def saveNetcdf():
modis_folder = '/users/global/cornkle/data/OBS/MSG_LSTA/lsta_raw_binary'
pool = multiprocessing.Pool(processes=7)
files = glob.glob(modis_folder + '/lsta_daily_2*.gra')
for f in files:
ds = rewrite_data.rewriteLSTA_toNetcdf(f, write=True)
def saveNetcdf():
modis_folder = '/users/global/cornkle/data/OBS/MSG_LSTA/lsta_raw_binary'
pool = multiprocessing.Pool(processes=7)
files = glob.glob(modis_folder+'/lsta_daily_2*.gra')
for f in files:
ds = rewrite_data.rewriteLSTA_toNetcdf(f, write=True)
def saveNetcdf_blobs():
modis_folder = '/users/global/cornkle/data/OBS/MSG_LSTA/lsta_raw_binary_new' #'/users/global/cornkle/data/OBS/MSG_LSTA/lsta_raw_binary'
td = pd.Timedelta('16 hours')
files = glob.glob(modis_folder + '/lsta_daily_*.gra') #2*.gra')
msgfile = '/users/global/cornkle/MCSfiles/blob_map_allscales_-50_JJAS_points_dominant_daily.nc'
msg = xr.open_dataarray(msgfile)
for f in files:
ds, out = rewrite_data.rewriteLSTA_toNetcdf(f, write=False)
ds['LSTA'].values[ds['LSTA'].values < -800] = np.nan
ds['LSTA'].values = ds['LSTA'].values - np.nanmean(ds['LSTA'].values)
m = msg[(msg['time'] - td).values == ds['time'].values]
m = m.sel(lat=slice(10.3, 19.7), lon=slice(-9.7, 9.7))
ds['cell'] = ds['LSTA'].copy() * np.nan
pos = np.where(np.isfinite(m.values))
if np.isnan(np.nansum(m.values)):
print('No blobs found')
try:
ds.to_netcdf(out)
except OSError:
print('Did not find ' + out)
print('Out directory not found')
print('Wrote ' + out)
continue
for y, x in zip(pos[1], pos[2]):
lat = m['lat'][y]
lon = m['lon'][x]
point = ds.sel(lat=lat, lon=lon, method='nearest')
plat = point['lat'].values
plon = point['lon'].values
xpos = np.where(ds['lon'].values == plon)
xpos = int(xpos[0])
ypos = np.where(ds['lat'].values == plat)
ypos = int(ypos[0])
ds['cell'].values[0, ypos, xpos] = m.values[0, y, x]
try:
ds.to_netcdf(out)
except OSError:
print('Did not find ' + out)
print('Out directory not found')
print('Wrote ' + out)
def saveNetcdf_blobs():
modis_folder = cnst.network_data + 'data/OBS/MSG_LSTA/lsta_raw_binary_new'#'/users/global/cornkle/data/OBS/MSG_LSTA/lsta_raw_binary'
td = pd.Timedelta('16 hours')
files = glob.glob(modis_folder + '/lsta_daily_*.gra') #2*.gra')
msgfile = cnst.network_data + 'MCSfiles/blob_map_allscales_-50_JJAS_points_dominant_daily.nc'
msg = xr.open_dataarray(msgfile)
ll = np.load(cnst.network_data + 'data/OBS/MSG_LSTA/lsta_728_348_lat_lon.npz')
blat = ll['lat']
blon = ll['lon']
latmin = np.min(blat)
latmax = np.max(blat)
lonmin = -9.98 #np.min(blon)
lonmax = 9.98 #np.max(blon)
dist = np.round(np.float(np.mean(blon[0,:][1::] - blon[0,:][0:-1])), decimals=4)
lat_regular = np.arange(latmin + 10*dist, latmax - 10*dist , dist)
lon_regular = np.arange(lonmin , lonmax , dist)
inds, weights, shape = uint.interpolation_weights(blon, blat, lon_regular, lat_regular)
int_input = {'x' : lon_regular, 'y' : lat_regular, 'inds' : inds, 'weights' : weights, 'shape' : shape}
for f in files:
ds,out = rewrite_data.rewriteLSTA_toNetcdf(f, int_input)
ds['LSTA'].values[ds['LSTA'].values < -800] = np.nan
#ds['LSTA'].values = ds['LSTA'].values - np.nanmean(ds['LSTA'].values)
m = msg[(msg['time']-td).values==ds['time'].values]
m = m.sel(lat=slice(10.3,19.7), lon=slice(-9.7,9.7))
ds['cell'] = ds['LSTA'].copy()*np.nan
pos = np.where(np.isfinite(m.values))
if np.isnan(np.nansum(m.values)):
print('No blobs found')
try:
ds.to_netcdf(out)
except OSError:
print('Did not find ' + out)
print('Out directory not found')
print('Wrote ' + out)
continue
for y, x in zip(pos[1], pos[2]):
lat = m['lat'][y]
lon = m['lon'][x]
point = ds.sel(lat=lat, lon=lon, method='nearest')
plat = point['lat'].values
plon = point['lon'].values
xpos = np.where(ds['lon'].values == plon)
xpos = int(xpos[0])
ypos = np.where(ds['lat'].values == plat)
ypos = int(ypos[0])
ds['cell'].values[0,ypos,xpos] = m.values[0,y,x]
try:
ds.to_netcdf(out)
except OSError:
print('Did not find ' + out)
print('Out directory not found')
print('Wrote ' + out)
def saveNetcdf_blobs():
modis_folder = cnst.network_data + 'data/OBS/MSG_LSTA/lsta_raw_binary_new'#'/users/global/cornkle/data/OBS/MSG_LSTA/lsta_raw_binary'
td = pd.Timedelta('16 hours')
files = glob.glob(modis_folder + '/lsta_daily_*.gra') #2*.gra')
msgfile = cnst.network_data + 'MCSfiles/blob_map_allscales_-50_JJAS_points_dominant_daily.nc'
msg = xr.open_dataarray(msgfile)
ll = np.load(cnst.network_data + 'data/OBS/MSG_LSTA/lsta_728_348_lat_lon.npz')
blat = ll['lat']
blon = ll['lon']
latmin = np.min(blat)
latmax = np.max(blat)
lonmin = -9.98 #np.min(blon)
lonmax = 9.98 #np.max(blon)
dist = np.round(np.float(np.mean(blon[0,:][1::] - blon[0,:][0:-1])), decimals=4)
lat_regular = np.arange(latmin + 10*dist, latmax - 10*dist , dist)
lon_regular = np.arange(lonmin , lonmax , dist)
inds, weights, shape = uint.interpolation_weights(blon, blat, lon_regular, lat_regular)
int_input = {'x' : lon_regular, 'y' : lat_regular, 'inds' : inds, 'weights' : weights, 'shape' : shape}
for f in files:
ds,out = rewrite_data.rewriteLSTA_toNetcdf(f, int_input)
ds['LSTA'].values[ds['LSTA'].values < -800] = np.nan
#ds['LSTA'].values = ds['LSTA'].values - np.nanmean(ds['LSTA'].values)
m = msg[(msg['time']-td).values==ds['time'].values]
m = m.sel(lat=slice(10.3,19.7), lon=slice(-9.7,9.7))
ds['cell'] = ds['LSTA'].copy()*np.nan
pos = np.where(np.isfinite(m.values))
if np.isnan(np.nansum(m.values)):
print('No blobs found')
try:
ds.to_netcdf(out)
except OSError:
print('Did not find ' + out)
print('Out directory not found')
print('Wrote ' + out)
continue
for y, x in zip(pos[1], pos[2]):
lat = m['lat'][y]
lon = m['lon'][x]
point = ds.sel(lat=lat, lon=lon, method='nearest')
plat = point['lat'].values
plon = point['lon'].values
xpos = np.where(ds['lon'].values == plon)
xpos = int(xpos[0])
ypos = np.where(ds['lat'].values == plat)
ypos = int(ypos[0])
ds['cell'].values[0,ypos,xpos] = m.values[0,y,x]
try:
ds.to_netcdf(out)
except OSError:
print('Did not find ' + out)
print('Out directory not found')
print('Wrote ' + out)