In [1]:

import sys
sys.path.append('/data/nsoontie/MEOPAR/mixing-paper/analysis')

import numpy as np
import xarray as xr
import matplotlib.pyplot as plt
import pandas as pd
import datetime

import mixing
from nowcast import analyze

%matplotlib inline

Load Nowcast-green¶

In [2]:

sdt = datetime.datetime(2016,4,1)
edt = datetime.datetime(2016,7,28)
xs=slice(150,300)
ys=slice(200,400)
files = analyze.get_filenames(sdt, edt, '1d', 'grid_T', '/results/SalishSea/nowcast-green/')
d = xr.open_dataset(files[0]).isel(x=xs, y=ys)
for f in files[1:]:
    with xr.open_dataset(f).isel(x=xs, y=ys) as d1:
        d =xr.concat([d,d1],dim='time_counter',data_vars='minimal')

In [3]:

salt_surface = d.sel(deptht=0,method='nearest').vosaline
salt_surface.resample('7D', dim='time_counter').plot(col='time_counter', col_wrap=6,x='x',y='y',
                      vmin=20,vmax=33)

Out[3]:

<xarray.plot.facetgrid.FacetGrid at 0x7f20fff270f0>

In [4]:

i=200
j=260
fig,ax= plt.subplots(1,1)
d.sel(deptht=0,method='nearest').vosaline.mean(dim='time_counter').plot( vmin=20,vmax=33,ax=ax)
ax.plot(i,j,'or')

Out[4]:

[<matplotlib.lines.Line2D at 0x7f20fd0edd30>]

In [5]:

fig,ax= plt.subplots(1,1,figsize=(10,5))
d.sel(x=i,y=j).vosaline.plot(x='time_counter',y='deptht',vmin=28,vmax=33,ax=ax)
ax.set_ylim([140,0])

Out[5]:

(140, 0)

In [6]:

fig,ax=plt.subplots(1,1)
profile = d.sel(x=i,y=j)
for t in np.arange(profile.vosaline.shape[0]):
    ax.plot(profile.vosaline.values[t,:], profile.deptht.values[:], )
ax.set_ylim([140,0])
ax.set_xlim([28,34])
ax.grid()
ax.set_ylabel('Depth [m]')
ax.set_xlabel('Salinity [g/kg]')

Out[6]:

<matplotlib.text.Text at 0x7f20fda77940>

We are actually pretty stratified in the daily averages. Maybe the other case was fully mixed because it didn't have the Hollingsworth correction...

August 2015 casts from PSF had about 33.5 psu at 140 m at 31 to 31.5 psu at surface.

In [ ]: