使用Matplotlib Quiver绘制风矢量叠加等值线报错求助

Hey there! I totally get the frustration of wrestling with quiver plots and NetCDF data—those dimension mismatches can be tricky to spot even after days of debugging. Let’s walk through the most common issues that might be causing your Quiver error with myfile1.nc and myfile2.nc:

1. Dimension Mismatch Between Coordinates and Wind Data #

The #1 culprit here is usually mismatched dimensions between your X/Y coordinates and U/V wind vectors. NetCDF datasets often store variables in multi-dimensional structures (like [time, level, lat, lon]), but quiver(X,Y,U,V) requires all four arrays to have identical shapes.

• Fix: Slice your wind data to a single time step and/or vertical level (adjust based on your dataset’s structure):

  # Example: Extract the first time step and lowest level
  U = nc_file1.variables['u'][0, 0, :, :]
  V = nc_file2.variables['v'][0, 0, :, :]
  

• If your lon and lat are 1D arrays (super common in NetCDF), convert them to 2D grids with meshgrid:

  import numpy as np
  lon = nc_file1.variables['lon'][:]
  lat = nc_file1.variables['lat'][:]
  X, Y = np.meshgrid(lon, lat)
  

  Always double-check shapes with print(X.shape, U.shape)—they should be identical!

2. Incorrect Variable Names or Missing Data #

It’s easy to misspell variable names (e.g., using 'wind_u' instead of 'ua' depending on your dataset’s conventions). Also, missing values (NaN) can break quiver if not handled properly.

• Fix:
  1. List all variables in your NetCDF files to confirm names:

     print(list(nc_file1.variables.keys()))
     

  2. Mask out NaN values to avoid plotting issues:

     from numpy.ma import masked_array
     U_masked = masked_array(U, np.isnan(U))
     V_masked = masked_array(V, np.isnan(V))
     

3. Overloading Quiver with Too Many Arrows #

If your grid is high-resolution, plotting every single vector can cause memory errors or render the plot unreadable (and sometimes trigger quiver-specific errors).

• Fix: Subsample your data to plot fewer arrows. For example, plot every 5th point:

  plt.quiver(X[::5, ::5], Y[::5, ::5], U[::5, ::5], V[::5, ::5], color='white')
  

Full Working Example Code #

Here’s a complete snippet that ties these fixes together, adapted to your setup:

from netCDF4 import Dataset as NetCDFFile
import matplotlib.pyplot as plt
import numpy as np
from numpy.ma import masked_array

# Load NetCDF files
nc_temp = NetCDFFile('myfile1.nc')
nc_wind = NetCDFFile('myfile2.nc')

# Get 1D coordinates and convert to 2D grid
lon = nc_temp.variables['lon'][:]
lat = nc_temp.variables['lat'][:]
X, Y = np.meshgrid(lon, lat)

# Extract subsampled, masked wind data (adjust indices to match your dataset)
U = masked_array(nc_wind.variables['u'][0, 0, :, :], np.isnan(nc_wind.variables['u'][0, 0, :, :]))
V = masked_array(nc_wind.variables['v'][0, 0, :, :], np.isnan(nc_wind.variables['v'][0, 0, :, :]))
U_sub = U[::5, ::5]
V_sub = V[::5, ::5]
X_sub = X[::5, ::5]
Y_sub = Y[::5, ::5]

# Plot contours first (e.g., temperature)
temp = nc_temp.variables['temp'][0, 0, :, :]
plt.contourf(X, Y, temp, cmap='coolwarm')

# Add wind vectors
plt.quiver(X_sub, Y_sub, U_sub, V_sub, color='black', scale=100)

# Add labels and colorbar
plt.xlabel('Longitude')
plt.ylabel('Latitude')
plt.colorbar(label='Temperature (K)')
plt.title('Temperature Contours with Wind Vectors')
plt.show()

If you’re still getting errors, share the full traceback message (not just "Quiver error")—that will help pinpoint exactly where the issue is!

内容的提问来源于stack exchange，提问作者bayouwxman