/
s3_sensity_of_t2m_to_SST.py
219 lines (183 loc) · 9.69 KB
/
s3_sensity_of_t2m_to_SST.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
#!/usr/bin/env python
# coding: utf-8
# 计算中国区域气温均值对不同地区海温异常的敏感性分布
# 中国区域的均值序列，和海温的每个格点的序列
# 以季节为时间单位，错开不同的时次求相关
# calculate the area-weighted mean of t2m in china
# and the correlations between the mean and the SST in the pacific
# given the different lead 4 seasons(unit)
import numpy as np
import xarray as xr
import matplotlib.pyplot as plt
import scipy.stats as stats
import sys
import warnings
warnings.filterwarnings('ignore')
import seaborn as sns
import cartopy.crs as ccrs
import cartopy.feature as cfeat
from cartopy.mpl.ticker import LongitudeFormatter, LatitudeFormatter
# load data
# Read SST using the xarray module.
sst_o = xr.open_dataset('./data/ersst.v5.1m.1854-2019.nc')['sst'].squeeze()
t2m_o = xr.open_dataset('./data/air.2m.mon.mean.nc')['air']
t2m = t2m_o[:,::-1,:]
sst_time = sst_o.isel(time = slice(1128,1992))
t2m_time = t2m.isel(time = slice(0,864))
# print(sst_time)
# print(t2m_time)
def mask(ds, label='land'):
landsea = xr.open_dataset('./data/landsea.nc')['lsm']
landsea = landsea[:,::-1,:].squeeze()
# interpolation of the leandsea data to the ds
landsea = landsea.interp(latitude=ds.lat.values, longitude=ds.lon.values)
ds.coords['mask'] = (('lat', 'lon'), landsea.values)
if label == 'land':
ds = ds.where(ds.mask < 0.8,np.nan)
elif label == 'ocean':
ds = ds.where(ds.mask > 0.2,np.nan)
return ds
t2m_mk = mask(t2m_time,label = 'ocean')
# select the china area
t2m_CHN = t2m_mk.sel(lat = slice(20,55),lon = slice(105,125))
# calculate the anomalies by removing the annual cycle
t2ma = t2m_CHN.groupby('time.month') - t2m_CHN.groupby('time.month').mean('time')
lat_t2m = t2m_CHN.coords['lat']
lon_t2m = t2m_CHN.coords['lon']
time = np.arange(1948,2019,1)
t2m_DJF = t2ma.sel(time=t2ma['time.season'] == 'DJF')
t2m_y = np.zeros((71,len(t2ma.coords['lat']),len(t2ma.coords['lon'])),np.float)
print(t2m_y.shape)
for ii in range(t2m_y.shape[0]):
t2m_temp = t2m_DJF[3*ii+2:3*ii+5,:,:]
t2m_y[ii,:,:] = np.mean(t2m_temp,axis = 0)
t2m_xr = xr.DataArray(t2m_y, dims=('time', 'lat', 'lon'), coords={'time': time,'lat':lat_t2m,'lon':lon_t2m})
# print(t2m_xr)
# calculate the area-weighted mean of t2m
cosy = np.cos(np.deg2rad(t2m_xr.coords['lat'].values)).clip(0., 1.)
# pay attention to the difference of np.mean and np.nanmean and the output type
t_ave_zonal = np.mean(t2m_xr,axis=2)
# Then take the weighted average of those using the weights we calculated earlier
t2m_ave = np.average(t_ave_zonal, axis=1, weights=cosy)
print(t2m_ave)
# calculate the SST data in different seasons
sst_nino = sst_time.sel(lat = slice(-60,60),lon = slice(120,280)) # it can be adjusted
lat_sst = sst_nino.coords['lat']
lon_sst = sst_nino.coords['lon']
ssta = sst_nino.groupby('time.month') - sst_nino.groupby('time.month').mean('time')
# 分别对提前0-4个季节的SST做季节内平均
sst_DJF = ssta.sel(time=ssta['time.season'] == 'DJF')
sst_SON = ssta.sel(time=ssta['time.season'] == 'SON')
sst_JJA = ssta.sel(time=ssta['time.season'] == 'JJA')
sst_MAM = ssta.sel(time=ssta['time.season'] == 'MAM')
sst_y_DJF = np.zeros((71,len(ssta.coords['lat']),len(ssta.coords['lon'])),np.float)
sst_y_SON = np.zeros((71,len(ssta.coords['lat']),len(ssta.coords['lon'])),np.float)
sst_y_JJA = np.zeros((71,len(ssta.coords['lat']),len(ssta.coords['lon'])),np.float)
sst_y_MAM = np.zeros((71,len(ssta.coords['lat']),len(ssta.coords['lon'])),np.float)
print(sst_y_DJF.shape)
# calculate the mean of seasons (i think it can be improved with more concise code)
for ii in range(sst_y_DJF.shape[0]):
DJF_temp = sst_DJF[3*ii+2:3*ii+5,:,:]
sst_y_DJF[ii,:,:] = np.mean(DJF_temp,axis = 0)
SON_temp = sst_SON[3*ii:3*ii+3,:,:]
sst_y_SON[ii,:,:] = np.mean(SON_temp,axis = 0)
JJA_temp = sst_JJA[3*ii:3*ii+3,:,:]
sst_y_JJA[ii,:,:] = np.mean(JJA_temp,axis = 0)
MAM_temp = sst_MAM[3*ii:3*ii+3,:,:]
sst_y_MAM[ii,:,:] = np.mean(MAM_temp,axis = 0)
time = np.arange(1948,2019,1)
# print(time)
sst_xr_DJF = xr.DataArray(sst_y_DJF, dims=('time', 'lat', 'lon'), coords={'time': time,'lat':lat_sst,'lon':lon_sst})
sst_xr_SON = xr.DataArray(sst_y_SON, dims=('time', 'lat', 'lon'), coords={'time': time,'lat':lat_sst,'lon':lon_sst})
sst_xr_JJA = xr.DataArray(sst_y_JJA, dims=('time', 'lat', 'lon'), coords={'time': time,'lat':lat_sst,'lon':lon_sst})
sst_xr_MAM = xr.DataArray(sst_y_MAM, dims=('time', 'lat', 'lon'), coords={'time': time,'lat':lat_sst,'lon':lon_sst})
# print(sst_xr_MAM.mean(axis = 0).max())
# print(sst_xr_MAM.mean(axis = 0).min())
# calculate the correlations
def corr_season(var, data, ci = 0.95):
corr = np.ma.masked_all(data.shape[1:])
for mm in range(corr.shape[0]):
for nn in range(corr.shape[1]):
check_nan = data[:,mm,nn]
if np.any(np.isnan(check_nan)):
continue
c,p = stats.pearsonr(var,data[:,mm,nn])
corr.data[mm,nn] = c # convert trend to per decade
sig = (p < (1-ci))
corr.mask[mm,nn] = ~sig
return corr
corr_DJF = corr_season(t2m_ave, sst_xr_DJF)
corr_SON = corr_season(t2m_ave, sst_xr_SON)
corr_JJA = corr_season(t2m_ave, sst_xr_JJA)
corr_MAM = corr_season(t2m_ave, sst_xr_MAM)
# plot, mind the projection which matters
sns.set_style('white', {'font.family': 'Arial'})
lon_formatter = LongitudeFormatter(zero_direction_label=True)
lat_formatter = LatitudeFormatter()
lonlbl = [u'160°E',u'10°W',u'40°W',u'70°W',u'100°W']
latlbl = [u'10°S',u'5°S',u'0°',u'5°N',u'10°N']
fig = plt.figure(figsize=(12,9),dpi=300)
ax0 = fig.add_subplot(221, projection=ccrs.PlateCarree(central_longitude = 180))
plt.subplots_adjust(wspace =0.5, hspace =0.2)# adjust the space of subplots
cs1 = ax0.contour(lon_sst, lat_sst, sst_xr_DJF.mean(axis = 0), np.linspace(-0.02,0.03,6), colors='k',transform=ccrs.PlateCarree())
cs2 = ax0.contourf(lon_sst, lat_sst, corr_DJF.data, np.linspace(-0.5,0.5,11),
cmap=plt.cm.RdBu_r, extend='both',transform=ccrs.PlateCarree())
ax0.contourf(lon_sst, lat_sst, corr_DJF.mask.astype('int'), [-0.5,0.5], hatches=['.','none'],
colors='none', zorder=10,transform=ccrs.PlateCarree())
ax0.clabel(cs1, inline=1, fontsize=8)
ax0.set_extent([120, 280, -60, 60],crs=ccrs.PlateCarree())
ax0.set_xticks(range(120, 281, 30), crs=ccrs.PlateCarree())
ax0.set_yticks(range(-60, 61, 20), crs=ccrs.PlateCarree())
ax0.xaxis.set_major_formatter(lon_formatter)
ax0.yaxis.set_major_formatter(lat_formatter)
ax0.add_feature(cfeat.COASTLINE, edgecolor='#333333')
ax0.set_title(r'The correlation between t2m and SST_DJF ( head = 0 season)')
ax0 = fig.add_subplot(222,projection=ccrs.PlateCarree(central_longitude = 180))
cs1 = ax0.contour(lon_sst, lat_sst, sst_xr_SON.mean(axis = 0), np.linspace(-0.03,0.01,5), colors='k',transform=ccrs.PlateCarree())
cs2 = ax0.contourf(lon_sst, lat_sst, corr_SON.data, np.linspace(-0.5,0.5,11),
cmap=plt.cm.RdBu_r, extend='both',transform=ccrs.PlateCarree())
ax0.contourf(lon_sst, lat_sst, corr_SON.mask.astype('int'), [-0.5,0.5], hatches=['.','none'],
colors='none', zorder=10,transform=ccrs.PlateCarree())
ax0.clabel(cs1, inline=1, fontsize=8)
ax0.set_extent([120, 280, -60, 60],crs=ccrs.PlateCarree())
ax0.set_xticks(range(120, 281, 30), crs=ccrs.PlateCarree())
ax0.set_yticks(range(-60, 61, 20), crs=ccrs.PlateCarree())
ax0.xaxis.set_major_formatter(lon_formatter)
ax0.yaxis.set_major_formatter(lat_formatter)
ax0.add_feature(cfeat.COASTLINE, edgecolor='#333333')
ax0.set_title(r'The correlation between t2m and SST_SON ( head = 1 season)')
ax0 = fig.add_subplot(223,projection=ccrs.PlateCarree(central_longitude = 180))
cs1 = ax0.contour(lon_sst, lat_sst, sst_xr_JJA.mean(axis = 0), np.linspace(-0.04,0.01,5),
colors='k',transform=ccrs.PlateCarree())
cs2 = ax0.contourf(lon_sst, lat_sst, corr_JJA.data, np.linspace(-0.5,0.5,11),
cmap=plt.cm.RdBu_r, extend='both',transform=ccrs.PlateCarree())
ax0.contourf(lon_sst, lat_sst, corr_JJA.mask.astype('int'), [-0.5,0.5], hatches=['.','none'],
colors='none', zorder=10,transform=ccrs.PlateCarree())
ax0.clabel(cs1, inline=1, fontsize=8)
ax0.set_extent([120, 280, -60, 60],crs=ccrs.PlateCarree())
ax0.set_xticks(range(120, 281, 30), crs=ccrs.PlateCarree())
ax0.set_yticks(range(-60, 61, 20), crs=ccrs.PlateCarree())
ax0.xaxis.set_major_formatter(lon_formatter)
ax0.yaxis.set_major_formatter(lat_formatter)
ax0.add_feature(cfeat.COASTLINE, edgecolor='#333333')
ax0.set_title(r'The correlation between t2m and SST_JJA ( head = 2 seasons)')
ax0 = fig.add_subplot(224,projection=ccrs.PlateCarree(central_longitude = 180))
cs1 = ax0.contour(lon_sst, lat_sst, sst_xr_MAM.mean(axis = 0), np.linspace(-0.04,0.001,5),
colors='k',transform=ccrs.PlateCarree())
cs2 = ax0.contourf(lon_sst, lat_sst, corr_MAM.data, np.linspace(-0.5,0.5,11),
cmap=plt.cm.RdBu_r, extend='both',transform=ccrs.PlateCarree())
ax0.contourf(lon_sst, lat_sst, corr_MAM.mask.astype('int'), [-0.5,0.5], hatches=['.','none'],
colors='none', zorder=10,transform=ccrs.PlateCarree())
ax0.clabel(cs1, inline=1, fontsize=8)
ax0.set_extent([120, 280, -60, 60],crs=ccrs.PlateCarree())
ax0.set_xticks(range(120, 281, 30), crs=ccrs.PlateCarree())
ax0.set_yticks(range(-60, 61, 20), crs=ccrs.PlateCarree())
ax0.xaxis.set_major_formatter(lon_formatter)
ax0.yaxis.set_major_formatter(lat_formatter)
ax0.add_feature(cfeat.COASTLINE, edgecolor='#333333')
ax0.set_title(r'The correlation between t2m and SST_MAM ( head = 3 seasons)')
# color bar with new axes
cxa = fig.add_axes([0.11, 0.08, 0.8, 0.01])
kwargs = {'cax': cxa,'orientation': 'horizontal', 'label': 'correlation coefficients'}
plt.colorbar(cs2,**kwargs)
plt.savefig("./pacific.pdf")