid, lat_grid)
gm = Apex(date=2005)
mlat,mlon = gm.convert(lat_grid,lon_grid,'geo','apex')
hc1 = m.contour(lon_grid,lat_grid,mlat,levels=np.arange(-90,91,10),
colors='k', zorder=3, linestyles='dashed',
linewidths=1, latlon=True)
# hc2 = m.contour(lon_grid,lat_grid,mlon,levels=np.arange(-180,181,45),
# colors='k', zorder=3, linestyles='dashed', latlon=True)
plt.clabel(hc1,inline=True,colors='k',fmt='%d')
# plt.clabel(hc2,inline=True,colors='k',fmt='%d')
# Calculate and plot x and y winds
lat = self.lat
lon = self.long
wind = self.wind
winde1 = self.winde
winde = winde1*wind
windn1 = self.windn
windn = windn1*wind
# only appropriate for the npstere and spstere
xwind = fc*winde*np.cos(lon/180*np.pi)-windn*np.sin(lon/180*np.pi)
ywind = winde*np.sin(lon/180*np.pi)+fc*windn*np.cos(lon/180*np.pi)
hq = m.quiver(np.array(lon),np.array(lat),xwind,ywind,color='blue',
scale=300, scale_units='inches',zorder=3, latlon=True)
#plt.quiverkey(hq,1.05,1.05,100,'100 m/s',coordinates='axes',labelpos='E')
#m.scatter(np.array(lon),np.array(lat),
# s=50, c=self.index.to_julian_date(),linewidths=0, zorder=4,latlon=True)
return m
def test_convert_mlt2geo():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
apex_out = Apex(date=2000, refh=300)
assert_allclose(
apex_out.convert(60,
15,
'mlt',
'geo',
height=100,
datetime=datetime,
precision=1e-2,
ssheight=2e5),
apex_out.apex2geo(60,
apex_out.mlt2mlon(15, datetime, ssheight=2e5),
100,
precision=1e-2)[:-1])
def mag_parallels(date, parallels=range(-75, 76, 15), height=350, N=1000):
"""
Return a mapping between magnetic latitudes specified by
*parallels* to the tuple of mapped geographic latitudes and
longitudes. The mapping is made across *N* uniformly spaced
geographic longitudes, on :class:`datetime` *date*, and at
*height* (in [km]) in apex geomagnetic coordinates. If *date* is
None, use the current date and time in the coordinate
transformation.
"""
apex = Apex(date=date)
parallel_map = OrderedDict()
lons = NP.linspace(-180, 180, N)
for parallel in parallels:
glat, glon = apex.convert(parallel, lons, source="apex", dest="geo")
# sort by geographic longitude
glat, glon = zip(*sorted(zip(glat, glon), key=lambda x: x[1]))
parallel_map[parallel] = glat, glon
return parallel_map
def mag_parallels(date, parallels=range(-75, 76, 15), height=350, N=1000):
"""
Return a mapping between magnetic latitudes specified by
*parallels* to the tuple of mapped geographic latitudes and
longitudes. The mapping is made across *N* uniformly spaced
geographic longitudes, on :class:`datetime` *date*, and at
*height* (in [km]) in apex geomagnetic coordinates. If *date* is
None, use the current date and time in the coordinate
transformation.
"""
apex = Apex(date=date)
parallel_map = OrderedDict()
lons = NP.linspace(-180, 180, N)
for parallel in parallels:
glat, glon = apex.convert(parallel, lons, source='apex', dest='geo')
# sort by geographic longitude
glat, glon = zip(*sorted(zip(glat, glon), key=lambda x: x[1]))
parallel_map[parallel] = glat, glon
return parallel_map
def get_model(tec_data, hemisphere, omni_file):
"""Get magnetic latitudes of the trough according to the model in Deminov 2017
for a specific time and set of magnetic local times.
"""
logger.info("getting model")
omni_data = xr.open_dataset(omni_file)
logger.info(f"{omni_data.time.values[0]=} {omni_data.time.values[-1]=}")
kp = _get_weighted_kp(tec_data.time, omni_data)
logger.info(f"{kp.shape=}")
apex = Apex(date=utils.datetime64_to_datetime(tec_data.time.values[0]))
mlat = 65.5 * np.ones((tec_data.time.shape[0], tec_data.mlt.shape[0]))
for i in range(10):
glat, glon = apex.convert(mlat, tec_data.mlt.values[None, :], 'mlt', 'geo', 350, tec_data.time.values[:, None])
mlat = _model_subroutine_lat(tec_data.mlt.values[None, :], glon, kp[:, None], hemisphere)
if hemisphere == 'south':
mlat = mlat * -1
tec_data['model'] = xr.DataArray(
mlat,
coords={'time': tec_data.time, 'mlt': tec_data.mlt},
dims=['time', 'mlt']
)
def satellite_position_lt_lat(self, ax, nlat=90, slat=0,
mag=False, plot_type='polar',
*args, **kwargs):
""" Show the (M)lt and (M)lat positions of the satellite in a polar
or rectangular coordinate axis.
Input:
ax: which axis to draw orbit on
nlat: northward latitude limit (default 90)
slat: southward latitude limit (default 0)
mag: if True, for MLT and Mlat position
Return:
hc: handle of the scatter
"""
if self.empty:
return
tmp = self
if mag:
from apexpy import Apex
import datetime as dt
a = Apex(date=self.index.year.mean())
mlat, mlt = a.convert(tmp.lat, tmp.long, 'geo','mlt',
height=tmp.height, datetime=tmp.index)
tmp['MLT'] = mlt
tmp['Mlat'] = mlat
ltp='MLT' if mag else 'LT'
latp='Mlat' if mag else 'lat'
ct = (self[latp]>slat) & (self[latp]<nlat)
if 'pol' in plot_type.lower():
if nlat*slat<0:
print('Error: For polar plot, nlat and slat'
' should have the same signs')
csign = 1 if nlat>0 else -1
theta = tmp.loc[ct,ltp]/12*np.pi
r = 90-csign*tmp.loc[ct,latp]
hc = ax.scatter(theta, r, linewidths=0, *args, **kwargs)
if 'rec' in plot_type.lower():
hc = ax.scatter(tmp.loc[ct, ltp], tmp.loc[ct, latp], linewidths=0,
*args, **kwargs)
return hc
def plot_den_win():
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
g = g2a
alts = [130, 400]
for ialt, alt in enumerate(alts):
alt_ind = np.argmin(np.abs(g['Altitude'][0, 0, 2:-2]/1000-alt))+2
alt_str = '%6.2f' % (g['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
1, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g, 'polar', useLT=True),
coastlines=False)
# density
lon0, lat0, zdata0 = g3ca.contour_data('Rho', g, alt=alt)
fp = (lat0[:,0]>slat) & (lat0[:,0]<nlat)
lon0, lat0, zdata0 = lon0[fp, :], lat0[fp,:], zdata0[fp,:]
hc = ax.contourf(lon0, lat0, zdata0, 21,
transform=ccrs.PlateCarree(), cmap='jet',
extend='both')
hc = plt.colorbar(hc, pad=0.17)
hc.set_label(r'$\rho$ (kg/m$^3$)')
# wind
lon0, lat0, ewind, nwind = g3ca.vector_data(g, 'neutral', alt=alt)
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection)
hq = ax.quiver(
lon0, lat0, ewind, nwind, scale=1500, scale_units='inches',
regrid_shape=20)
ax.quiverkey(hq, 0.93, -0.1, 1000, '1000 m/s')
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+titletime, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
plt.show()
return
def test_convert_invalid_transformation():
A = Apex(date=2000, refh=300)
with pytest.raises(NotImplementedError):
A.convert(0, 0, 'foobar', 'geo')
with pytest.raises(NotImplementedError):
A.convert(0, 0, 'geo', 'foobar')
def test_convert_mlt2qd():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'mlt', 'qd', height=100, datetime=datetime, ssheight=2e5), A.apex2qd(60, A.mlt2mlon(15, datetime, ssheight=2e5), height=100))
def test_convert_mlt2geo():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'mlt', 'geo', height=100, datetime=datetime, precision=1e-2, ssheight=2e5), A.apex2geo(60, A.mlt2mlon(15, datetime, ssheight=2e5), 100, precision=1e-2)[:-1])
def test_convert_qd2mlt():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'qd', 'mlt', height=100, datetime=datetime, ssheight=2e5)[1], A.mlon2mlt(15, datetime, ssheight=2e5))
def test_convert_geo2mlt():
datetime = dt.datetime(2000, 3, 9, 14, 25, 58)
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'geo', 'mlt', height=100, ssheight=2e5, datetime=datetime)[1], A.mlon2mlt(A.geo2apex(60, 15, 100)[1], datetime, ssheight=2e5))
def test_convert_geo2mlt_nodate():
A = Apex(date=2000, refh=300)
with pytest.raises(ValueError):
A.convert(60, 15, 'geo', 'mlt')
def test_convert_geo2qd():
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'geo', 'qd', height=100), A.geo2qd(60, 15, 100))
def plot_time_constant(show=False, f107=150, which_alt=600):
stime1 = pd.Timestamp('2003-03-22 00:10:00')
etime1 = pd.Timestamp('2003-03-22 06:00:00')
stime2 = pd.Timestamp('2003-03-23 00:10:00')
etime2 = pd.Timestamp('2003-03-23 06:00:00')
timeidx1 = pd.DatetimeIndex(start=stime1, end=etime1, freq='10min')
timeidx2 = pd.DatetimeIndex(start=stime2, end=etime2, freq='10min')
if f107==150:
fp1 = '/media/guod/wd2t/simulation_output/momentum_analysis/'\
+ 'run_no_shrink_iondrift_4_1/data/'
fn1 = [glob.glob(fp1+'3DALL_t'+k.strftime('%y%m%d_%H%M')+'*.bin')[0]
for k in timeidx1]
fp2 = '/media/guod/wd2t/simulation_output/momentum_analysis/'\
+ 'run_no_shrink_iondrift_4_all_day/data/'
fn2 = [glob.glob(fp2+'3DALL_t'+k.strftime('%y%m%d_%H%M')+'*.bin')[0]
for k in timeidx2]
if which_alt==200:
rholevels=np.linspace(1.8, 2.6, 21)*1e-10
if which_alt==400:
rholevels=np.linspace(2.7, 10, 21)*1e-12
if which_alt==600:
rholevels=np.linspace(0.1, 0.9, 21)*1e-12
else:
fp1 = '/home/guod/simulation_output/momentum_analysis/'\
+ 'run_shrink_70_continue/data/'
fn1 = [glob.glob(fp1+'3DALL_t'+k.strftime('%y%m%d_%H%M')+'*.bin')[0]
for k in timeidx]
fp2 = '/home/guod/simulation_output/momentum_analysis/'\
+ 'run_no_shrink_70/data/'
fn2 = [glob.glob(fp2+'3DALL_t'+k.strftime('%y%m%d_%H%M')+'*.bin')[0]
for k in timeidx]
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
fig = plt.figure(figsize=[8,9])
def animate_den_wind(i):
g1, g2 = [gitm.GitmBin(k[i]) for k in [fn1, fn2]]
# create axis
ax = list(range(6))
projection = ax.copy()
for ins in range(2):
nlat, slat = [90, 40] if ins==0 else [-40, -90]
for irun in range(3):
ax[ins+irun*2], projection[ins+irun*2] = gcc.create_map(
3, 2, 1+ins+irun*2, 'polar', nlat=nlat, slat=slat,
dlat=10, centrallon=g3ca.calculate_centrallon(
g1, 'polar', useLT=True),
coastlines=False)
# Density
lon1, lat1, zdata1 = g3ca.contour_data('Rho', g1, alt=which_alt)
lon2, lat2, zdata2 = g3ca.contour_data('Rho', g2, alt=which_alt)
hc = [ax[k].contourf(
lon1, lat1, zdata1, transform=ccrs.PlateCarree(),
levels=rholevels,
cmap='jet', extend='both') for k in [0, 1]]
ax[0].set_title(g1['time'].strftime('%d-%b-%y %H:%M')+' UT', y=1.05)
ax[1].set_title(g1['time'].strftime('%d-%b-%y %H:%M')+' UT', y=1.05)
hc = [ax[k].contourf(
lon2, lat2, zdata2, transform=ccrs.PlateCarree(),
levels=rholevels,
cmap='jet', extend='both') for k in [2, 3]]
# diff density
diffzdata = 100*(zdata2-zdata1)/zdata1
hc = [ax[k].contourf(
lon2, lat2, diffzdata, 21, transform=ccrs.PlateCarree(),
levels=np.linspace(-30, 30, 21), cmap='seismic',
extend='both') for k in [4, 5]]
hc = [ax[k].contour(
lon2, lat2, diffzdata, [-10], transform=ccrs.PlateCarree(),
colors='g',linestyles='-') for k in [4, 5]]
# wind
lon1, lat1, ewind1, nwind1 = \
g3ca.vector_data(g1, 'neutral', alt=which_alt)
lon2, lat2, ewind2, nwind2 = \
g3ca.vector_data(g2, 'neutral', alt=which_alt)
for iax in range(6):
if iax == 0 or iax == 1:
lon0, lat0, ewind, nwind = (
lon1.copy(), lat1.copy(), ewind1.copy(), nwind1.copy())
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection[iax])
elif iax == 2 or iax == 3:
lon0, lat0, ewind, nwind = (
lon2.copy(), lat2.copy(), ewind2.copy(), nwind2.copy())
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection[iax])
elif iax == 4 or iax == 5:
lon0, lat0, ewind, nwind = (
lon1.copy(), lat1.copy(), ewind2-ewind1, nwind2-nwind1)
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection[iax])
hq = ax[iax].quiver(
lon0, lat0, ewind, nwind, scale=1500, scale_units='inches',
color='k', regrid_shape=20)
ax[iax].scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
# ax.quiverkey(hq, 0.93, 0, 1000, '1000 m/s')
# hc = plt.colorbar(hc, ticks=np.arange(3, 7)*1e-12)
# hc.set_label(r'$\rho$ (kg/m$^3$)')
# ax[iax].scatter(
# qlonn, qlatn, color='k', transform=ccrs.PlateCarree())
# ax[iax].scatter(
# qlons, qlats, color='k', transform=ccrs.PlateCarree())
return
anim = animation.FuncAnimation(
fig, animate_den_wind, frames=len(fn1))
Writer = animation.writers['ffmpeg']
writer = Writer(fps=15, bitrate=1800)
anim.save(path+'02_time_const_%d_%d.wmv' % (f107, which_alt),writer=writer)
return
def test_convert_qd2geo():
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'qd', 'geo', height=100, precision=1e-2), A.qd2geo(60, 15, 100, precision=1e-2)[:-1])
def test_convert_qd2apex():
A = Apex(date=2000, refh=300)
assert_allclose(A.convert(60, 15, 'qd', 'apex', height=100), A.qd2apex(60, 15, height=100))
def plot_animation_den_win(show=False):
stime = pd.Timestamp('2003-03-22 00:00:00')
etime = pd.Timestamp('2003-03-22 06:00:00')
timeidx = pd.DatetimeIndex(start=stime, end=etime, freq='5min')
fp1 = '/home/guod/simulation_output/momentum_analysis/'\
+ 'run_shrink_iondrift_4_c1/data/'
fn1 = [glob.glob(fp1+'3DALL_t'+k.strftime('%y%m%d_%H%M')+'*.bin')[0]
for k in timeidx]
fp2 = '/home/guod/simulation_output/momentum_analysis/'\
+ 'run_no_shrink_iondrift_4_1/data/'
fn2 = [glob.glob(fp2+'3DALL_t'+k.strftime('%y%m%d_%H%M')+'*.bin')[0]
for k in timeidx]
alts = [130, 300, 600]
rholevels = [
np.linspace(4, 8.5, 21)*1e-9,
np.linspace(1.35,3.5, 21)*1e-11,
np.linspace(0.1,0.9,21)*1e-12]
nlat, slat = -40, -90
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
fig = plt.figure(figsize=[8,9])
def animate_den_wind(i):
g1, g2 = [gitm.GitmBin(k[i]) for k in [fn1, fn2]]
# create axis
ax = [1,2,3,4,5,6,7,8,9]
projection = ax.copy()
for ialts in range(3):
for ird in range(2): # run2 or diff
centrallon=g3ca.calculate_centrallon(g1, 'polar', useLT=True)
ax[ialts*2+ird], projection[ialts*2+ird] = gcc.create_map(
3, 2, 1+ialts*2+ird, 'polar', nlat=nlat, slat=slat,
dlat=10, centrallon=centrallon, coastlines=False)
ax[ialts*2+ird].scatter(
qlon, qlat, color='k', transform=ccrs.PlateCarree())
ax[0].set_title(g1['time'].strftime('%d-%b-%y %H:%M')+' UT', y=1.05)
ax[1].set_title(g1['time'].strftime('%d-%b-%y %H:%M')+' UT', y=1.05)
#ax[2].set_title(g1['time'].strftime('%d-%b-%y %H:%M')+' UT', y=1.05)
# Density and diff
for ialt, alt in enumerate(alts):
lon1, lat1, zdata1 = g3ca.contour_data('Rho', g1, alt=alt)
lon2, lat2, zdata2 = g3ca.contour_data('Rho', g2, alt=alt)
diffzdata = 100*(zdata2-zdata1)/zdata1
hc = ax[ialt*2].contourf(
lon1, lat1, zdata2, transform=ccrs.PlateCarree(),
levels=rholevels[ialt], cmap='jet', extend='both')
hc = ax[ialt*2+1].contourf(
lon2, lat2, diffzdata, transform=ccrs.PlateCarree(),
levels=np.linspace(-20, 20, 21),
cmap='seismic', extend='both')
# density change rate
# for ialt, alt in enumerate(alts):
# lon1 = np.array(g1['Longitude'])
# lat1 = np.array(g1['Latitude'])
# alt1 = np.array(g1['Altitude'])
# Re = 6371*1000 # Earth radius, unit: m
# RR = Re+alt1
# omega = 2*np.pi/(24*3600)
# rho1 = np.array(g1['Rho'])
# nwind1 = np.array(g1['V!Dn!N (north)'])
# ewind1 = np.array(g1['V!Dn!N (east)']) + omega*RR*np.cos(lat1)
# uwind1 = np.array(g1['V!Dn!N (up)'])
# div_rhov1 = \
# (cr.calc_div_hozt(lon1, lat1, alt1, rho1*nwind1, rho1*ewind1)\
# +cr.calc_div_vert(alt1, rho1*uwind1))/rho1
# lon2 = np.array(g2['Longitude'])
# lat2 = np.array(g2['Latitude'])
# alt2 = np.array(g2['Altitude'])
# Re = 6371*1000 # Earth radius, unit: m
# RR = Re+alt2
# omega = 2*np.pi/(24*3600)
# rho2 = np.array(g2['Rho'])
# nwind2 = np.array(g2['V!Dn!N (north)'])
# ewind2 = np.array(g2['V!Dn!N (east)']) + omega*RR*np.cos(lat2)
# uwind2 = np.array(g2['V!Dn!N (up)'])
# div_rhov2 = \
# (cr.calc_div_hozt(lon2, lat2, alt2, rho2*nwind2, rho2*ewind2)\
# +cr.calc_div_vert(alt2, rho2*uwind2))/rho2
# g1['divrhov_diff'] = div_rhov1-div_rhov2
# lon2, lat2, zdata2 = g3ca.contour_data('divrhov_diff', g1, alt=alt)
# hc = ax[ialt*3+2].contourf(
# lon2, lat2, zdata2, transform=ccrs.PlateCarree(),
# levels=np.linspace(-1,1,21)*1e-4, cmap='seismic', extend='both')
# wind
for ialt, alt in enumerate(alts):
lon1, lat1, ewind1, nwind1 = \
g3ca.vector_data(g1, 'neutral', alt=alt)
lon2, lat2, ewind2, nwind2 = \
g3ca.vector_data(g2, 'neutral', alt=alt)
lon0, lat0, ewind, nwind = (
lon2.copy(), lat2.copy(), ewind2.copy(), nwind2.copy())
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection[ialt*2])
hq = ax[ialt*2].quiver(
lon0, lat0, ewind, nwind, scale=1500, scale_units='inches',
color='k', regrid_shape=20)
lon0, lat0, ewind, nwind = (
lon1.copy(), lat1.copy(), ewind2-ewind1, nwind2-nwind1)
lon0, lat0, ewind, nwind = g3ca.convert_vector(
lon0, lat0, ewind, nwind, plot_type='polar',
projection=projection[ialt*2+1])
hq = ax[ialt*2+1].quiver(
lon0, lat0, ewind, nwind, scale=1500, scale_units='inches',
color='k', regrid_shape=20)
return
anim = animation.FuncAnimation(fig, animate_den_wind, frames=len(fn1))
writera = animation.writers['ffmpeg']
writerb = writera(fps=15, bitrate=1800)
anim.save(path+'01_den_wind.wmv' ,writer=writerb)
return
def plot_divrhov_rho_diff(show=True, save=True):
# density change (shrink)
lon1 = np.array(g1a['Longitude'])
lat1 = np.array(g1a['Latitude'])
alt1 = np.array(g1a['Altitude'])
Re = 6371*1000 # Earth radius, unit: m
RR = Re+alt1
omega = 2*np.pi/(24*3600)
rho1 = np.array(g1a['Rho'])
nwind1 = np.array(g1a['V!Dn!N (north)'])
ewind1 = np.array(g1a['V!Dn!N (east)']) + omega*RR*np.cos(lat1)
uwind1 = np.array(g1a['V!Dn!N (up)'])
div_rhov1 = \
(cr.calc_div_hozt(lon1, lat1, alt1, rho1*nwind1, rho1*ewind1)\
+cr.calc_div_vert(alt1, rho1*uwind1))/rho1
# density change (no shrink)
lon2 = np.array(g2a['Longitude'])
lat2 = np.array(g2a['Latitude'])
alt2 = np.array(g2a['Altitude'])
Re = 6371*1000 # Earth radius, unit: m
RR = Re+alt2
omega = 2*np.pi/(24*3600)
rho2 = np.array(g2a['Rho'])
nwind2 = np.array(g2a['V!Dn!N (north)'])
ewind2 = np.array(g2a['V!Dn!N (east)']) + omega*RR*np.cos(lat2)
uwind2 = np.array(g2a['V!Dn!N (up)'])
div_rhov2 = \
(cr.calc_div_hozt(lon2, lat2, alt2, rho2*nwind2, rho2*ewind2)\
+cr.calc_div_vert(alt2, rho2*uwind2))/rho2
g1a['divrhov_diff'] = div_rhov1-div_rhov2
apex = Apex(date=2003)
qlat, qlon = apex.convert(-90, 0, source='apex', dest='geo', height=400)
plt.close('all')
plt.figure(figsize=(7.26, 9.25))
for ialt, alt in enumerate([130, 200, 300, 400, 500, 600]):
alt_ind = np.argmin(np.abs(g1a['Altitude'][0, 0, :]/1000-alt))
alt_str = '%6.2f' % (g1a['Altitude'][0, 0, alt_ind]/1000)
ax, projection = gcc.create_map(
3, 2, ialt+1, 'polar', nlat=nlat, slat=slat, dlat=10,
centrallon=g3ca.calculate_centrallon(g1a, 'polar', useLT=True),
coastlines=False)
lon0, lat0, zdata0 = g3ca.contour_data('divrhov_diff', g1a, alt=alt)
fp = (lat0[:,0]>slat) & (lat0[:,0]<nlat)
lon0, lat0, zdata0 = lon0[fp, :], lat0[fp,:], zdata0[fp,:]
hc = ax.contourf(
lon0, lat0, zdata0, np.linspace(-1,1,21)*1e-4,
transform=ccrs.PlateCarree(), cmap='seismic', extend='both')
hc = plt.colorbar(hc, pad=0.17)
hc.formatter.set_powerlimits((0,0))
hc.update_ticks()
hc.set_label(r'$-\frac{\nabla\cdot(\rho\vec{u})}{\rho}$')
# wind difference
lon1, lat1, ewind1, nwind1 = g3ca.vector_data(g1a, 'neutral', alt=alt)
lon2, lat2, ewind2, nwind2 = g3ca.vector_data(g2a, 'neutral', alt=alt)
lon0, lat0 = lon1, lat1
lon0, lat0, ewind0, nwind0 = g3ca.convert_vector(
lon0, lat0, ewind2-ewind1, nwind2-nwind1, plot_type='polar',
projection=projection)
hq = ax.quiver(
lon0, lat0, ewind0, nwind0, scale=1000, scale_units='inches',
regrid_shape=20, headwidth=5)
ax.quiverkey(hq, 0.93, -0.1, 500, '500 m/s')
ax.scatter(qlon, qlat, color='k', transform=ccrs.PlateCarree())
plt.title('%s km' % alt_str, y=1.05)
plt.text(0.5, 0.95, 'Time: '+tstring, fontsize=15,
horizontalalignment='center', transform=plt.gcf().transFigure)
if show:
plt.show()
if save:
plt.savefig(spath+'08_divrhov_rho_diff_%s%s.pdf' % (tstrday,tstring))
return
