def parse_date(year, datestr):
months = ['January', 'February', 'March', 'April', 'May', 'June',
'July', 'August', 'September', 'October', 'November',
'December']
month_lookup = {month : m + 1 for m, month in enumerate(months)}
day, month = datestr.split()
month = month_lookup[month]
day = int(day)
jday = atm.mmdd_to_jday(month, day, year)
return jday
def onset_OCI(u, latlon = (5, 15, 40, 80), mmdd_thresh=(6,1),
ndays=7, yearnm='Year', daynm='Day'):
"""Return monsoon Onset Circulation Index.
Parameters
----------
u : xray.DataArray
850 hPa zonal wind.
latlon : 4-tuple of floats, optional
Tuple of (lat1, lat2, lon1, lon2) defining South Arabian Sea
region to average over.
mmdd_thres : 2-tuple of ints, optional
Tuple of (month, day) defining climatological mean onset date
to use for threshold value of u.
ndays : int, optional
Number of consecutive days threshold must be exceeded to
define onset.
yearnm, daynm : str, optional
Name of year and day dimensions in DataArray
Returns
-------
oci : xray.Dataset
OCI daily timeseries for each year and monsoon onset day for
each year.
Reference
---------
Wang, B., Ding, Q., & Joseph, P. V. (2009). Objective Definition
of the Indian Summer Monsoon Onset. Journal of Climate, 22(12),
3303-3316.
"""
days = atm.get_coord(u, coord_name=daynm)
years = atm.get_coord(u, coord_name=yearnm)
nyears = len(years)
# Average over South Arabian Sea region
lat1, lat2, lon1, lon2 = latlon
ubar = atm.mean_over_geobox(u, lat1, lat2, lon1, lon2)
# Find values at climatological onset
m0, d0 = mmdd_thresh
d0 = [atm.mmdd_to_jday(m0, d0, year) for year in years]
u0 = [ubar.sel(**{daynm : day, yearnm : year}).values
for year, day in zip(years, d0)]
u0 = np.array(u0).flatten()
uthreshold = np.mean(u0)
# Find first day when OCI exceeds threshold and stays above the
# threshold for consecutive ndays
def onset_day(tseries, uthreshold, ndays, daynm):
above = (tseries.values > uthreshold)
d0 = above.argmax()
while not above[d0:d0+ndays].all():
d0 += 1
return tseries[daynm].values[d0]
# Find onset day for each year
onset = [onset_day(ubar[y], uthreshold, ndays, daynm)
for y in range(nyears)]
# Pack into dataset
oci = xray.Dataset()
oci['tseries'] = ubar
oci['onset'] = xray.DataArray(onset, coords={yearnm : years})
oci.attrs['latlon'] = latlon
oci.attrs['mmdd_thresh'] = mmdd_thresh
oci.attrs['ndays'] = ndays
return oci
for nm in index_pts.data_vars:
ind = index[nm].sel(year=index_pts['year'])
pts_reg[nm] = atm.regress_field(index_pts[nm], ind, axis=0)
pts_reg[nm]['pts_mask'] = (pts_reg[nm]['p'] >= 0.05)
# Mask out grid points where CHP index is ill-defined
def applymask(ds, mask_in):
for nm in ds.data_vars:
mask = atm.biggify(mask_in, ds[nm], tile=True)
vals = np.ma.masked_array(ds[nm], mask=mask).filled(np.nan)
ds[nm].values = vals
return ds
if ptsmaskfile is not None:
fracmin = 0.5
day1 = atm.mmdd_to_jday(6, 1)
day2 = atm.mmdd_to_jday(9, 30)
with xray.open_dataset(ptsmaskfile) as ds:
pcp = ds['PREC'].sel(lat=index_pts.lat).sel(lon=index_pts.lon).load()
pcp_ssn = atm.subset(pcp, {'day' : (day1, day2)})
pcp_frac = pcp_ssn.sum(dim='day') / pcp.sum(dim='day')
mask = pcp_frac < fracmin
index_pts = applymask(index_pts, mask)
for key in pts_reg:
pts_reg[key] = applymask(pts_reg[key], mask)
# MFC budget
with xray.open_dataset(mfcbudget_file) as mfc_budget:
mfc_budget.load()
mfc_budget = mfc_budget.rename({'DWDT' : 'dw/dt'})
mfc_budget['P-E'] = mfc_budget['PRECTOT'] - mfc_budget['EVAP']