Python删除所有especific变量的坐标-netcdf

Question

我有几个netcdf文件，想删除具有某些特征的数据并保存到新的netcdf文件。

文件具有纬度，经度和时间坐标，并且温度和风是可变的。这个想法是，如果风变量小于6 m / s，则将删除风和温度变量的所有信息（关于此纬度，经度和时间）。如何在python中执行此操作？

目前为止我有

from netCDF4 import Dataset
from datetime import * 
import os

#-----------------------------------------------------------------------------------------------------------------------------------
path = '/home/Downloads/Arquivos_GOES_2018_OSISAF_NC/'

data_ini = datetime(2018,1,1,1,0,0) 
data_end = datetime(2018,7,1,1,0,0) 

IDval = 6 # minimo quality_level aceito

#-----------------------------------------------------------------------------------------------------------------------------------
DateTimeNow = data_ini 

while DateTimeNow<=data_end:


   namefile_net = DateTimeNow.strftime('%Y%m%d%H%M')+'00-OSISAF-L3C_GHRSST-SSTsubskin-GOES16-ssteqc_goes16_'+DateTimeNow.strftime('%Y%m%d_%H%M')+'00-v02.0-fv01.0.nc'


   if os.path.isfile(path+namefile_net) == True:


      data_net = Dataset(path+namefile_net)
      tsm = data_net.variables['sea_surface_temperature'][0,:]-273.15 
      lat = data_net.variables['lat'][:]
      lon = data_net.variables['lon'][:]
      wind = data_net.variables['wind_speed'][0,:]

      wind = np.where(qlv.mask==True,-1.0, qlv) 

      lon, lat = np.meshgrid(lon, lat) # criar uma matriz com lat e lon


      tsm_area = tsm[yid,xid] 
      lon_area = lon[yid,xid] 
      lat_area = lat[yid,xid] 
      wind_area = wind[yid,xid] 

      if np.sum(wind>=IDval) > 0:

         lon_area = np.where(qlv_area<IDval,-999.9,lon_area)
         lat_area = np.where(qlv_area<IDval,-999.9,lat_area)

         yid2, xid2 = seach_point(lon_area, lat_area, lon_point, lat_point)
         ```

Answer 1

似乎您的数据已经是numpy。这是一个主意（没有CDF，所以让我们想象第3列是风）：

>>> y = np.zeros((4,4))
>>> data = np.zeros((4,4))
>>> data[:,3] = [1,12, 4, 9]
>>> data
array([[  0.,   0.,   0.,   1.],
       [  0.,   0.,   0.,  12.],
       [  0.,   0.,   0.,   4.],
       [  0.,   0.,   0.,   9.]])
>>> data[ data[:,3] > 6]
array([[  0.,   0.,   0.,  12.],
       [  0.,   0.,   0.,   9.]])