PSM for ice core d18O

In this tutorial, we introduce the PSM for ice core d18O in cfr.

%load_ext autoreload
%autoreload 2

import cfr
print(cfr.__version__)

0.6.0

Data preparation

Proxy

pdb = cfr.ProxyDatabase().fetch('PAGES2kv2')

pobj = pdb.records['Arc_011']
fig, ax = pobj.plot()

Model

%%time

model_tas = cfr.ClimateField().fetch('iCESM_past1000historical/tas')
model_pr = cfr.ClimateField().fetch('iCESM_past1000historical/pr')
model_psl = cfr.ClimateField().fetch('iCESM_past1000historical/psl')
model_d18O = cfr.ClimateField().fetch('iCESM_past1000historical/d18O')

>>> The target file seems existed at: ./data/tas_sfc_Amon_iCESM_past1000historical_085001-200512.nc . Loading from it instead of downloading ...
>>> The target file seems existed at: ./data/pr_sfc_Amon_iCESM_past1000historical_085001-200512.nc . Loading from it instead of downloading ...
>>> The target file seems existed at: ./data/psl_sfc_Amon_iCESM_past1000historical_085001-200512.nc . Loading from it instead of downloading ...
>>> The target file seems existed at: ./data/d18O_sfc_Amon_iCESM_past1000historical_085001-200512.nc . Loading from it instead of downloading ...
CPU times: user 105 ms, sys: 14.1 ms, total: 119 ms
Wall time: 130 ms

Get climate data for a specific ProxyRecord

%%time

pobj.del_clim()
pobj.get_clim(model_tas, tag='model')
pobj.get_clim(model_pr, tag='model')
pobj.get_clim(model_psl, tag='model')
pobj.get_clim(model_d18O, tag='model')

CPU times: user 11.9 ms, sys: 299 ms, total: 311 ms
Wall time: 3.47 s

pobj.clim['model.d18O'].da

<xarray.DataArray 'd18O' (time: 13872)>
array([-24.155558, -30.799866, -27.157188, ..., -27.73261 , -28.163136,
       -29.739498], dtype=float32)
Coordinates:
  * time     (time) object 0850-01-17 00:00:00 ... 2005-12-17 00:00:00
    lat      float32 72.95
    lon      float32 322.5
Attributes:
    long_name:  Ratio of the total mass precip rate H2O
    units:      permil

xarray.DataArray

'd18O'

• time: 13872

• -24.16 -30.8 -27.16 -28.47 -22.71 ... -26.09 -27.73 -28.16 -29.74

  array([-24.155558, -30.799866, -27.157188, ..., -27.73261 , -28.163136,
         -29.739498], dtype=float32)

• Coordinates: (3)
  • time
    (time)
    object
    0850-01-17 00:00:00 ... 2005-12-...

    description :

    time

    long_name :

    time

    standard_name :

    time

    array([cftime.DatetimeNoLeap(850, 1, 17, 0, 0, 0, 0, has_year_zero=True),
           cftime.DatetimeNoLeap(850, 2, 14, 0, 0, 0, 0, has_year_zero=True),
           cftime.DatetimeNoLeap(850, 3, 17, 0, 0, 0, 0, has_year_zero=True), ...,
           cftime.DatetimeNoLeap(2005, 10, 17, 0, 0, 0, 0, has_year_zero=True),
           cftime.DatetimeNoLeap(2005, 11, 16, 0, 0, 0, 0, has_year_zero=True),
           cftime.DatetimeNoLeap(2005, 12, 17, 0, 0, 0, 0, has_year_zero=True)],
          dtype=object)

  • lat
    ()
    float32
    72.95

    description :

    latitude

    long_name :

    latitude coordinate

    standard_name :

    latitude

    units :

    degrees_north

    array(72.947365, dtype=float32)

  • lon
    ()
    float32
    322.5

    description :

    longitude

    long_name :

    longitude coordinate

    standard_name :

    longitude

    units :

    degrees_east

    array(322.5, dtype=float32)

• Indexes: (1)
  • time
    PandasIndex

    PandasIndex(CFTimeIndex([0850-01-17 00:00:00, 0850-02-14 00:00:00, 0850-03-17 00:00:00,
                 0850-04-16 00:00:00, 0850-05-17 00:00:00, 0850-06-16 00:00:00,
                 0850-07-17 00:00:00, 0850-08-17 00:00:00, 0850-09-16 00:00:00,
                 0850-10-17 00:00:00,
                 ...
                 2005-03-17 00:00:00, 2005-04-16 00:00:00, 2005-05-17 00:00:00,
                 2005-06-16 00:00:00, 2005-07-17 00:00:00, 2005-08-17 00:00:00,
                 2005-09-16 00:00:00, 2005-10-17 00:00:00, 2005-11-16 00:00:00,
                 2005-12-17 00:00:00],
                dtype='object', length=13872, calendar='noleap', freq='None'))

• Attributes: (2)

  long_name :

  Ratio of the total mass precip rate H2O

  units :

  permil

Create a PSM object

mdl = cfr.psm.Ice_d18O(pobj)

%%time
pp = mdl.forward(nproc=1)

Warning: convolution kernel length (zp) is approaching that of half the length of timeseries. Kernel being clipped.

100%|██████████| 10602/10602 [00:03<00:00, 3000.93it/s]

CPU times: user 3.51 s, sys: 1.37 s, total: 4.87 s
Wall time: 4.88 s

%%time
pp = mdl.forward()

Warning: convolution kernel length (zp) is approaching that of half the length of timeseries. Kernel being clipped.
CPU times: user 1.39 s, sys: 342 ms, total: 1.73 s
Wall time: 1.8 s

fig, ax = pp.plot()