def main(config_path):
config = {}
with open(config_path) as f_config:
config = json.load(f_config)
fmt_date = 'wrfout_d01_{}'.format(config['fmt']['date'])
path_in = config['output-wrf-raw']
path_out = config['output-gw']
mh = config['coords']['Galway']
out = pd.DataFrame()
f_paths = sorted(g.glob(os.path.join(path_in, 'wrfout_d01*')))
rainmm = {}
rainmmt = {}
snowmm = {}
snowmmt = {}
for i, f_path in enumerate(f_paths):
index = [pd.Timestamp(dt.datetime.strptime(
os.path.basename(f_path), fmt_date
))]
if i == 0:
index0 = index[0]
dayofyear = to_dayofyear(index[0])
tmp = {}
d = xa.open_dataset(f_path).sel(Time=0)[VARS]
# Code to pick out the defined gridcell
diffarraylon=np.asarray(d.XLONG[:]-mh['lon'])
diffarraylat=np.asarray(d.XLAT[:]-mh['lat'])
diffarrayabs=xau.sqrt(xau.square(diffarraylon)+xau.square(diffarraylat))
#note: the array is stored in array dims [lat, lon]
ixlat,ixlon=np.where(diffarrayabs == np.min(diffarrayabs))
idx = {
'x': ixlon,
'y': ixlat
}
tmp_p, tmp_rh = pressure_rh(d)
d = d.isel(west_east=idx['x'], south_north=idx['y'])
# windspeed
tmp['dayofyear'] = dayofyear
# tmp['winddirection_deg'] = np.asscalar(xau.rad2deg(
# xau.arctan2(-d.U10, -d.V10)
# ).values)
tmp['pressure_hPa'] = tmp_p[idx['y'], idx['x']]
tmp['qcloud'] = d.isel(bottom_top=0).QCLOUD[0,:].values
tmp['qgraup'] = d.isel(bottom_top=0).QGRAUP[0,:].values
tmp['qicd'] = d.isel(bottom_top=0).QICE[0,:].values
tmp['qrain'] = d.isel(bottom_top=0).QRAIN[0,:].values
tmp['qsnow'] = d.isel(bottom_top=0).QSNOW[0,:].values
if i == 0:
rainmm[i] = np.asscalar((d.RAINNC.values + d.RAINC.values))
rainmmt[i] = np.asscalar((d.RAINNC.values + d.RAINC.values))
snowmm[i] = np.asscalar(d.SNOWNC.values)
snowmmt[i] = np.asscalar(d.SNOWNC.values)
else:
rainmmt[i] = np.asscalar((d.RAINNC.values + d.RAINC.values))
rainmm[i] = np.asscalar(d.RAINNC.values + d.RAINC.values)-rainmmt[i-1]
snowmmt[i] = np.asscalar(d.SNOWNC.values)
snowmm[i] = np.asscalar(d.SNOWNC.values)-snowmmt[i-1]
tmp['rain_mm'] = rainmm[i]
tmp['relativehumidity_percent'] = tmp_rh[idx['y'], idx['x']]
tmp['temperature2m_C'] = np.asscalar((d.T2 - 273.15).values)
tmp['winddirection_deg'] = np.asscalar(
270 - xau.rad2deg(xau.arctan2(d.V10, d.U10)).values) % 360
tmp['windspeed_mPs'] = np.asscalar(xau.sqrt(d.U10**2 + d.V10**2).values)
tmp['zcldfra'] = d.isel(bottom_top=0).CLDFRA[0,:].values
tmp['znox'] = d.isel(bottom_top=0).no2[0,:].values + d.isel(bottom_top=0).no[0,:].values
tmp['zo3'] = d.isel(bottom_top=0).o3[0,:].values*1000.0
tmp['zpm25'] = d.isel(bottom_top=0).PM2_5_DRY[0,:].values
tmp['zpm10'] = d.isel(bottom_top=0).PM10[0,:].values
tmp['zbc1'] = d.isel(bottom_top=0).BC1[0,:].values
tmp['zbc2'] = d.isel(bottom_top=0).BC2[0,:].values
tmp['zoc1'] = d.isel(bottom_top=0).OC1[0,:].values
tmp['zoc2'] = d.isel(bottom_top=0).OC2[0,:].values
tmp['zso2'] = d.isel(bottom_top=0).so2[0,:].values
tmp['zpblh'] = np.asscalar(d.PBLH.values)
tmp['zuv_index'] = np.sum((d.o3*d.PB/6950.0).values)
tmp = pd.DataFrame(tmp, index=index).sort_index(axis=1)
out = pd.concat((out, tmp))
print('done {}'.format(f_path))
#'BC1', 'BC2', 'OC1', 'OC2']
# out = out.rename({'pressure_sea_hPa': 'pressure_hPa'})
# (out[['dayofyear', 'pressure_hPa', 'relativehumidity_percent',
# 'temperature2m_C', 'winddirection_deg', 'windspeed_mPs']]
# .to_dataframe()
# .to_csv(os.path.join(path_out,
# '{}.csv'.format(index0.strftime('%Y%m%d%H%M')))))
out.to_csv(os.path.join(
path_out,
'{}.csv'.format(index0.strftime('%Y%m%d%H%M'))
))
E_PM25_ugperm2pers = E_PM_10_ugperm2pers * .75
E_BC_1_ugperm2pers = E_PM_10_ugperm2pers * .13
E_OC_DOM_ugperm2pers = E_PM25_ugperm2pers * .65
E_OC_25_10_ugperm2pers = E_PM_10_ugperm2pers * .65 - E_OC_DOM_ugperm2pers
E_OIN_10_ugperm2pers = E_PM_10_ugperm2pers * .22
E_OIN_25_ugperm2pers = E_PM25_ugperm2pers * .22
#link_dest=os.path.join(emis_preproc_dir, 'wrfinput_d01')
loc_lon, loc_lat = city_list[ss]['lon'], city_list[ss]['lat']
## Get the index of dublin grid cell from the wrfinput file as the produced wrfchemi files don't contain xlat, xlon
lat = np.asarray(dwrf.XLAT[0, :, :])
longg = np.asarray(dwrf.XLONG[0, :, :])
diffarray_lats = lat[:] - [loc_lat]
diffarray_lons = longg[:] - [loc_lon]
diffarrayabs = xu.sqrt(
xu.square(diffarray_lons) + xu.square(diffarray_lats))
ixlat, ixlon = np.where(diffarrayabs == np.min(diffarrayabs))
idxx = {'x': ixlon, 'y': ixlat}
lat_ix = np.argmin((arraylat - loc_lat)**2)
lon_ix = np.argmin((arraylon - loc_lon)**2)
print(lon_ix, lat_ix) #verified as corresponding with the location
idx = {'x': lon_ix, 'y': lat_ix}
dub_temp = data.var235[:, lat_ix, lon_ix].values - 273.15
print('Todays temp in ', ss, ' ', dub_temp)
## firstly, apply fix to NO, NO2 ratios
ds.variables['E_NO2'][:,
0, :, :] = ds.variables['E_NO'][:,
0, :, :] * 0.22
ds.variables['E_NO'][:,
0, :, :] = ds.variables['E_NO'][:, 0, :, :] * 0.88
if dub_temp < 12.0:
print('di', i, f_path)
d = xa.open_dataset(f_path)[VARS]
index = [
pd.Timestamp(
dt.datetime.strptime(os.path.basename(f_path),
'd01_%Y%m%d%H00.nc'))
]
if i == 0:
index0 = index[0]
dayofyear = to_dayofyear(index[0])
mh = coords[kk]
diffarraylon = np.asarray(d['lon'][:] - mh['lon'])
diffarraylat = np.asarray(d['lat'][:] - mh['lat'])
diffarrayabs = xau.sqrt(
xau.square(diffarraylon) + xau.square(diffarraylat))
#note: the array is stored in array dims [lat, lon]
ixlat, ixlon = np.where(diffarrayabs == np.min(diffarrayabs))
idx = {'x': ixlon, 'y': ixlat}
d = d.isel(x=idx['x'], y=idx['y'])
tmp['dayofyear'] = dayofyear
for vv in VARS:
if d[vv].ndim == 3:
tmp[vv] = d[vv][0, 0, 0].values
else:
tmp[vv] = d[vv][0, 0].values
#tmp[vv] = 0.0
# first: put in the rain loop
#if vv=='rain':
#print('look out! its the', vv, 'vbl')
wrfrundir = os.path.join(work_root_dir, 'WRFV3/run')
# link the operational wrfinput file to the emission preprocessor directory. This is necessary to generate the hourly emissions.
wrfinfile = os.path.join(wrfrundir, 'wrfinput_d01')
link_dest = os.path.join(emis_preproc_dir, 'wrfinput_d01')
dub_lon, dub_lat = -6.249910, 53.426448
dwrf = xr.open_dataset(wrfinfile)
# Get the index of dublin grid cell from the wrfinput file as the produced wrfchemi files don't contain xlat, xlon
lat = np.asarray(dwrf.XLAT[0, :, :])
longg = np.asarray(dwrf.XLONG[0, :, :])
diffarray_lats = lat[:] - [dub_lat]
diffarray_lons = longg[:] - [dub_lon]
print(diffarray_lats.shape, diffarray_lons.shape)
#diffarraylon1=longg-dub_lon
diffarrayabs = xu.sqrt(xu.square(diffarray_lons) + xu.square(diffarray_lats))
ixlat, ixlon = np.where(diffarrayabs == np.min(diffarrayabs))
idxx = {'x': ixlon, 'y': ixlat}
print(idxx)
if os.path.exists(link_dest):
os.remove(link_dest)
print('removed the input file')
os.symlink(wrfinfile, link_dest)
#ii='datestr'
ii = cdo.showdate(input=os.path.join(work_root_dir, 'WRFV3/run/wrfinput_d01'))
i = dt.datetime.strptime(ii[0], '%Y-%m-%d')
#i=dt.datetime.strptime('2018-11-02', '%Y-%m-%d')
#2. Update the namelist with current date
upd_file = os.path.join(emis_preproc_dir, 'namelist')
def main(config_path):
config = {}
with open(config_path) as f_config:
config = json.load(f_config)
#fmt_date = 'd01_{}'.format('%Y%m%d%H00.nc')
path_in = ('/mnt/raid/wrf-chem/wrfchem_v39_cri/data_back/output-wrf')
path_out = '../data/output-mh-archive'
mh = config['coords']['Mace Head']
out = pd.DataFrame()
f_paths = sorted(g.glob(os.path.join(path_in, 'd01_*')))
rainmm = {}
rainmmt = {}
snowmm = {}
snowmmt = {}
ddstr = {}
for i, f_path in enumerate(f_paths):
index = [
pd.Timestamp(
dt.datetime.strptime(os.path.basename(f_path),
'd01_%Y%m%d%H00.nc'))
]
print('innnndex', index)
print(i, f_path)
if i == 0:
index0 = index[0]
print('1st inx', index0)
dayofyear = to_dayofyear(index[0])
tmp = {}
#d = xa.open_dataset(f_path).sel(time=0)[VARS]
d = xa.open_dataset(f_path)[VARS]
# Code to pick out the defined gridcell
diffarraylon = np.asarray(d['lon'][:] - mh['lon'])
diffarraylat = np.asarray(d['lat'][:] - mh['lat'])
diffarrayabs = xau.sqrt(
xau.square(diffarraylon) + xau.square(diffarraylat))
#note: the array is stored in array dims [lat, lon]
ixlat, ixlon = np.where(diffarrayabs == np.min(diffarrayabs))
idx = {'x': ixlon, 'y': ixlat}
#tmp_p, tmp_rh = pressure_rh(d)
print(d.dims)
d = d.isel(x=idx['x'], y=idx['y'])
# windspeed
tmp['dayofyear'] = dayofyear
for vv in VARS:
#print('variable', vv)
# first: put in the rain loop
if vv == 'rain':
print('look out! its the', vv, 'vbl')
print(d[vv].shape)
cmmd = ('stat -c %y ' + str(f_path))
dstr = str(os.system(cmmd))
ddstr[i] = os.popen(cmmd).read()[:10]
if i == 0:
print('the first datapoint')
rainmm[i] = d['rain'][0, 0].values
tmp[vv] = np.nan
print(tmp[vv])
else:
if ddstr[i - 1] == ddstr[i]:
print('same dataset')
print(d[vv].shape)
rainmm[i] = d['rain'][0, 0] - rainmm[i - 1]
tmp[vv] = rainmm[i].values
else:
print(
'new dataset alert!!!!!! must disregard current rain rate value'
)
rainmm[i] = d['rain'][0, 0].values
tmp[vv] = np.nan
print('fuuuuuuuuuuuuuuucl', tmp[vv])
#print('uuup', ddstr)
## if i=0: then disregard the first value, and subtract all subsequent. if new date (find out by executing stat -c '%y' d01_201909110000.nc), then deal with it the same way etc.
else:
if d[vv].ndim == 3:
#print(d[vv].shape)
tmp[vv] = d[vv][0, 0, 0].values
else:
#print(vv, d[vv].shape)
tmp[vv] = d[vv][0, 0].values
## extract windspeed and wind direction
#winddirection_deg = np.asscalar(
#270 - xau.rad2deg(xau.arctan2(d.V10, d.U10)).values) % 360
tmp['winddirection_deg'] = np.asscalar(
270 - xau.rad2deg(xau.arctan2(d.v10, d.u10)).values) % 360
tmp['windspeed_mPs'] = np.asscalar(
xau.sqrt(d.u10**2 + d.v10**2).values)
tmp = pd.DataFrame(tmp, index=index).sort_index(axis=1)
#print(tmp.keys())
#for kk in tmp.keys():
#print(kk, tmp[kk].values)
out = pd.concat((out, tmp))
#print('done {}'.format(f_path))
#'BC1', 'BC2', 'OC1', 'OC2']
# out = out.rename({'pressure_sea_hPa': 'pressure_hPa'})
# (out[['dayofyear', 'pressure_hPa', 'relativehumidity_percent',
# 'temperature2m_C', 'winddirection_deg', 'windspeed_mPs']]
# .to_dataframe()
# .to_csv(os.path.join(path_out,
# '{}.csv'.format(index0.strftime('%Y%m%d%H%M')))))
out.to_csv(
os.path.join(path_out, '{}.csv'.format(index0.strftime('%Y%m%d%H%M'))))
