fig = plt.figure(figsize=(11,8))

ax = fig.add_axes([0.1,0.1,0.8,0.8],projection=ccrs.PlateCarree())

a = ds_merra2_mam.dust_trend_mam.sel(lon=slice(60,100), lat=slice(5,40)).plot(ax=ax, \

vmin=-0.01,\

vmax=0.01,extend='both', \

cmap='twilight_shifted')

ds_merra2_mam['dust_p_value_manm_'] = ds_merra2_mam['dust_p_value_mam']<0.05

#ds_merra2_mam['dust_p_value_manm_'].sel(lon=slice(60,100), lat=slice(5,40)).plot.scatter(ax=ax)

[m,n] = np.where(ds_merra2_mam['dust_p_value_mam'].sel(lon=slice(60,100), lat=slice(5,40)).values<0.05 )

z1 = np.zeros(ds_merra2_mam['dust_p_value_mam'].sel(lon=slice(60,100), lat=slice(5,40)).values.shape)

z1[m, n] = 99

x = ds_merra2_mam.sel(lon=slice(60,100), lat=slice(5,40)).lon.values

y = ds_merra2_mam.sel(lon=slice(60,100), lat=slice(5,40)).lat.values

cs = ax.contourf(x, y, z1 ,1 , hatches=['', '..'], alpha=0)

ax.coastlines(resolution='10m')

plt.title('Dust AOD(550nm) MAM trend from MERRA2 reanalysis 2002-2020')

ax_ = [ax]

for i in range(1):

gl = ax_[i].gridlines()

gl.xlabels_bottom = True

gl.ylabels_left = True

gl.xformatter = LONGITUDE_FORMATTER

gl.yformatter = LATITUDE_FORMATTER

gl.xlabel_style = {'size': 'medium'}

gl.ylabel_style = {'size': 'medium'}

#plt.savefig('merra2_mam_dust_22_34n_65_90e_2002_2014_spatial_reproduce_vinoj.png')