def print_cape_for_timestamp(wrf_data, timestamp, filepath):
slp = pressure_lib.get_sea_level_pressure(wrf_data)
cinfo = getvar(wrf_data, "cape_2d", missing=0.0)
cape = cinfo[0, :, :].fillna(0)
lat, lon = latlon_coords(slp)
lat_normal = to_np(lat)
lon_normal = to_np(lon)
# rain sum
cape_res = get_pyngl(cinfo)
cape_res.nglDraw = False # don't draw plot
cape_res.nglFrame = False # don't advance frame
cape_res.cnFillOn = True # turn on contour fill
cape_res.cnLinesOn = False # turn off contour lines
cape_res.cnLineLabelsOn = False # turn off line labels
cape_res.cnFillMode = "RasterFill" # These two resources
cape_res.cnLevelSelectionMode = "ExplicitLevels"
cape_res.cnFillColors = numpy.array([ [255,255,255], [ 0,255, 0], [ 0,128, 0], \
[240,230,140], [255,255, 0], [255,140, 0], \
[255, 0, 0], [139, 0, 0], [186, 85,211],\
[153, 50,204], [139, 0,139], ],'f') / 255.
cape_res.cnLevels = numpy.array(
[.1, 500, 1000, 1500, 2000, 2500, 3000, 3500, 4000, 4500])
cape_res = geography_lib.initialize_geography(cape_res, "gray50")
cape_res.lbTitleString = "Convective available potential energy [CAPE] in (J/kg)"
cape_res.lbOrientation = "horizontal"
cape_res.lbTitleFontHeightF = 0.015
cape_res.lbLabelFontHeightF = 0.015
cape_res.tiMainString = "Thunderstorm probability (%s)" % timestamp.strftime(
"%b %d %Y %HUTC")
cape_res.trGridType = "TriangularMesh" # can speed up plotting.
cape_res.tfDoNDCOverlay = True # required for native projection
# pressure
p_res = pressure_lib.get_pressure_resource(lat_normal, lon_normal)
wk_res = Ngl.Resources()
wk_res.wkWidth = 2500
wk_res.wkHeight = 2500
output_path = "%scape_%s" % (filepath, timestamp.strftime("%Y_%m_%d_%H"))
wks_comp = Ngl.open_wks("png", output_path, wk_res)
# creating plots for the measurands
capeplot = Ngl.contour_map(wks_comp, cape, cape_res)
pplot = Ngl.contour(wks_comp, slp, p_res)
Ngl.overlay(capeplot, pplot)
Ngl.maximize_plot(wks_comp, capeplot)
Ngl.draw(capeplot)
Ngl.frame(wks_comp)
Ngl.delete_wks(wks_comp) # delete currently used workstation
def get_3h_rainsum(previous_data, current_data, timestamp, filepath):
slp = pressure_lib.get_sea_level_pressure(current_data)
previous_sum, rain_con = get_cumulated_rain_sum(previous_data)
current_sum, rain_con = get_cumulated_rain_sum(current_data)
rain_sum = current_sum - previous_sum
lat, lon = latlon_coords(rain_con)
lat_normal = to_np(lat)
lon_normal = to_np(lon)
# rain sum
rr_res = initialize_rain_resource(rain_con)
rr_res.lbTitleString = "3h rainsum in (mm)"
rr_res.lbOrientation = "horizontal"
rr_res.lbTitleFontHeightF = 0.015
rr_res.lbLabelFontHeightF = 0.015
rr_res.tiMainString = "3h rainsum (%s)" % timestamp.strftime(
"%b %d %Y %HUTC")
rr_res.trGridType = "TriangularMesh" # can speed up plotting.
rr_res.tfDoNDCOverlay = True # required for native projection
# pressure
p_res = pressure_lib.get_pressure_resource(lat_normal, lon_normal)
wk_res = Ngl.Resources()
wk_res.wkWidth = 2500
wk_res.wkHeight = 2500
output_path = "%srain_3h_%s" % (filepath,
timestamp.strftime("%Y_%m_%d_%H"))
wks_comp = Ngl.open_wks("png", output_path, wk_res)
# creating plots for the measurands
rrplot = Ngl.contour_map(wks_comp, rain_sum, rr_res)
pplot = Ngl.contour(wks_comp, slp, p_res)
Ngl.overlay(rrplot, pplot)
Ngl.maximize_plot(wks_comp, rrplot)
Ngl.draw(rrplot)
Ngl.frame(wks_comp)
Ngl.delete_wks(wks_comp) # delete currently used workstation
def plot_coherence(cohsq, phase1, phase2, symmetry=("symm"), source="", vars1="", vars2="", plotpath="./", wkstype="png",
flim=0.5, nwaveplt=20, cmin=0.05, cmax=0.55, cspc=0.05, plotxy=[1, 1], N=[1, 2]):
dims = cohsq.shape
nplot = dims[0]
FillMode = "AreaFill"
# text labels
abc = list(string.ascii_lowercase)
# plot resources
#wkstype = wkstype
reswks = ngl.Resources()
reswks.wkWidth = 3000
reswks.wkHeight = 3000
wks = ngl.open_wks(wkstype, plotpath + "SpaceTimeCoherence_", reswks)
plots = []
# coherence2 plot resources
res = coh_resources(cmin, cmax, cspc, FillMode, flim, nwaveplt)
# phase arrow resources
res_a = phase_resources(flim, nwaveplt)
# dispersion curve resources
dcres = ngl.Resources()
dcres.gsLineThicknessF = 2.0
dcres.gsLineDashPattern = 0
# text box resources
txres = ngl.Resources()
txres.txPerimOn = True
txres.txFontHeightF = 0.013
txres.txBackgroundFillColor = "Background"
# panel plot resources
res_p = panel_resources(nplot, abc)
# plot contours and phase arrows
pp = 0
while pp < nplot:
coh2 = cohsq[pp, :, :]
phs1 = phase1[pp, :, :]
phs2 = phase2[pp, :, :]
if len(symmetry) == nplot:
Symmetry = symmetry[pp]
else:
Symmetry = symmetry
var1 = vars1[pp]
var2 = vars2[pp]
res.tiMainString = source + " coh^2(" + var1 + "," + var2 + ") " + Symmetry
plot = ngl.contour(wks, coh2, res)
plot_a = ngl.vector(wks, phs1, phs2, res_a)
ngl.overlay(plot, plot_a)
(matsuno_names, textlabels, textlocsX, textlocsY) = text_labels(Symmetry)
nlabel = len(textlocsX)
# generate matsuno mode dispersion curves
if Symmetry == "midlat":
He = [3000, 7000, 10000]
matsuno_modes = matsuno_modes_wk_bg(he=He, n=N, latitude=0., max_wn=nwaveplt, n_wn=500, u=30)
else:
He = [12, 25, 50]
matsuno_modes = mp.matsuno_modes_wk(he=He, n=N, latitude=0., max_wn=nwaveplt, n_wn=500)
# add polylines for dispersion curves
for he in matsuno_modes:
df = matsuno_modes[he]
wn = df.index.values
for wavename in df:
for matsuno_name in matsuno_names:
if wavename == (matsuno_name + str(he) + "m)"):
wave = df[wavename].values
wnwave = wn[~np.isnan(wave)]
wave = wave[~np.isnan(wave)]
ngl.add_polyline(wks, plot, wnwave, wave, dcres)
# add text boxes
ll = 0
while ll < nlabel:
ngl.add_text(wks, plot, textlabels[ll], textlocsX[ll], textlocsY[ll], txres)
ll += 1
plots.append(plot)
pp += 1
# panel plots
ngl.panel(wks, plots, plotxy, res_p)
ngl.delete_wks(wks)
ngl.end()
return
clres.cnLineThicknessF = 3. # triple thick contour lines clres.cnInfoLabelOrthogonalPosF = -0.15 # Move info label up. # # Create the various plots. They will not get drawn because # nglDraw is set to False for all of them. # map_plot = Ngl.map(wks,mpres) vector_plot = Ngl.vector(wks,ua,va,vcres) line_contour_plot = Ngl.contour(wks,pa,clres) fill_contour_plot = Ngl.contour(wks,ta,cfres) # # Overlay everything on the map plot. # Ngl.overlay(map_plot,fill_contour_plot) Ngl.overlay(map_plot,line_contour_plot) Ngl.overlay(map_plot,vector_plot) # # Change the title. # srlist = Ngl.Resources() srlist.tiMainString = "vectors, line, and filled contours" Ngl.set_values(map_plot,srlist) # # Draw the map plot, which now contains the vectors and # filled/line contours. # Ngl.maximize_plot(wks,map_plot) # Maximize size of plot in frame.
vcres.vcMinDistanceF = 0.02 vcres.vcRefAnnoOrthogonalPosF = -0.20 vcres.vcRefAnnoFontHeightF = 0.005 vcres.vcLineArrowThicknessF = 2.0 # create vector plot for analysis data and overlay on colour contours level 1 uv_plot1 = ngl.vector(wks,u_1,v_1,vcres) # create vector plot for analysis data and overlay on colour contours level2 vcres.vcLineArrowColor = "Red" uv_plot2 = ngl.vector(wks,u_2,v_2,vcres) ngl.overlay(shear_plot,uv_plot1) ngl.overlay(shear_plot,uv_plot2) ngl.maximize_plot(wks, shear_plot) ngl.draw(shear_plot) ngl.frame(wks) ngl.destroy(wks) del res del vcres ################################################################################################### # open forecast file f_fili = "GFS_48h_forecast_%s_%s.nc" % (init_dt[:8], init_dt[8:10])
def large_scale_prf (ptype,cseason, ncases, cases, casenames, nsite, lats, lons, filepath, filepathobs, casedir):
# ncases, the number of models
# cases, the name of models
# casename, the name of cases
# filepath, model output filepath
# filepathobs, filepath for observational data
# inptrs = [ncases]
if not os.path.exists(casedir):
os.mkdir(casedir)
_Font = 25
interp = 2
extrap = False
mkres = Ngl.Resources()
mkres.gsMarkerIndex = 2
mkres.gsMarkerColor = 'Red'
mkres.gsMarkerSizeF = 15.
infiles = ['' for x in range(ncases)]
ncdfs = ['' for x in range(ncases)]
nregions = nsite
localnames=np.append('ERAI',casenames)
varis = [ 'CLOUD' , 'CLDLIQ', 'THETA','RELHUM']
varisobs = ['CC_ISBL', 'CLWC_ISBL', 'THETA','RELHUM' ]
alphas = ['a) ','b) ','c) ','d) ','e) ','f) ','g) ','h) ','i) ', 'j) ', 'k) ', 'l) ', 'm) ', 'n) ', 'o) ', 'p) ', 'q) ', 'r) ', 's) ', 't) ', 'u) ', 'v) ', 'w) ', 'x) ', 'y) ', 'z) ']
nvaris = len(varis)
cunits = ['%','g/kg','K', '%' ]
cscale = [100, 1000 , 1., 1 ]
cscaleobs = [100, 1000 , 1., 1, 1, 1000, 1,1,1,1,1,1,1]
obsdataset =['ERAI', 'ERAI', 'ERAI', 'ERAI', 'ERAI', 'ERAI', 'ERAI', 'ERAI','ERAI','ERAI']
plotlgs=['' for x in range(nsite)]
for ire in range (0, nsite):
if not os.path.exists(casedir+'/'+str(lons[ire])+'E_'+str(lats[ire])+'N'):
os.mkdir(casedir+'/'+str(lons[ire])+'E_'+str(lats[ire])+'N')
plotname = casedir+'/'+str(lons[ire])+'E_'+str(lats[ire])+'N/Largescale_'+str(lons[ire])+'E_'+str(lats[ire])+'N_'+cseason
plotlgs[ire] = 'Largescale_'+str(lons[ire])+'E_'+str(lats[ire])+'N_'+cseason
wks= Ngl.open_wks(ptype,plotname)
Ngl.define_colormap(wks,'GMT_paired')
plot = []
res = Ngl.Resources()
res.nglMaximize = False
res.nglDraw = False
res.nglFrame = False
res.lgPerimOn = False # no box around
res.vpWidthF = 0.30 # set width and height
res.vpHeightF = 0.30
res.tiYAxisString = "Pressure [hPa]"
if (lats[ire] > 0 and lons[ire]<230 ):
res.trYMinF = 400.
res.trYMaxF = 1000.
else:
res.trYMinF = 700.
res.trYMaxF = 1000.
res.tiMainFont = _Font
res.tmYLLabelFont = _Font
res.tmXBLabelFont = _Font
res.tiYAxisFont = _Font
res.tiXAxisFont = _Font
res.tmXBLabelFontHeightF = 0.01
res.tmXBLabelFontThicknessF = 1.0
res.xyMarkLineMode = 'Lines'
res.xyLineThicknesses = [3.0, 3.0, 3.0, 3.0, 3.0, 3.0,3.,3.,3.,3.,3,3,3,3,3,3,3]
res.xyDashPatterns = np.arange(0,24,1)
pres = Ngl.Resources()
pres.nglFrame = False
pres.txFont = _Font
pres.nglPanelYWhiteSpacePercent = 5
pres.nglPanelXWhiteSpacePercent = 5
# pres.nglPanelTop = 0.88
pres.wkPaperWidthF = 17 # in inches
pres.wkPaperHeightF = 28 # in inches
pres.nglMaximize = True
pres.nglPanelTop = 0.935
pres.nglPanelFigureStrings = alphas
pres.nglPanelFigureStringsJust = 'Topleft'
pres.nglPanelFigureStringsFontHeightF = 0.015
for iv in range (0, nvaris):
if(iv == nvaris-1):
res.pmLegendDisplayMode = 'NEVER'
res.xyExplicitLegendLabels = casenames[:]
res.pmLegendSide = 'top'
res.pmLegendParallelPosF = 0.6
res.pmLegendOrthogonalPosF = -0.5
res.pmLegendWidthF = 0.10
res.pmLegendHeightF = 0.10
res.lgLabelFontHeightF = .02
res.lgLabelFontThicknessF = 1.5
res.lgPerimOn = True
else:
res.pmLegendDisplayMode = 'NEVER'
if(obsdataset[iv] =='CCCM'):
if(cseason == 'ANN'):
fileobs = '/Users/guoz/databank/CLD/CCCm/cccm_cloudfraction_2007-'+cseason+'.nc'
else:
fileobs = '/Users/guoz/databank/CLD/CCCm/cccm_cloudfraction_2007-2010-'+cseason+'.nc'
inptrobs = Dataset(fileobs,'r')
latobs=inptrobs.variables['lat'][:]
latobs_idx = np.abs(latobs - lats[ire]).argmin()
lonobs=inptrobs.variables['lon'][:]
lonobs_idx = np.abs(lonobs - lons[ire]).argmin()
B=inptrobs.variables[varisobs[iv]][:,latobs_idx,lonobs_idx]
else:
if (varisobs[iv] =='PRECT'):
fileobs = filepathobs+'/GPCP_'+cseason+'_climo.nc'
else:
fileobs = filepathobs + obsdataset[iv]+'_'+cseason+'_climo.nc'
inptrobs = Dataset(fileobs,'r')
if (varisobs[iv] =='THETA'):
latobs=inptrobs.variables['lat'][:]
latobs_idx = np.abs(latobs - lats[ire]).argmin()
lonobs=inptrobs.variables['lon'][:]
lonobs_idx = np.abs(lonobs - lons[ire]).argmin()
B = inptrobs.variables['T'][0,:,latobs_idx,lonobs_idx]
pre1 = inptrobs.variables['lev'][:]
for il1 in range (0, len(pre1)):
B[il1] = B[il1]*(1000/pre1[il1])**0.286
else:
pre1 = inptrobs.variables['lev'][:]
latobs=inptrobs.variables['lat'][:]
latobs_idx = np.abs(latobs - lats[ire]).argmin()
lonobs=inptrobs.variables['lon'][:]
lonobs_idx = np.abs(lonobs - lons[ire]).argmin()
B = inptrobs.variables[varisobs[iv]][0,:,latobs_idx,lonobs_idx]
B[:]=B[:] * cscaleobs[iv]
for im in range (0,ncases):
ncdfs[im] = './data/'+cases[im]+'_site_location.nc'
infiles[im]= filepath[im]+cases[im]+'/'+cases[im]+'_'+cseason+'_climo.nc'
inptrs = Dataset(infiles[im],'r') # pointer to file1
lat=inptrs.variables['lat'][:]
nlat=len(lat)
lon=inptrs.variables['lon'][:]
nlon=len(lon)
lev=inptrs.variables['lev'][:]
nlev=len(lev)
ncdf= Dataset(ncdfs[im],'r')
n =ncdf.variables['n'][:]
idx_cols=ncdf.variables['idx_cols'][:,:]
ncdf.close()
if (im ==0 ):
A_field = np.zeros((ncases,nlev),np.float32)
for subc in range( 0, n[ire]):
npoint=idx_cols[ire,n[subc]-1]-1
if (varis[iv] == 'THETA'):
tmp = inptrs.variables['T'][0,:,npoint]
hyam =inptrs.variables['hyam'][:]
hybm =inptrs.variables['hybm'][:]
ps=inptrs.variables['PS'][0,npoint]
ps=ps
p0=inptrs.variables['P0']
pre = np.zeros((nlev),np.float32)
for il in range (0, nlev):
pre[il] = hyam[il]*p0 + hybm[il] * ps
tmp[il] = tmp[il] * (100000/pre[il])**0.286
theunits=str(cscale[iv])+inptrs.variables['T'].units
else:
tmp=inptrs.variables[varis[iv]][0,:,npoint]
theunits=str(cscale[iv])+inptrs.variables[varis[iv]].units
A_field[im,:] = (A_field[im,:]+tmp[:]/n[ire]).astype(np.float32 )
A_field[im,:] = A_field[im,:] *cscale[iv]
inptrs.close()
textres = Ngl.Resources()
textres.txFontHeightF = 0.02
textres.txFont = _Font
textres.txFontHeightF = 0.015
res.tiXAxisString = varis[iv]+ " Unit= " + cunits[iv]
res.trXMinF = min(np.min(A_field[0, :]),np.min(B))
res.trXMaxF = max(np.max(A_field[0, :]),np.max(B))
if(varis[iv] == 'CLOUD') :
if (lats[ire] >= 0 and lons[ire]<230):
res.trXMinF = 0.
res.trXMaxF = 50.
else:
res.trXMinF = 0.
res.trXMaxF = 100.
if(varis[iv] =='CLDLIQ') :
if (lats[ire] >= 0):
res.trXMinF = 0.
res.trXMaxF = 0.1
else:
res.trXMinF = 0.
res.trXMaxF = 0.1
if(varis[iv] =='RELHUM') :
res.trXMinF = 0.
res.trXMaxF = 100.
if(varis[iv] == 'THETA') :
if (lats[ire] >= 0 and lons[ire]<230 ):
res.trXMinF = 290.
res.trXMaxF = 340.
else:
res.trXMinF = 280
res.trXMaxF = 320
if(varis[iv] == 'T') :
res.trXMinF = 180
res.trXMaxF = 300
if(varis[iv] == 'U') :
res.trXMinF = -40
res.trXMaxF = 40
res.trYReverse = True
res.xyLineColors = np.arange(3,20,2)
res.xyMarkerColors = np.arange(2,20,2)
p = Ngl.xy(wks,A_field,lev,res)
res.trYReverse = False
res.xyLineColors = ['black']
pt = Ngl.xy(wks,B,pre1,res)
Ngl.overlay(p,pt)
plot.append(p)
Ngl.panel(wks,plot[:],[nvaris/2,2],pres)
txres = Ngl.Resources()
txres.txFontHeightF = 0.02
txres.txFont = _Font
Ngl.text_ndc(wks,'Large-scale VAR at '+ str(lons[ire])+'E,'+str(lats[ire])+'N',0.5,0.93,txres)
Common_functions.create_legend(wks,localnames,0.02,np.append('black',np.arange(3,20,2)),0.25,0.85)
Ngl.frame(wks)
Ngl.destroy(wks)
return plotlgs
def create_meteogram_for(filepath, filename, timestamp):
with open(filepath + filename) as f:
head, input = read_file(f)
# Create numpy 2D array
cdf = numpy.array(input)
# Collect meteorological data
pressure = cdf[:, 9] / 100
rain_cum = cdf[:, 16]
rain_expl = cdf[:, 17]
tempht = cdf[:, 5] - 273.15
hum = cdf[:, 6]
taus = cdf[:, 1]
u = cdf[:, 7]
v = cdf[:, 8]
rain_sum = numpy.add(rain_cum, rain_expl)
# set up a color map and open an output workstation.
colors = numpy.array([ \
[255,255,255], [255,255,255], [255,255,255], \
[240,255,240], [220,255,220], [190,255,190], \
[120,255,120], [ 80,255, 80], [ 50,200, 50], \
[ 20,150, 20] \
],'f') / 255.
wks_type = "png"
output_file = filepath + timestamp.strftime(
"%m_%d_%H") + "_meteogram_" + filename.split(".")[0]
wks = Ngl.open_wks(wks_type, output_file)
# calculate secondary data
count_xdata = taus[-1]
main_hours, sec_hours, labels = generate_xlegend(timestamp, count_xdata)
# dew points data and relative humidity
rel_hum = humidity_lib.calculate_relative_humidity(hum, tempht, pressure,
len(taus))
dew_point = temperature_lib.calculate_dewpoint(tempht, rel_hum, len(taus))
# 3 hour rain sums
rain3h_sum, rain3h_time = rain_lib.calculate_3hrain_data(rain_sum, taus)
# wind speed
wind_speed = wind_lib.calculate_windspeed(u, v, len(taus))
# wind direction
wind_direction = wind_lib.calculate_winddirection(u, v, len(taus))
# generate measurand resources
# pressure resource
sealevel_pressure = pressure_lib.reduce_pressure_to_sealevel(
pressure, cdf[:, 5], float(head[13]))
pres_res = pressure_lib.get_pressure_resource(count_xdata,
sealevel_pressure)
pres_res.tiMainString = format_title(head[0], timestamp)
pres_res = config_xaxis_legend(pres_res, main_hours, sec_hours, labels)
# relative humidity
relhum_res = humidity_lib.get_relhumidity_resource(count_xdata)
relhum_res = config_xaxis_legend(relhum_res, main_hours, sec_hours, labels)
# wind speed recource
wind_res = wind_lib.get_windspeed_resource(count_xdata, wind_speed)
wind_res = config_xaxis_legend(wind_res, main_hours, sec_hours, labels)
# wind direction recource
direction_res = wind_lib.get_winddirection_resource(
count_xdata, wind_speed)
direction_res = config_xaxis_legend(direction_res, main_hours, sec_hours,
labels)
# rain sum resources
rainsum_res = rain_lib.get_rainsum_resource(count_xdata)
rainsum_res = config_xaxis_legend(rainsum_res, main_hours, sec_hours,
labels)
# 3 hour rain sum bar charts
rain3h_res = rain_lib.get_3hrain_resource(rain3h_time)
rainhist = create_rain_bar_plot(wks, rain3h_time, rain3h_sum, rain3h_res)
# ground temperature resource
tempsfc_res = temperature_lib.get_temperature_resource(
count_xdata, tempht, dew_point)
tempsfc_res = config_xaxis_legend(tempsfc_res, main_hours, sec_hours,
labels)
# generate plot results
pressmsz = Ngl.xy(wks, taus, sealevel_pressure, pres_res)
relhummsz = Ngl.xy(wks, taus, rel_hum, relhum_res)
windmsz = Ngl.xy(wks, taus, wind_speed, wind_res)
dirmsz = Ngl.xy(wks, taus, wind_direction, direction_res)
rainsum = Ngl.xy(wks, taus, rain_sum, rainsum_res)
temptmsz = Ngl.xy(wks, taus, tempht, tempsfc_res)
tempsfc_res.xyLineColor = "blue" # line color for dew point
dewpmsz = Ngl.xy(wks, taus, dew_point, tempsfc_res)
Ngl.draw(pressmsz)
Ngl.draw(relhummsz)
Ngl.overlay(rainsum, rainhist)
Ngl.draw(rainsum)
Ngl.draw(windmsz)
Ngl.draw(dirmsz)
Ngl.overlay(temptmsz, dewpmsz)
Ngl.draw(temptmsz)
Ngl.frame(wks)
Ngl.delete_wks(wks) # delete currently used workstation
bres.tmXBLabelFontHeightF = 0.015 # Make labels smaller. This
# will affect Y labels too.
#---Set the values and labels for the Y axis of blank plot.
bres.tmYLMode = "Explicit"
bres.tmYLValues = lat_values
bres.tmYLLabels = lat_labels
#---Align four corners of both plots, that is, don't do data transformation
bres.tfDoNDCOverlay = True
#---Create the blank plot.
blank = Ngl.blank_plot(wks,bres)
#---Draw blank plot for debugging purposes.
#Ngl.draw(blank)
#Ngl.frame(wks)
#---Overlay blank plot on existing map plot.
Ngl.overlay(map.base,blank)
#---Resize map so it fits within frame
Ngl.maximize_plot(wks, map)
#---Drawing the original map also draws the overlaid blank plot
Ngl.draw(map)
Ngl.frame(wks)
Ngl.end()
cnres2.nglDraw = False
cnres2.cnFillOn = False
cnres2.cnLinesOn = True
cnres2.cnLineLabelsOn = False
cnres2.cnInfoLabelOn = False
#cnres2.cnLineLabelInterval = 4
#cnres2.cnLineLabelBackgroundColor = -1
#cnres2.cnLineLabelFontColor = "gray20"
#cnres2.cnLineLabelFontHeightF = 0.014
#cnres2.cnLabelMasking = True
cnres2.cnLineDashPatterns = 3
cnres2.cnLineThicknessF = 4
cnres2.cnLineColor = "gray20"
cnres2.cnLevelSelectionMode = "ManualLevels"
cnres2.cnMinLevelValF = 0.
cnres2.cnMaxLevelValF = 10.
cnres2.cnLevelSpacingF = 1.
cnres2.sfXArray = lon.data
cnres2.sfYArray = lat.data
contour = Ngl.contour_map(wks, var1[:, :], cnres)
contour2 = Ngl.contour(wks, var1[:, :], cnres2)
tx = add_labels_lcm(wks, contour, 5, 5)
Ngl.overlay(contour, contour2)
Ngl.draw(contour)
Ngl.frame(wks)
print('end')
#Ngl.end()
res.cnMonoLineLabelFontColor = True
res.lbLabelFontHeightF = 0.01
res.cnLineDashPattern = 11
res.cnLineThicknessF = 5.0
res.cnLineColor = "purple"
res.cnLevelSelectionMode = "ManualLevels"
res.cnMinLevelValF = -85.0
res.cnMaxLevelValF = 115.0
res.cnLevelSpacingF = 100.0
# plot ITD and overlay on colour contours
dp_plot2 = ngl.contour(wks, dewpoint, res)
ngl.overlay(dp_plot, dp_plot2)
ngl.maximize_plot(wks, dp_plot)
ngl.draw(dp_plot)
ngl.frame(wks)
ngl.destroy(wks)
del res
del dewpoint
###################################################################################################
# open forecast file
f_fili = "GFS_forecast_%s_%s.nc" % (init_dt[:8], init_dt[8:10])
forecast = nio.open_file(diri + f_fili)
res.cnLineLabelsOn = True
res.cnLinesOn = True
res.cnMonoLineLabelFontColor = True
res.lbLabelFontHeightF = 0.01
res.cnLevelSelectionMode = "ManualLevels"
res.cnMinLevelValF = geo_thresh - 200.0
res.cnMaxLevelValF = geo_thresh + 200.0
res.cnLevelSpacingF = 200.0
res.cnLineThicknessF = 2.5
# plot ITD and overlay on colour contours
HL_plot = ngl.contour(wks, geo_diff, res)
ngl.overlay(dp_plot, HL_plot)
ngl.maximize_plot(wks, dp_plot)
ngl.draw(dp_plot)
ngl.frame(wks)
ngl.destroy(wks)
del res
del dp
del geo1
del geo2
del geo_diff
###################################################################################################
os.system('mogrify -trim *' + region + '_*DPandHL_2M_SNGL_' + init_dt[0:10] +
# Create base plot upon which we will overlay the rest of the
# "spaghetti" contours.
#
plot_base = Ngl.contour_map(wks, hgt0, mpres)
#
# Copy just the contour resources from mpres to a new resource list (cnres).
#
cnres = Ngl.Resources()
for t in dir(mpres):
if (t[0:2] == 'cn' or t[0:2] == 'sf' or t[0:3] == 'ngl'):
setattr(cnres, t, getattr(mpres, t))
#
# Loop over other 19 fields but only do contour plot.
# Note the color index is changing. In loop, we select
# a new color from the default color map.
#
for i in range(19):
cnres.cnLineColor = 2 + i # Change line color.
hgt0 = Ngl.add_cyclic(hgt[i + 1, :, :]) # Add a cyclic pt to data.
plot = Ngl.contour(wks, hgt0, cnres) # Generate contours.
Ngl.overlay(plot_base, plot) # Overlay this contour on map.
# Draw the plot, add the lon labels, and advance the frame.
Ngl.draw(plot_base)
add_lon_labels(wks)
Ngl.frame(wks)
Ngl.end()
resv.vfXArray = lon resv.vfYArray = lat # set the annotation string. ~C~ is line break resv.vcRefAnnoString1 = "Wind Speed~C~ (50 m/s)" resv.vcRefAnnoArrowSpaceF = 0.65 # reduces white space resv.vcRefAnnoString2On = False # remove the string "Reference vector" resv.vcRefMagnitudeF = 50.0 # speed of the reference arrow resv.vcRefLengthF = 0.08 # length of the reference arrow resv.vcMinDistanceF = 0.02 # min. dist. between arrows # Draw a vector plot. Note that here, the vector_map method is not used, # since map projection will be managed vplot = Ngl.vector(wks, u, v, resv) Ngl.overlay(m, vplot) # draws the map Ngl.draw(m) # add a page to the pdf output Ngl.frame(wks) # add the colormap to see the colors # Ngl.draw_colormap(wks) # Ngl.end() # - # <img src="figs/pyngl_examples.000008.png" width="50%">
map_plot = Ngl.map(wks, resmp)
plot_fill = Ngl.contour(wks, 3.6E6 * field[:, :, 0], resources)
#plt.hist(3.6E6*field[:,:,0].flatten(),range(1,20))
#plt.savefig('RC.png')
fst_data = read_fst(file, "PN", 0, fhour, yy)
field = fst_data[0]
lat = fst_data[1]
lon = fst_data[2]
respn.sfXArray = lon
respn.sfYArray = lat
plot_line = Ngl.contour(wks, field[:, :, 0], respn)
Ngl.overlay(map_plot, plot_fill)
Ngl.overlay(map_plot, plot_line)
Ngl.draw(map_plot)
Ngl.frame(wks)
Ngl.delete_wks(wks)
del ([resources, respn, resmp])
print 'done'
os.system("convert -density 300 RT.ps RT.png")
cmd = "convert -trim -geometry 2000x1500 +repage -border 8 -bordercolor white -background white -flatten RT.png RT.png"
os.system(cmd)
del ([field, lat, lon])
if var == "RC" and fhour <= 144 and model_name != "HRDPS":
print "calculating mid-level precip rate"
taus_above_zero = numpy.take(taus, ind_above_zero) px[0::5] = (taus_above_zero - dx / 2.).astype(taus.dtype.char) px[1::5] = (taus_above_zero - dx / 2.).astype(taus.dtype.char) px[2::5] = (taus_above_zero + dx / 2.).astype(taus.dtype.char) px[3::5] = (taus_above_zero + dx / 2.).astype(taus.dtype.char) px[4::5] = (taus_above_zero - dx / 2.).astype(taus.dtype.char) py[0::5] = rain_res.trYMinF py[1::5] = numpy.take(rain03, ind_above_zero) py[2::5] = numpy.take(rain03, ind_above_zero) py[3::5] = rain_res.trYMinF py[4::5] = rain_res.trYMinF polyg = Ngl.add_polygon(wks, rainhist, px, py, pgres) # # For the outlines, we don't need the fifth point. # polyl = Ngl.add_polyline(wks, rainhist, px, py, pgres) temptmsz = Ngl.xy(wks, taus, tempht, tempsfc_res) # ---------------------- overlay, draw, and advance frame --------- Ngl.overlay(rhfill, templine) # Overlay temperature contour on rh plot. Ngl.overlay(rhfill, windlayer) # Overlay windbarbs on rh plot. Ngl.draw(rhfill) Ngl.draw(rainhist) Ngl.draw(temptmsz) Ngl.frame(wks) Ngl.end()
mpres.mpMaskAreaSpecifiers = mask_specs # areas to mask (protect)
#
# This set of resources deals with a map labelbar.
#
# Note that this is different from a contour labelbar
# (which we will turn off below), and we have more
# control over it (i.e. we can set the number of boxes,
# the fill # colors, etc) than we do with a contour labelbar.
#
mpres.pmLabelBarDisplayMode = "Always" # Turn on a map labelbar
# Labelbar resources.
mpres.lbLabelStrings = ["Ocean","Land","< 0","0-10","10-20","20-30",\
"30-40","40-50","50-60","60-70","70-80", \
"80-90","90-100","> 100"]
mpres.lbFillColors = mp_fill_colrs + cn_fill_clrs
# Fill resources
cnres.cnFillColors = cn_fill_clrs
cnres.pmLabelBarDisplayMode = "Never" # turn off, b/c map has one turned on
contour = Ngl.contour(wks,z[:,:],cnres)
map = Ngl.map(wks,mpres)
Ngl.overlay(map,contour)
Ngl.draw(map)
Ngl.frame(wks)
Ngl.end()
res.cnMonoLineLabelFontColor = True res.lbLabelFontHeightF = 0.0075 res.cnLineThicknessF = 2.5 res.cnInfoLabelOn = True res.cnInfoLabelString = "CIN Contours at -50, -100 and -250 J/Kg" res.cnInfoLabelOrthogonalPosF = -0.06 res.cnInfoLabelParallelPosF = 0.505 res.cnLevelSelectionMode = "ExplicitLevels" res.cnLevels = [-250.0, -100.0, -50.0] # plot CIN and overlay on colour contours CIN_plot = ngl.contour(wks,CIN,res) ngl.overlay(CAPE_plot,CIN_plot) ngl.maximize_plot(wks, CAPE_plot) ngl.draw(CAPE_plot) ngl.frame(wks) ngl.destroy(wks) del res del CAPE del CIN ################################################################################################### # open forecast file f_fili = "GFS_forecast_%s_%s.nc" % (init_dt[:8], init_dt[8:10])
def contourwithoverlay(wks,resMP,varin, plotoverlay):
plottemp = Ngl.contour_map(wks,varin,resMP)
Ngl.overlay(plottemp,plotoverlay)
return(plottemp)
mpres.mpLabelsOn = False mpres.mpLeftCornerLatF = 10. mpres.mpLeftCornerLonF = -180. mpres.mpLimitMode = "corners" mpres.mpProjection = "Stereographic" mpres.mpRightCornerLatF = 10. mpres.mpRightCornerLonF = 0. mpres.tiMainFontHeightF = 0.02 mpres.tiMainString = "Busy graphic with contours and streamlines" stream = Ngl.streamline(wks, data[0, :, :], data[1, :, :], stres) contour = Ngl.contour(wks, data[2, :, :], cnres) map = Ngl.map(wks, mpres) Ngl.overlay(map, stream) Ngl.overlay(map, contour) # Since we overlaid 2 plots, we need to resize them to make sure # they fit in frame. Ngl.maximize_plot(wks, map) Ngl.draw(map) Ngl.frame(wks) del cnres.nglDraw del cnres.nglFrame del mpres.nglDraw del mpres.nglFrame del stres.nglDraw del stres.nglFrame
res.vcLineArrowColor = "Black"
vcres.vcRefMagnitudeF = 30.0 # define vector ref mag
vcres.vcRefLengthF = 0.03 # define length of vec ref
vcres.vcMinFracLengthF = 0.3
vcres.vcMinDistanceF = 0.02
vcres.vcRefAnnoOrthogonalPosF = -0.20
vcres.vcRefAnnoFontHeightF = 0.005
vcres.vcLineArrowThicknessF = 2.0
# create vector plot for analysis data and overlay on colour contours
uv_plot1 = ngl.vector(wks, max_shear_u2_u1, max_shear_v2_v1, vcres)
# create vector plot for analysis data and overlay on colour contours level2
ngl.overlay(PRATE_plot, uv_plot1)
ngl.maximize_plot(wks, PRATE_plot)
ngl.draw(PRATE_plot)
ngl.frame(wks)
ngl.destroy(wks)
del res.mpProjection
del res.mpLimitMode
del res.mpMinLonF
del res.mpMaxLonF
del res.mpMinLatF
del res.mpMaxLatF
# del res.mpPerimOn
del res.mpOutlineBoundarySets
del res.mpNationalLineColor
volplot = Ngl.contour(wks,zplot[ja,:,:],res)
##
## Draw hydrothermal stream lines on top
##
res.pmLabelBarDisplayMode = 'Never'
res.cnFillOn = False
res.cnLinesOn = True
res.cnLineLabelsOn = True
res.cnInfoLabelOn = False
res.cnLevels = psi_levels
psi2 = psi * 10**(-9)
psiplot = Ngl.contour(wks,psi2,res)
Ngl.overlay(volplot,psiplot)
Ngl.draw(volplot)
print '-------------------------------------------'
print 'mass('+rts_label[jb]+','+rts_label[jc]+')'
print ' '+title[ja]
print ' Total mass: '+repr(np.sum(volrr[jb,jc,ja,:,:]))
tres = Ngl.Resources()
fhei = 0.03
tres.txFontHeightF = fhei
Ngl.text_ndc(wks,'b)',0.82,0.67,tres)
Ngl.frame(wks)
def grid_order_contourplot(model, grid, first_varying_var, cut_domain):
# define paths #
path = dict(base='/', plots='data/plots/experimental/gridpoint_order/')
print('plotting model: {}'.format(model))
# load icosahedral grid information #
if grid == 'icosahedral':
clat, clon, vlat, vlon = read_grid_coordinates(model, grid)
data_array1 = np.arange(clat.shape[0], dtype='float32')
else:
clat, clon = read_grid_coordinates(model, grid)
clon = clon - 180
if first_varying_var == 'lon':
data_array1 = np.arange(clat.shape[0] * clon.shape[0],
dtype='float32').reshape(
(clat.shape[0], clon.shape[0]))
elif first_varying_var == 'lat':
data_array1 = np.arange(clat.shape[0] * clon.shape[0],
dtype='float32').reshape(
(clon.shape[0], clat.shape[0])).T
print('data_array1:', data_array1.shape, 'clat:', clat.shape, 'clon:',
clon.shape)
if grid == 'icosahedral':
print('vlat:', vlat.shape, 'vlon:', vlon.shape)
print('data_array1 min,max:', data_array1.min(), data_array1.max())
print('clat min,max:', clat.min(), clat.max())
print('clon min,max:', clon.min(), clon.max())
if grid == 'icosahedral':
print('vlat min,max:', vlat.min(), vlat.max())
print('vlon min,max:', vlon.min(), vlon.max())
print('------------------------------------------')
if cut_domain['name'] == 'uncut':
data_array1_cut = data_array1
else:
margin_deg = 0
if grid == 'icosahedral':
data_array1_cut, clat, clon, vlat, vlon \
= cut_by_domain(cut_domain, grid, data_array1, clat, clon, vlat, vlon, margin_deg)
else:
data_array1_cut, clat, clon \
= cut_by_domain(cut_domain, grid, data_array1, clat, clon, None, None, margin_deg)
print('data_array1:', data_array1_cut.shape, 'clat:', clat.shape, 'clon:',
clon.shape)
if grid == 'icosahedral':
print('vlat:', vlat.shape, 'vlon:', vlon.shape)
print('data_array1_cut min,max:', data_array1_cut.min(),
data_array1_cut.max())
print('clat min,max:', clat.min(), clat.max())
print('clon min,max:', clon.min(), clon.max())
if grid == 'icosahedral':
print('vlat min,max:', vlat.min(), vlat.max())
print('vlon min,max:', vlon.min(), vlon.max())
if grid == 'icosahedral':
plot_name = 'gridpoint_order_{}_{}_global_{}'.format(
model, grid, cut_domain['name'])
else:
plot_name = 'gridpoint_order_{}_{}_{}-varying-first_global_{}'.format(
model, grid, first_varying_var, cut_domain['name'])
# plot basic map with borders #
wks_res = Ngl.Resources()
x_resolution = 800
y_resolution = 800
wks_res.wkWidth = x_resolution
wks_res.wkHeight = y_resolution
wks_res.wkColorMap = 'BkBlAqGrYeOrReViWh200'
levels1 = np.linspace(data_array1.min(), data_array1.max(), 200)
wks_type = 'png'
wks = Ngl.open_wks(wks_type, path['base'] + path['plots'] + plot_name,
wks_res)
mpres = Ngl.Resources()
mpres.mpProjection = 'CylindricalEquidistant'
mpres.mpLimitMode = 'LatLon'
mpres.mpCenterLonF = 0.0
mpres.mpCenterLatF = 0.0
mpres.mpMinLonF = -180.
mpres.mpMaxLonF = 180.
mpres.mpMinLatF = -90.
mpres.mpMaxLatF = 90.
mpres.nglMaximize = False
mpres.vpXF = 0.003
mpres.vpYF = 1.00
mpres.vpWidthF = 0.86
mpres.vpHeightF = 1.00
mpres.mpMonoFillColor = 'True'
mpres.mpFillColor = -1
#Ngl.set_values(wks,mpres)
mpres.mpFillOn = True
#resources.cnFillDrawOrder = 'Predraw' # draw contours first
mpres.mpGridAndLimbOn = False
mpres.mpGreatCircleLinesOn = False
#mpres.mpOutlineDrawOrder = 'PreDraw'
mpres.mpDataBaseVersion = 'LowRes'
mpres.mpDataSetName = 'Earth..4'
mpres.mpOutlineBoundarySets = 'national'
mpres.mpGeophysicalLineColor = 'black'
mpres.mpNationalLineColor = 'black'
mpres.mpGeophysicalLineThicknessF = 2. * x_resolution / 1000
mpres.mpNationalLineThicknessF = 2. * x_resolution / 1000
mpres.mpPerimOn = True
mpres.mpPerimLineColor = 'black'
mpres.mpPerimLineThicknessF = 8.0 * x_resolution / 1000
mpres.tmXBOn = False
mpres.tmXTOn = False
mpres.tmYLOn = False
mpres.tmYROn = False
mpres.nglDraw = False #-- don't draw plot
mpres.nglFrame = False #-- don't advance frame
basic_map = Ngl.map(wks, mpres)
# plot variable1 in shading/contours #
v1res = Ngl.Resources()
v1res.sfDataArray = data_array1_cut
v1res.sfXArray = clon
v1res.sfYArray = clat
if grid == 'icosahedral':
v1res.cnFillMode = 'CellFill'
v1res.sfXCellBounds = vlon
v1res.sfYCellBounds = vlat
else:
v1res.cnFillMode = 'RasterFill'
v1res.sfMissingValueV = 9999
v1res.cnLinesOn = False # Turn off contour lines.
v1res.cnFillOn = True
#v1res.cnFillOpacityF = 0.5
#v1res.cnFillDrawOrder = 'Predraw'
v1res.cnLineLabelsOn = False
v1res.cnLevelSelectionMode = 'ExplicitLevels'
v1res.cnLevels = levels1
# set resources for a nice label bar #
v1res.lbLabelBarOn = True
v1res.lbAutoManage = False
v1res.lbOrientation = 'vertical'
v1res.lbLabelOffsetF = 0.04 # minor axis fraction: the distance between colorbar and numbers
v1res.lbBoxMinorExtentF = 0.20 # minor axis fraction: width of the color boxes when labelbar down
v1res.lbTopMarginF = 0.05 # make a little more space at top for the unit label
v1res.lbRightMarginF = 0.0
v1res.lbBottomMarginF = 0.0
v1res.lbLeftMarginF = -0.35
v1res.cnLabelBarEndStyle = 'ExcludeOuterBoxes'
#v1res.cnLabelBarEndStyle = 'IncludeOuterBoxes'
#v1res.cnExplicitLabelBarLabelsOn = True
#v1res.pmLabelBarDisplayMode = 'Always'
v1res.pmLabelBarWidthF = 0.10
#v1res.lbLabelStrings = label_str_list
v1res.lbLabelFontHeightF = 0.010
#v1res.lbBoxCount = 40
v1res.lbBoxSeparatorLinesOn = False
v1res.lbBoxLineThicknessF = 4
#v1res.lbBoxEndCapStyle = 'TriangleBothEnds'
v1res.lbLabelAlignment = 'InteriorEdges'
v1res.lbLabelStride = 1
v1res.nglFrame = False
v1res.nglDraw = False
v1plot = Ngl.contour(wks, data_array1_cut, v1res)
# plot label bar unit #
text_str = 'Index'
text_res_1 = Ngl.Resources()
text_res_1.txFontColor = 'black'
text_x = 0.975
text_y = 0.72
text_res_1.txFontHeightF = 0.012
Ngl.overlay(basic_map, v1plot)
Ngl.draw(basic_map)
Ngl.text_ndc(wks, text_str, text_x, text_y, text_res_1)
Ngl.frame(wks)
Ngl.destroy(wks)
#Ngl.end()
# cut top and bottom whitespace of plot #
im = Image.open(path['base'] + path['plots'] + plot_name + '.png')
image_array = np.asarray(im.convert('L'))
image_array = np.where(image_array < 255, 1, 0)
image_filter = np.amax(image_array, axis=1)
vmargins = [
np.nonzero(image_filter)[0][0],
np.nonzero(image_filter[::-1])[0][0]
]
#print(vmargins)
#print(im.size)
im_cropped = Image.new(
'RGB', (im.size[0], im.size[1] - vmargins[0] - vmargins[1]),
(255, 255, 255))
im_cropped.paste(
im.crop((0, vmargins[0], im.size[0], im.size[1] - vmargins[1])),
(0, 0))
#print(im_cropped.size)
im.close()
im_cropped.save(path['base'] + path['plots'] + plot_name + '.png', 'png')
im_cropped.close()
del data_array1, data_array1_cut, clat, clon
if grid == 'icosahedral':
del vlat, vlon
return
vcres.vfYArray = lat vcres.vcFillArrowsOn = True vcres.vcRefMagnitudeF = 30.0 # define vector ref mag vcres.vcRefLengthF = 0.02 # define length of vec ref vcres.vcMinFracLengthF = 0.3 vcres.vcMinDistanceF = 0.02 vcres.vcRefAnnoOrthogonalPosF = -0.20 vcres.vcRefAnnoFontHeightF = 0.005 cnres.tiMainString = "Fully opaque filled vectors over filled contours" #---Draw fully opaque vectors uv_plot = Ngl.vector(wks,u,v,vcres) spd_plot = Ngl.contour_map(wks,spd,cnres) Ngl.overlay(spd_plot,uv_plot) Ngl.maximize_plot(wks, spd_plot) Ngl.draw(spd_plot) Ngl.frame(wks) #---This time make vectors partially transparent vcres.vcGlyphOpacityF = 0.3 cnres.tiMainString = "Partially transparent vectors over filled contours" uv_plot = Ngl.vector(wks,u,v,vcres) spd_plot = Ngl.contour_map(wks,spd,cnres) Ngl.overlay(spd_plot,uv_plot) Ngl.maximize_plot(wks, spd_plot) Ngl.draw(spd_plot) Ngl.frame(wks)
res.cnMonoLineLabelFontColor = True
res.lbLabelFontHeightF = 0.01
res.cnLineDashPattern = 0
res.cnLineThicknessF = 5.0
res.cnLineColor = "steelblue1"
res.cnLevelSelectionMode = "ManualLevels"
res.cnMinLevelValF = -85.0
res.cnMaxLevelValF = 115.0
res.cnLevelSpacingF = 100.0
# plot ITD and overlay on colour contours
ws_plot2 = ngl.contour(wks, ws, res)
ngl.overlay(ws_plot, ws_plot2)
ngl.overlay(ws_plot, uv_plot)
ngl.maximize_plot(wks, ws_plot)
ngl.draw(ws_plot)
ngl.frame(wks)
ngl.destroy(wks)
del res
del vcres
###################################################################################################
# open forecast file
f_fili = "GFS_forecast_%s_%s.nc" % (init_dt[:8], init_dt[8:10])
taus_above_zero = Numeric.take(taus,ind_above_zero) px[0::5] = (taus_above_zero - dx/2.).astype(taus.typecode()) px[1::5] = (taus_above_zero - dx/2.).astype(taus.typecode()) px[2::5] = (taus_above_zero + dx/2.).astype(taus.typecode()) px[3::5] = (taus_above_zero + dx/2.).astype(taus.typecode()) px[4::5] = (taus_above_zero - dx/2.).astype(taus.typecode()) py[0::5] = rain_res.trYMinF py[1::5] = Numeric.take(rain03,ind_above_zero) py[2::5] = Numeric.take(rain03,ind_above_zero) py[3::5] = rain_res.trYMinF py[4::5] = rain_res.trYMinF polyg = Ngl.add_polygon(wks,rainhist,px,py,pgres) # # For the outlines, we don't need the fifth point. # polyl = Ngl.add_polyline(wks,rainhist,px,py,pgres) temptmsz = Ngl.xy(wks,taus,tempht,tempsfc_res) # ---------------------- overlay, draw, and advance frame --------- Ngl.overlay(rhfill,templine) # Overlay temperature contour on rh plot. Ngl.overlay(rhfill,windlayer) # Overlay windbarbs on rh plot. Ngl.draw(rhfill) Ngl.draw(rainhist) Ngl.draw(temptmsz) Ngl.frame(wks) Ngl.end()
def plot_psirr_lines(vpx=0.2, vpy=0.8, hei=0.4, xlab=1,ylab=1):
psi_levels = np.array([-500,-50])
if(1):
if(1):
res = Ngl.Resources()
res.nglMaximize = False
res.nglPaperOrientation = 'Portrait'
res.nglFrame = False
res.nglDraw = False
res.cnFillOn = False
res.cnLinesOn = True
res.cnLineLabelsOn = True
res.cnLineLabelDensityF = 2
res.cnLineLabelBackgroundColor = -1
res.cnLevelSelectionMode = 'ExplicitLevels'
res.cnLevels = psi_levels
res.cnMonoLineColor = True
res.cnLineColor = 'blue'
res.cnLineThicknessF = 2
res.cnInfoLabelOn = False
res.sfXArray = vx
res.sfYArray = vy
res.trXMinF = -3
res.trXMaxF = 55
res.trYMinF = 250
res.trYMaxF = 360
res.tiXAxisString = 'Latent heat [kJ/kg]'
res.tiYAxisString = 'Dry static energy [kJ/kg]'
res.tiMainString = title
if (xlab == 1):
res.tiXAxisOn = True
res.tmXBLabelsOn = True
res.tiXAxisSide = 'Bottom'
elif (xlab == -1):
res.tiXAxisOn = True
res.tmXBLabelsOn = False
res.tmXTLabelsOn = True
res.tiXAxisSide = 'Top'
elif (xlab == 0):
res.tiXAxisOn = False
res.tmXBLabelsOn = False
res.tmXTLabelsOn = False
if (ylab == 1):
res.tiYAxisOn = True
res.tmYLLabelsOn = True
res.tmYRLabelsOn = False
res.tiYAxisSide = 'Left'
elif (ylab == -1):
res.tiYAxisOn = True
res.tmYLLabelsOn = False
res.tmYRLabelsOn = True
res.tiYAxisSide = 'Right'
elif (ylab == 0):
res.tiYAxisOn = False
res.tmYLLabelsOn = False
res.tmYRLabelsOn = False
res.vpWidthF = hei
res.vpHeightF = hei
res.vpYF = vpy
res.vpXF = vpx
lsmooth = False
if (lsmooth):
sigma = 2
order = 0
for ji in range(0,2):
ndimage.filters.gaussian_filter(zplot1[:,:],sigma,order=order,\
output=zplot[:,:],\
mode='reflect', cval=0.0)
cont1 = Ngl.contour(wks,zplot1[:,:],res)
## Clausius-Clapeyron line
vy_cc = np.linspace(240,360,101)
press = 101300.
vx_cc,es = cclap(vy_cc*1000/1004.,press)
vx_cc = 2.5 * 10**3 * vx_cc
#ccstring = 'C-C, RH=100%'
lres = Ngl.Resources()
lres.gsLineDashPattern = 15
lres.gsLineLabelFontHeightF = 0.009
lres.gsLineThicknessF = 3
#lres.gsLineLabelString = ccstring
line = Ngl.add_polyline(wks,cont1,vx_cc,vy_cc,lres)
## MSE profiles
lres = Ngl.Resources()
lres.gsLineDashPattern = 1
lres.gsLineThicknessF = 7
lres.gsLineColor = 'blue'
line = Ngl.add_polyline(wks,cont1,vlh1,vdse1,lres)
res.cnLineColor = 'red'
cont2 = Ngl.contour(wks,zplot2[:,:],res)
## MSE profiles
lres = Ngl.Resources()
lres.gsLineDashPattern = 1
lres.gsLineThicknessF = 7
lres.gsLineColor = 'red'
line = Ngl.add_polyline(wks,cont2,vlh2,vdse2,lres)
Ngl.overlay(cont2,cont1)
Ngl.draw(cont2)
def draw_gh_500hPa(sel_year, sel_month):
print("darwing 500hPa gh plot for " + str(sel_month).zfill(2) + " " +
str(sel_year))
files_cli = sorted(
glob.glob(
os.path.join(
'/home/alley/work/Dong/mongo/seasonal_analysis/data/data/download_from_mongo/cli/',
'gh_500_*.grb')))
f_cli = xr.open_mfdataset(files_cli,
concat_dim="time",
combine="nested",
engine="cfgrib",
parallel=True)
h = f_cli["z"]
h = h.mean(dim="time")
h_cli = h / 98
lat = f_cli["latitude"].values
lon = f_cli["longitude"].values
file_cur = "/home/alley/work/Dong/mongo/seasonal_analysis/data/data/download_from_mongo/cur/gh_500_" + str(
sel_year) + str(sel_month).zfill(2) + ".grb"
# print(file_cur)
f_cur = xr.open_mfdataset(file_cur, engine="cfgrib", parallel=True)
h_cur = f_cur["z"] / 98
h_ano = h_cur - h_cli
leftString = "500hPa Height & Anomaly for " + str(sel_month) + str(
sel_year)
rightString = "dagpm"
wks_type = 'png'
wks = Ngl.open_wks(
wks_type,
'/home/alley/work/Dong/mongo/seasonal_analysis/images/gh_500hPa_' +
str(sel_year) + str(sel_month).zfill(2) + '.png')
res = Ngl.Resources()
res.nglFrame = False
res.nglDraw = False
res.mpLimitMode = "LatLon"
res.mpMinLonF = 60
res.mpMaxLonF = 180
res.mpMinLatF = 0
res.mpMaxLatF = 60
plot_map = Ngl.map(wks, res)
fres = Ngl.Resources()
fres.nglFrame = False
fres.nglDraw = False
fres.cnFillOn = True
fres.sfXArray = lon
fres.sfYArray = lat
fres.lbOrientation = "Horizontal" # horizontal labelbar
fres.cnLinesOn = False
fres.cnLineLabelsOn = False
fres.cnLevelSelectionMode = "ExplicitLevels"
fres.cnFillPalette = "BlueDarkRed18"
fres.cnLevels = np.arange(-5, 6, 1)
plot_cn = Ngl.contour(wks, h_ano, fres)
cnres = Ngl.Resources()
cnres.nglDraw = False
cnres.nglFrame = False
cnres.cnFillOn = False
cnres.cnLinesOn = True
cnres.cnLineLabelsOn = True
cnres.cnInfoLabelOn = False
cnres.sfXArray = lon
cnres.sfYArray = lat
cnres.cnLevelSelectionMode = "ExplicitLevels"
cnres.cnLevels = np.arange(500, 600, 4)
cnres.cnLineColor = "black"
cnres.cnLineThicknessF = 5
plot2 = Ngl.contour(wks, h_cur, cnres)
Ngl.overlay(plot_map, plot_cn)
Ngl.overlay(plot_map, plot2)
#
txres = Ngl.Resources()
txres.txFontHeightF = 0.024
#
Ngl.text_ndc(wks, leftString, 0.3, 0.83, txres)
Ngl.text_ndc(wks, rightString, 0.85, 0.83, txres)
#
Ngl.maximize_plot(wks, plot_map)
Ngl.draw(plot_map)
Ngl.frame(wks)
Ngl.end()
print("Finish darwing 500hPa gh plot plot for " + str(sel_month).zfill(2) +
" " + str(sel_year))
# Set resources for contour plot only ores = set_common_resources() ores.cnFillOpacityF = 1.0 # Fully opaque ores.lbLabelBarOn = False # Turn off labelbar ores.sfMissingValueV = u._FillValue ores.sfXArray = lon['lon|-120:120'] ores.sfYArray = lat['lat|-30:30'] # Use PyNIO's selection syntax to subset the lat/lon area of interest overlay_plot = Ngl.contour(wks, u['time|i1 lat|-30:30 lon|-120:120'], ores) # Overlay the contour plot on the contour/map plot Ngl.overlay(base_plot, overlay_plot) # Drawing the base plot draws both plots Ngl.draw(base_plot) Ngl.frame(wks) # Create the contour/plot again, but with no transparency del bres.cnFillOpacityF plot = Ngl.contour_map(wks, u[1, :, :], bres) # Set resources for a partially transparent polygon. gnres = Ngl.Resources() gnres.gsFillOpacityF = 0.6 # mostly opaque gnres.gsFillColor = "white" lat_box = [-30, -30, 30, 30, -30]
cnres.mpMinLonF = 60 cnres.mpMaxLonF = 180 cnres.mpMinLatF = 0 cnres.mpMaxLatF = 60 cnres.sfXArray = lon cnres.sfYArray = lat cnres.cnLinesOn = False cnres.cnLineLabelsOn = False cnres.cnLevelSelectionMode = "ExplicitLevels" cnres.cnFillPalette = "BlueDarkRed18" cnres.cnLevels = np.arange(500, 600, 4) cn #res.cnMonoLineThickness = False cn #res.cnLineThicknesses = value cnres.cnLineThicknessF = 5 cn # res.cnInfoLabelOn=False cn #res.cnLineColor="black" plot1 = Ngl.contour_map(wks, h_cur, res) Ngl.overlay(plot1, plot) txres = Ngl.Resources() txres.txFontHeightF = 0.024 Ngl.text_ndc(wks, leftString, 0.3, 0.78, txres) Ngl.text_ndc(wks, rightString, 0.9, 0.78, txres) Ngl.frame(wks) Ngl.end() # In[ ]:
mpres.mpMaskAreaSpecifiers = mask_specs # areas to mask (protect)
#
# This set of resources deals with a map labelbar.
#
# Note that this is different from a contour labelbar
# (which we will turn off below), and we have more
# control over it (i.e. we can set the number of boxes,
# the fill # colors, etc) than we do with a contour labelbar.
#
mpres.pmLabelBarDisplayMode = "Always" # Turn on a map labelbar
# Labelbar resources.
mpres.lbLabelStrings = ["Ocean","Land","< 0","0-10","10-20","20-30",\
"30-40","40-50","50-60","60-70","70-80", \
"80-90","90-100","> 100"]
mpres.lbFillColors = mp_fill_colrs + cn_fill_clrs
# Fill resources
cnres.cnFillColors = cn_fill_clrs
cnres.pmLabelBarDisplayMode = "Never" # turn off, b/c map has one turned on
contour = Ngl.contour(wks, z[:, :], cnres)
map = Ngl.map(wks, mpres)
Ngl.overlay(map, contour)
Ngl.draw(map)
Ngl.frame(wks)
Ngl.end()
vcres.vcRefMagnitudeF = 30.0 # define vector ref mag vcres.vcRefLengthF = 0.03 # define length of vec ref vcres.vcMinFracLengthF = 0.3 vcres.vcMinDistanceF = 0.02 vcres.vcRefAnnoOrthogonalPosF = -0.20 vcres.vcRefAnnoFontHeightF = 0.005 vcres.vcLineArrowThicknessF = 2.0 # create vector plot for analysis data and overlay on colour contours level 1 uv_plot1 = ngl.vector(wks, max_shear_u2_u1, max_shear_v2_v1, vcres) PWAT_plot = ngl.contour(wks, PWAT, res) ngl.overlay(LI_plot, uv_plot1) ngl.overlay(LI_plot, PWAT_plot) ngl.maximize_plot(wks, LI_plot) ngl.draw(LI_plot) ngl.frame(wks) ngl.destroy(wks) del res del LI del PWAT del vcres del u1 del v1 del u2 del v2
bres.tmXBLabelFontHeightF = 0.01 # Make these labels smaller.
bres.tmXBMajorOutwardLengthF = 0.0 # Don't draw tickmarks b/c they
bres.tmXBMajorLengthF = 0.0 # were drawn on previous plot.
bres.tmXBLabelDeltaF = 0.6 # Move label away from tickmarks.
bres.tmXBLabelFontColor = "Brown"
bres.tmYROn = False # Turn off right tickmarks.
bres.tmXTOn = False # Turn off top tickmarks.
bres.tmYLOn = False # Turn off left tickmarks.
#---Create the blank plot.
blank = Ngl.blank_plot(wks, bres)
#---Overlay on existing XY plot.
Ngl.overlay(plot, blank)
#---Drawing the original plot also draws the overlaid blank plot
Ngl.draw(plot)
Ngl.frame(wks)
#
# Second plot section.
#
#---Create dummy data for additional curve.
npts2 = 40
y2 = numpy.zeros([npts2], 'f')
for i in range(npts2):
y2[i] = 3.0 - 6. * random.random()
res.tmYRLabelFont = res.tmYLLabelFont
res.tmXBLabelFontHeightF = 0.015
res.tmYLLabelFontHeightF = res.tmXBLabelFontHeightF
res.tmYRLabelFontHeightF = res.tmYLLabelFontHeightF
if ~('res_d' in locals()):
res_d = copy.deepcopy(res)
res_d.xyLineColor = "red"
res_d.xyDashPattern = 2
fire = np.arange(1000.)
xf = np.empty(shape=len(fire), dtype=np.float64)
xf[:] = fire_i
plot = []
plot.append(Ngl.xy(wks, xx, vv[:, i], res))
plot.append(Ngl.xy(wks, xf, fire, res_d))
Ngl.overlay(plot[0], plot[1])
Ngl.draw(plot[0])
Ngl.frame(wks)
del res
Ngl.end()
end_time = cpu_time()
end_time = (end_time - start_time) / 60.0
print(
os.path.basename(__file__) +
" has done!\nTime elapsed: {:.2f}".format(end_time), "mins.")
#-- generate tplot, but don't draw it yet
print("-- map --")
map = Ngl.map(wks, mpres)
#-- generate tplot, but don't draw it yet
print("-- tplot --")
tplot = Ngl.contour(wks, t, tres)
#-- generate plot2, but don't draw it yet
print("-- rplot --")
rplot = Ngl.contour(wks, rhum, rres)
#-- overlay rplot on tplot
print("-- overlay tplot --")
Ngl.overlay(map, tplot)
print("-- overlay rplot --")
Ngl.overlay(map, rplot)
#-- draw the plot
Ngl.draw(map)
#-- write variable long_name and units to the plot
txres = Ngl.Resources()
txres.txFontHeightF = 0.014
Ngl.text_ndc(wks, f.variables["t"].attributes['long_name'], 0.17, 0.88, txres)
Ngl.text_ndc(wks, f.variables["t"].attributes['units'], 0.95, 0.88, txres)
#-- advance the frame
Ngl.frame(wks)
def large_scale_prf(ptype, cseason, ncases, cases, casenames, nsite, lats,
lons, filepath, filepathobs, casedir, dofv):
# ncases, the number of models
# cases, the name of models
# casename, the name of cases
# filepath, model output filepath
# filepathobs, filepath for observational data
# inptrs = [ncases]
if not os.path.exists(casedir):
os.mkdir(casedir)
_Font = 25
interp = 2
extrap = False
mkres = Ngl.Resources()
mkres.gsMarkerIndex = 2
mkres.gsMarkerColor = "Red"
mkres.gsMarkerSizeF = 15.
infiles = ["" for x in range(ncases)]
ncdfs = ["" for x in range(ncases)]
nregions = nsite
varis = [
"CLOUD", "OMEGA", "Q", "CLDLIQ", "THETA", "RELHUM", "U", "CLDICE", "T"
]
varisobs = [
"CC_ISBL", "OMEGA", "SHUM", "CLWC_ISBL", "THETA", "RELHUM", "U",
"CIWC_ISBL", "T"
]
nvaris = len(varis)
cunits = [
"%", "mba/day", "g/kg", "g/kg", "K", "%", "m/s", "g/kg", "m/s", "m/s",
"K", "m"
]
cscale = [100, 864, 1000, 1000, 1., 1, 1, 1000, 1, 1, 1, 1, 1, 1, 1]
cscaleobs = [100, 1, 1, 1000, 1., 1, 1, 1000, 1, 1, 1, 1, 1, 1, 1]
obsdataset = [
"ERAI", "ERAI", "ERAI", "ERAI", "ERAI", "ERAI", "ERAI", "ERAI", "ERAI",
"ERAI"
]
plotlgs = ["" for x in range(nsite)]
for ire in range(0, nsite):
if not os.path.exists(casedir + '/' + str(lons[ire]) + 'E_' +
str(lats[ire]) + 'N'):
os.mkdir(casedir + '/' + str(lons[ire]) + 'E_' + str(lats[ire]) +
'N')
plotname = casedir + '/' + str(lons[ire]) + 'E_' + str(
lats[ire]) + 'N/Largescale_' + str(lons[ire]) + "E_" + str(
lats[ire]) + "N_" + cseason
plotlgs[ire] = 'Largescale_' + str(lons[ire]) + "E_" + str(
lats[ire]) + "N_" + cseason
wks = Ngl.open_wks(ptype, plotname)
Ngl.define_colormap(wks, "GMT_paired")
plot = []
res = Ngl.Resources()
res.nglMaximize = False
res.nglDraw = False
res.nglFrame = False
res.lgPerimOn = False # no box around
res.vpWidthF = 0.30 # set width and height
res.vpHeightF = 0.30
res.tiYAxisString = "Pressure [hPa]"
res.tiMainFont = _Font
res.tmYLLabelFont = _Font
res.tmXBLabelFont = _Font
res.tiYAxisFont = _Font
res.tiXAxisFont = _Font
res.tmXBLabelFontHeightF = 0.01
res.tmXBLabelFontThicknessF = 1.0
res.xyMarkLineMode = 'Lines'
# res.tmXBLabelAngleF = 45
res.xyLineThicknesses = [
3.0, 3.0, 3.0, 3.0, 3.0, 3.0, 3., 3., 3., 3., 3, 3, 3, 3, 3, 3, 3
]
res.xyDashPatterns = np.arange(0, 24, 1)
pres = Ngl.Resources()
pres.nglFrame = False
pres.txString = "Large-scale VAR at" + str(lons[ire]) + "E," + str(
lats[ire]) + "N"
pres.txFont = _Font
pres.nglPanelYWhiteSpacePercent = 5
pres.nglPanelXWhiteSpacePercent = 5
pres.nglPanelTop = 0.88
pres.wkPaperWidthF = 17 # in inches
pres.wkPaperHeightF = 28 # in inches
pres.nglMaximize = True
pres.wkWidth = 10000
pres.wkHeight = 10000
for iv in range(0, nvaris):
if (iv == nvaris - 1):
res.pmLegendDisplayMode = "NEVER"
res.xyExplicitLegendLabels = casenames[:]
res.pmLegendSide = "top"
res.pmLegendParallelPosF = 0.6
res.pmLegendOrthogonalPosF = -0.5
res.pmLegendWidthF = 0.10
res.pmLegendHeightF = 0.10
res.lgLabelFontHeightF = .02
res.lgLabelFontThicknessF = 1.5
res.lgPerimOn = True
else:
res.pmLegendDisplayMode = "NEVER"
if (obsdataset[iv] == "CCCM"):
if (cseason == "ANN"):
fileobs = "/Users/guoz/databank/CLD/CCCm/cccm_cloudfraction_2007-" + cseason + ".nc"
else:
fileobs = "/Users/guoz/databank/CLD/CCCm/cccm_cloudfraction_2007-2010-" + cseason + ".nc"
inptrobs = Dataset(fileobs, 'r')
latobs = inptrobs.variables['lat'][:]
latobs_idx = np.abs(latobs - lats[ire]).argmin()
lonobs = inptrobs.variables['lon'][:]
lonobs_idx = np.abs(lonobs - lons[ire]).argmin()
B = inptrobs.variables[varisobs[iv]][:, latobs_idx, lonobs_idx]
else:
if (varisobs[iv] == "PRECT"):
fileobs = filepathobs + '/GPCP_' + cseason + '_climo.nc'
else:
fileobs = filepathobs + obsdataset[
iv] + '_' + cseason + '_climo.nc'
inptrobs = Dataset(fileobs, 'r')
if (varisobs[iv] == "THETA"):
latobs = inptrobs.variables['lat'][:]
latobs_idx = np.abs(latobs - lats[ire]).argmin()
lonobs = inptrobs.variables['lon'][:]
lonobs_idx = np.abs(lonobs - lons[ire]).argmin()
B = inptrobs.variables['T'][0, :, latobs_idx, lonobs_idx]
pre1 = inptrobs.variables['lev'][:]
for il1 in range(0, len(pre1)):
B[il1] = B[il1] * (1000 / pre1[il1])**0.286
else:
pre1 = inptrobs.variables['lev'][:]
latobs = inptrobs.variables['lat'][:]
latobs_idx = np.abs(latobs - lats[ire]).argmin()
lonobs = inptrobs.variables['lon'][:]
lonobs_idx = np.abs(lonobs - lons[ire]).argmin()
B = inptrobs.variables[varisobs[iv]][0, :, latobs_idx,
lonobs_idx]
B[:] = B[:] * cscaleobs[iv]
for im in range(0, ncases):
ncdfs[im] = './data/' + cases[im] + '_site_location.nc'
infiles[im] = filepath[im] + '/' + cases[
im] + '_' + cseason + '_climo.nc'
inptrs = Dataset(infiles[im], 'r') # pointer to file1
lat = inptrs.variables['lat'][:]
nlat = len(lat)
lon = inptrs.variables['lon'][:]
nlon = len(lon)
lev = inptrs.variables['lev'][:]
nlev = len(lev)
ncdf = Dataset(ncdfs[im], 'r')
n = ncdf.variables['n'][:]
idx_cols = ncdf.variables['idx_cols'][:, :]
if (dofv):
idx_lats = ncdf.variables['idx_coord_lat'][:, :]
idx_lons = ncdf.variables['idx_coord_lon'][:, :]
ncdf.close()
theunits = "[units]"
if (im == 0):
A_field = np.zeros((ncases, nlev), np.float32)
for subc in range(0, n[ire]):
npoint = idx_cols[ire, n[subc] - 1] - 1
if (dofv):
npointlat = idx_lats[ire, 0]
npointlon = idx_lons[ire, 0]
if (varis[iv] == 'THETA'):
if (dofv):
tmp = inptrs.variables['T'][0, :, npointlat,
npointlon]
else:
tmp = inptrs.variables['T'][0, :, npoint]
hyam = inptrs.variables['hyam'][:]
hybm = inptrs.variables['hybm'][:]
if (dofv):
ps = inptrs.variables['PS'][0, npointlat,
npointlon]
else:
ps = inptrs.variables['PS'][0, npoint]
ps = ps
p0 = 100000.0 #CAM uses a hard-coded p0
pre = np.zeros((nlev), np.float32)
for il in range(0, nlev):
pre[il] = hyam[il] * p0 + hybm[il] * ps
tmp[il] = tmp[il] * (100000 / pre[il])**0.286
theunits = str(
cscale[iv]) + "x" + inptrs.variables['T'].units
else:
if (dofv):
tmp = inptrs.variables[varis[iv]][0, :, npointlat,
npointlon]
else:
tmp = inptrs.variables[varis[iv]][0, :, npoint]
theunits = str(cscale[iv]) + "x" + inptrs.variables[
varis[iv]].units
##import pdb; pdb.set_trace()
A_field[im, :] = (A_field[im, :] + tmp[:] / n[ire]).astype(
np.float32)
A_field[im, :] = A_field[im, :] * cscale[iv]
inptrs.close()
res.tiMainString = varis[iv] + " " + theunits
res.trXMinF = min(np.min(A_field[0, :]), np.min(B))
res.trXMaxF = max(np.max(A_field[0, :]), np.max(B))
if (varis[iv] == "THETA"):
res.trXMinF = 270.
res.trXMaxF = 400.
if (varis[iv] == "CLOUD" or varis[iv] == "RELHUM"):
res.trXMinF = 0.
res.trXMaxF = 100.
if (varis[iv] == "T"):
res.trXMinF = 180
res.trXMaxF = 300
if (varis[iv] == "U"):
res.trXMinF = -40
res.trXMaxF = 40
res.trYReverse = True
res.xyLineColors = np.arange(3, 20, 2)
res.xyMarkerColors = np.arange(2, 20, 2)
p = Ngl.xy(wks, A_field, lev, res)
res.trYReverse = False
res.xyLineColors = ["black"]
pt = Ngl.xy(wks, B, pre1, res)
Ngl.overlay(p, pt)
plot.append(p)
Ngl.panel(wks, plot[:], [nvaris / 3, 3], pres)
txres = Ngl.Resources()
txres.txFontHeightF = 0.02
txres.txFont = _Font
Ngl.text_ndc(
wks, "Large-scale VAR at" + str(lons[ire]) + "E," +
str(lats[ire]) + "N", 0.5, 0.92 + ncases * 0.01, txres)
Common_functions.create_legend(wks, casenames, 0.02,
np.arange(3, 20,
2), 0.1, 0.89 + ncases * 0.01)
Ngl.frame(wks)
Ngl.destroy(wks)
return plotlgs
