def return_obs_shear(lev1,lev2,obspath):
# Load data from text file
D = load_sounding_data(obspath)
# Interpolate wspd and wdir for lev2 from sounding
# Wdir interpolation might be rubbish if it crosses north...
wspd2 = N.interp(lev2,D['hght'],D['sknt']*0.514444)
wdir2 = N.interp(lev2,D['hght'],D['drct'])
u2,v2 = generalmet.decompose_wind(wspd2,wdir2)
if lev1 == 0:
u1,v1 = generalmet.decompose_wind(D['sknt'][0]*0.514444,D['drct'][0])
else:
pass
# Compute shear
shear = N.sqrt((u1-v1)**2 + (u2-v2)**2)
return shear
# hrStr = ["%02d" %t for t in hrIter] # dayStr = ["%02d" %t for t in dayIter] # Create figure fig = plt.figure(figsize=(width, height)) ST._isotherms() ST._isobars() ST._dry_adiabats() ST._moist_adiabats() # Observational data pres = D["obs"]["pres"] * 100 # hPa temp = D["obs"]["temp"] dewp = D["obs"]["dwpt"] wspd = generalmet.convert_kt2ms(D["obs"]["sknt"]) uwind, vwind = generalmet.decompose_wind(wspd, D["obs"]["drct"]) """ # Thin out 5000m and less thinme = N.where(D['obs']['obs0'] < 5000)[0][::5] leaveme = N.where(D['obs']['obs0'] > 5000) uwindlow = D['obs']['obs13'][thinme] vwindlow = D['obs']['obs14'][thinme] plow = pres[thinme] uwindhigh = D['obs']['obs13'][leaveme] vwindhigh = D['obs']['obs14'][leaveme] phigh = pres[leaveme] uwind = N.hstack((uwindlow, uwindhigh)) vwind = N.hstack((vwindlow, vwindhigh)) pthin = N.hstack((plow,phigh)) """
def xzPlot(nest,datetuple,XS,saveoutput=0):
outdirxs = outdir + 'xs/'
nc = openWRF(nest)
time = convert_time(nest,datetuple) # JRL function to convert time stamp to time
Nx,Ny,Nz,longitude,_lats,_dx,_dy,x_nr,y_nr = getDimensions(nc)
lats = _lats.astype(N.float32)
lons = longitude.astype(N.float32)
# Create folder for plot if necessary
try:
os.stat(outdirxs)
except:
os.makedirs(outdirxs)
for xs in XS.keys():
# x-y position of start and end points
stxy = getXYxs(lons[Ny/2,:],lats[:,Nx/2],XS[xs]['startll'])
endxy = getXYxs(lons[Ny/2,:],lats[:,Nx/2],XS[xs]['endll'])
# Joe Kington method for slice through data
hyppts = int(N.hypot(endxy[0]-stxy[0], endxy[1]-stxy[1])) # Number of pts along hypotenuse
xx = N.linspace(stxy[0],endxy[0],hyppts)
yy = N.linspace(stxy[1],endxy[1],hyppts)
xint, yint = xx.astype(int), yy.astype(int)
angle = N.arctan((yy[-1]-yy[0])/(xx[-1]-xx[0])) # In radians
# Get terrain heights?
heightground_x,heighthalf_xz = _getHeight(nc,nest, time, yy, xx, Nz, hyppts)
# Set up plot
# Length of hypotenus in km
hlen = (1/1000.0) * N.sqrt((-1.0*hyppts*nc.DX*N.cos(angle))**2 + (hyppts*nc.DY*N.sin(angle))**2)
# Hypotenuse grid length ticks: locs/labels along cross-section
hgl_ticks = N.arange(0,hlen,hlen/hyppts)
hgl_labels = [r"%3.0f" %x for x in hgl_ticks]
grid = N.swapaxes(N.repeat(N.array(hgl_ticks).reshape(hyppts,1),Nz,axis=1),0,1)
# PLOTTING
for v in XS[xs]['vars']:
plt.figure(figsize=(width, height))
if nc.DX == nc.DY: # Make sure boxes are square...
plt.axis([0,(hyppts*nc.DX/1000.0)-1,z_min,z_max+dz])
else:
print r"Domain not square - fix me"
break
if (v == 'parawind'):
u_xz = nc.variables['U'][time,:,yint,xint]
v_xz = nc.variables['V'][time,:,yint,xint]
cfdata = (N.cos(angle)*u_xz) - (N.sin(angle)*v_xz)
cf = plt.contourf(grid, heighthalf_xz, cfdata, alpha=0.6, levels=u_wind_levels,cmap=jet_sw,norm=Unorm,extend='both')
#varCBlabel = 'Rotated zonal wind speed (ms${-1}$)'
varCBlabel = r'Wind speed (ms$^{-1}$)'
if (v == 'perpwind'): # Currently reversed to give negative wind for easterly.
u_xz = nc.variables['U'][time,:,yint,xint]
v_xz = nc.variables['V'][time,:,yint,xint]
cfdata = (N.cos(angle)*v_xz) + (N.sin(angle)*u_xz)
cf = plt.contourf(grid, heighthalf_xz, cfdata*-1, alpha=0.6, levels=u_wind_levels,cmap=jet_sw,norm=Unorm,extend='both')
#varCBlabel = 'Rotated perpendicular wind speed (ms${-1}$)'
varCBlabel = r'Wind speed (ms$^{-1}$)'
if (v == 'windmag'):
u_xz = nc.variables['U'][time,:,yint,xint]
v_xz = nc.variables['V'][time,:,yint,xint]
cfdata = N.sqrt(u_xz**2 + v_xz**2)
cf = plt.contourf(grid, heighthalf_xz, cfdata, alpha=0.6, levels=xz_wind_levels,cmap=jet_sw,norm=Unorm)
if (v == 'uwind'):
u_xz = nc.variables['U'][time,:,yint,xint]
jet_sw = M.colors.LinearSegmentedColormap('jet_sw',customcmaps.jet_sw,256)
Unorm = M.colors.BoundaryNorm(u_wind_levels,256)
cf = plt.contourf(grid, heighthalf_xz, u_xz, alpha=0.6, levels=u_wind_levels,cmap=jet_sw,norm=Unorm,extend='both')
varCBlabel = r'Wind speed (ms${-1}$)'
if (v == 'uwind_sr'):
u_xz = nc.variables['U'][time,:,yint,xint]
storm_spd,storm_dir = XS[xs]['storm_motion'] # Get u-component from this
zonal_storm_motion, dummy = generalmet.decompose_wind(storm_spd,storm_dir)
u_sr = u_xz - zonal_storm_motion
jet_sw = M.colors.LinearSegmentedColormap('jet_sw',customcmaps.jet_sw,256)
Unorm = M.colors.BoundaryNorm(u_wind_levels,256)
cf = plt.contourf(grid, heighthalf_xz, u_sr, alpha=0.6, levels=u_wind_levels,cmap=jet_sw,norm=Unorm,extend='both')
varCBlabel = r'Storm-relative wind speed (ms${-1}$)'
if (v == 'vwind'):
v_xz = nc.variables['V'][time,:,yint,xint]
if (v == 'W'):
cfdata = nc.variables['W'][time,:-1,yint,xint]
Wnorm = M.colors.BoundaryNorm(w_wind_levels,256)
w_jet_sw = M.colors.LinearSegmentedColormap('w_jet_sw',customcmaps.w_jet_sw,256)
cf = plt.contourf(grid, heighthalf_xz, cfdata, alpha=0.6, levels=w_wind_levels,cmap=w_jet_sw,norm=Wnorm,extend='both')
varCBlabel = r'Vertical wind speed (ms${-1}$)'
if (v == 'T'):
pass
#cfdata = nc.variables['T'][time,:-1,yint,xint]
if (v == 'RH'):
pass
#cfdata = nc.variables['W'][time,:-1,yint,xint]
if (v == 'theta'):
theta = nc.variables['T'][time,:,yint,xint] + T_base
theta_int = N.arange(260.0,400.0,2.0)
cs = plt.contour(grid, heighthalf_xz, theta, theta_int, colors='black')
plt.clabel(cs, inline=1, fmt='%3.0f', fontsize=12, colors='black')
if (v == 'pertpres'):
pertpres = nc.variables['P'][time,:,yint,xint]
#cf = plt.contourf(grid, heighthalf_xz, cfdata, alpha=0.6, levels=w_wind_levels,cmap=w_jet_sw,norm=Wnorm,extend='both')
cf = plt.contourf(grid, heighthalf_xz, pertpres)
varCBlabel = r'Perturbation pressure (Pa)'
if (v == 'qvapor'):
qvapor = nc.variables['QVAPOR'][time,:,yint,xint]
#cf = plt.contourf(grid, heighthalf_xz, cfdata, alpha=0.6, levels=w_wind_levels,cmap=w_jet_sw,norm=Wnorm,extend='both')
cf = plt.contourf(grid, heighthalf_xz, qvapor)
varCBlabel = r'Water vapor mixing ratio (kg kg$^{-1}$)'
plt.plot(hgl_ticks,heightground_x,color='black')
plt.fill_between(hgl_ticks,heightground_x,0,facecolor='lightgrey')
labeldelta = 15
#plt.xticks(N.arange(0,hyppts,labeldelta),hgl_labels[::labeldelta])
plt.yticks(N.arange(z_min,z_max+dz,dz)) # JRL: edited in z_min
plt.xlabel(r'Distance along cross-section (km)')
plt.ylabel(r'Height above sea level (m)')
fname = '_'.join(('xs',xs,v,"%02d" %datetuple[1],"%02d" %datetuple[2],"%02d" %datetuple[3],"%02d" %datetuple[4])) + '.png'
plt.savefig(outdirxs+fname,bbox_inches='tight',pad_inches=0.5)
plt.close()
# Draw colorbar the first time through
try:
with open(outdirxs+v+'CB.png'): pass
except IOError:
# Plot colorbar
fig = plt.figure(figsize=(8,6))
cbar_ax = fig.add_axes([0.15,0.05,0.7,0.02])
cb = plt.colorbar(cf,cax=cbar_ax,orientation='horizontal')
cb.set_label(varCBlabel)
plt.savefig(outdirxs+v+'CB.png',bbox_inches='tight')
# Save all data to pickle file
if saveoutput:
D = {'day':datetuple[2], 'hour':datetuple[3], 'theta':theta,
'grid':grid, 'heighthalf_xz':heighthalf_xz, 'hgl_ticks':hgl_ticks,
'heightground_x':heightground_x, 'hyppts':hyppts, 'hgl_labels':hgl_labels,
'xs':xs, 'v':v, 'data':cfdata, 'DX':nc.DX}
pickledir = '/uufs/chpc.utah.edu/common/home/u0737349/dsws/thesis/xspickle/' + ens + '/'
fname = '_'.join(('data',xs,v,str(datetuple[1]),str(datetuple[2]),str(datetuple[3]))) + '.p'
with open(pickledir+fname,'wb') as p:
pickle.dump(D,p)
