使用 MetPy 的声明性语法在卫星图像上绘制纬度和纬度

Question

我正试图找到一种方法来可视化MJO的单独区域/阶段。我相信这样做的一种方法是绘制分隔每个相位区域的经线(大约在60E、80E、100E、120E、140E、160E、180(，但我不确定是否可以添加到我现有的图中。

我使用的是NCEI的GRID卫星B1数据。以下是我当前的代码：

import matplotlib.pyplot as plt
from metpy.plots import declarative, colortables
import cartopy.crs as ccrs
import xarray as xr
file = "GRIDSAT-B1.2003.11.23.00.v02r01.nc"
dataset = xr.open_dataset(file)
vtime = dataset.time.values.astype('datetime64[s]').astype('O')
date_long = vtime[0]
date = date_long.strftime("%d-%b-%Y-%HZ")
# Create water vapor image
img = declarative.ImagePlot()
img.data = dataset
img.field = 'irwvp'
img.colormap = 'WVCIMSS_r'
img.image_range = (180, 280)
panel = declarative.MapPanel()
panel.layers = ['coastline', 'borders']
panel.title = f'GridSat-B1 (Water Vapor Imagery): {date}'
panel.projection = (ccrs.Mollweide(central_longitude=-240))
panel.area = ([-370, -140, -30, 30])
panel.layout = (2, 1, 2)
panel.plots = [img]
# Create the IR image
img2 = declarative.ImagePlot()
img2.data = dataset
img2.field = 'irwin_cdr'
img2.colormap = 'turbo_r' #maybe use cubehelix instead?
img2.image_range = (180, 300)
panel2 = declarative.MapPanel()
panel2.layers = ['coastline', 'borders']
panel2.title = f'GridSat-B1 (Infrared Imagery): {date}'
panel2.projection = (ccrs.Mollweide(central_longitude=-240))
panel2.area = ([-370, -140, -30, 30])
panel2.layout = (2, 1, 1)
panel2.plots = [img2]
# Plot both panels in one figure
pc = declarative.PanelContainer()
pc.size = (20, 14)
pc.panels = [panel, panel2]
pc.show()

以下是运行脚本时创建的当前输出：Nov03.png

感谢您的任何帮助/建议-提前感谢！

Answer 1

MetPy的声明性接口中没有内置任何内容，但幸运的是，MapPanel对象公开了一个.ax属性，该属性为您提供了一个MatplotlibAxes对象及其所有绘图方法：

import cartopy.crs as ccrs
import matplotlib.pyplot as plt
import metpy.plots as mpplots
import numpy as np
import xarray as xr
file = "/Users/rmay/Downloads/GRIDSAT-B1.2003.11.23.00.v02r01.nc"
dataset = xr.open_dataset(file)
vtime = dataset.time.values.astype('datetime64[s]').astype('O')
date_long = vtime[0]
date = date_long.strftime("%d-%b-%Y-%HZ")
# Create water vapor image
img = mpplots.ImagePlot()
img.data = dataset
img.field = 'irwvp'
img.colormap = 'WVCIMSS_r'
img.image_range = (180, 280)
panel = mpplots.MapPanel()
panel.layers = ['coastline', 'borders']
panel.title = f'GridSat-B1 (Water Vapor Imagery): {date}'
panel.projection = ccrs.Mollweide(central_longitude=-240)
panel.area = (-370, -140, -30, 30)
panel.layout = (2, 1, 2)
panel.plots = [img]
# Create the IR image
img2 = mpplots.ImagePlot()
img2.data = dataset
img2.field = 'irwin_cdr'
img2.colormap = 'turbo_r' #maybe use cubehelix instead?
img2.image_range = (180, 300)
panel2 = mpplots.MapPanel()
panel2.layers = ['coastline', 'borders']
panel2.title = f'GridSat-B1 (Infrared Imagery): {date}'
panel2.projection = ccrs.Mollweide(central_longitude=-240)
panel2.area = (-370, -140, -30, 30)
panel2.layout = (2, 1, 1)
panel2.plots = [img2]
# Plot both panels in one figure
pc = mpplots.PanelContainer()
pc.size = (20, 14)
pc.panels = [panel, panel2]
lons = np.array([60, 80, 100, 120, 140, 160, 180]).reshape(1, -1)
lats = np.linspace(-90, 90).reshape(-1, 1)
# Match up the arrays into 2xN arrays fit to plot in call
lons, lats = np.broadcast_arrays(lons, lats)
# Needs to be *after* the panels are assigned to a PanelContainer
# Using Geodetic gives lines interpolated on the curved globe
panel.ax.plot(lons, lats, transform=ccrs.Geodetic(), color='black', linewidth=3)
panel2.ax.plot(lons, lats, transform=ccrs.Geodetic(), color='black', linewidth=3)
pc.show()

(注意：不建议直接从metpy的声明性模块导入，因为这是一个可能会更改的实现细节——只需从metpy.plots中获取内容即可(。因此，这是使用Matplotlib对plot的标准调用来绘制线条。另一种选择是使用CartoPy的Gridliner。