How to plot sample point values overlaid onto a map

[edrewitz]

Hi all!

I am wondering how I can plot the numerical value of sample points on a map. Let's say my map is a contourf map for temperature and there is a spot that is 50F. How can I plot a small number 50 on the point that is 50F? I haven't found much online on how to do this so figured I'd reach out in here if anyone has any ideas. Thanks!

Regards,

Eric

[dopplershift]

• There's the clabel() matplotlib plotting method if you want to label the contours
• You could use plt.text() or ax.text() with an appropriate transform argument to put the text at that location
• If there are a lot of points, you might want to use MetPy's StationPlot class, which you could do something like:

sp = StationPlot(ax, lon, lat, transform=ccrs.PlateCarree())
sp.plot_parameter('C', my_data_vals)

[edrewitz]

Thank you for the quick response! I tried the StationPlot method and it worked. The only thing to note is in order to do this with a data array from either a GRIB or netCDF file is to first convert the data array to a dataframe. I'll share my section of code here in case if anyone else comes here with the same question.

        mask = mask
        local_time, utc_time = standard.plot_creation_time()

        rtma_data, rtma_time = da.UCAR_THREDDS_SERVER_OPENDAP_Downloads.CONUS.get_rtma_data_24_hour_difference(utc_time, 'Temperature_Analysis_height_above_ground')


        rtma_time_24 = rtma_time - timedelta(hours=24)

        from_zone = tz.tzutc()
        to_zone = tz.tzlocal()
        rtma_time = rtma_time.replace(tzinfo=from_zone)
        rtma_time = rtma_time.astimezone(to_zone)
        rtma_time_24 = rtma_time_24.replace(tzinfo=from_zone)
        rtma_time_24 = rtma_time_24.astimezone(to_zone)
        rtma_time_utc = rtma_time.astimezone(from_zone)
        rtma_time_24_utc = rtma_time_24.astimezone(from_zone)

        rtma_data = calc.unit_conversion.Temperature_Or_Dewpoint_Change_to_Fahrenheit(rtma_data)

        rtma_df = rtma_data.to_dataframe()
        
        mapcrs = ccrs.LambertConformal(central_longitude=central_longitude, central_latitude=central_latitude, standard_parallels=(first_standard_parallel,second_standard_parallel))
        datacrs = ccrs.PlateCarree()

        plot_proj = rtma_data.metpy.cartopy_crs

        fig = plt.figure(figsize=(fig_x_length, fig_y_length))
        fig.set_facecolor('aliceblue')

        ax = fig.add_subplot(1, 1, 1, projection=plot_proj)
        ax.set_extent((western_bound, eastern_bound, southern_bound, northern_bound), crs=ccrs.PlateCarree())
        ax.add_feature(cfeature.COASTLINE.with_scale('50m'), linewidth=0.75)
        ax.add_feature(cfeature.LAND, color='beige', zorder=1)
        ax.add_feature(cfeature.OCEAN, color='lightcyan', zorder=4)
        ax.add_feature(cfeature.LAKES, color='lightcyan', zorder=4)
        if show_rivers == True:
            ax.add_feature(cfeature.RIVERS, color='lightcyan', zorder=4)
        else:
            pass
        ax.add_feature(cfeature.STATES, linewidth=2, edgecolor='lime', zorder=6)
        ax.add_feature(USCOUNTIES, linewidth=county_border_linewidth, zorder=5)

        cs = ax.contourf(rtma_data.metpy.x, rtma_data.metpy.y, rtma_data, 
                         transform=rtma_data.metpy.cartopy_crs, levels=np.arange(-25, 26, 1), cmap='seismic', alpha=1, extend='both', zorder=2)

        stn = mpplots.StationPlot(ax, rtma_df['longitude'][::mask], rtma_df['latitude'][::mask],
                                         transform=ccrs.PlateCarree(), fontsize=8, zorder=3, clip_on=True)

        stn.plot_parameter('C', rtma_df['Temperature_Analysis_height_above_ground'][::mask], color='black', path_effects=[withStroke(linewidth=1, foreground='black')], zorder=3)


        cbar = fig.colorbar(cs, shrink=color_table_shrink, pad=colorbar_pad)
        cbar.set_label(label="Temperature Change (\N{DEGREE SIGN}F)", size=colorbar_label_font_size, fontweight='bold')

[edrewitz]

Another question I have regarding this, is when I repeat this process when downloading the data from NCEP/NOMADS instead of UCAR/THREDDS, this same process doesn't work. If I convert the ds to df, it is very sloppy when using the same dataset from NCEP/NOMADS and when trying to use the sp.plot_parameter('C', my_data_vals) with a data array (i.e. rtma_data[0, :, :]) I get the error saying I need to use .any() or .all() but when I do that, it still doesn't work. Any suggestions for this when working with a dataset downloaded from NOMADS rather than THREDDS? Thanks!

Here is what I am trying to do with the RTMA data from NOMADS:

stn = mpplots.StationPlot(ax, rtma_data['lon'][::decimate], rtma_data['lat'][::decimate],
                                 transform=ccrs.PlateCarree(), fontsize=12, zorder=7, clip_on=True)

stn.plot_parameter('C', rtma_data[0, ::decimate, ::decimate], color='black', path_effects=[withStroke(linewidth=1, foreground='black')], zorder=7)

[dopplershift]

My first guess is that the problem is because rtma_data[0, ::decimate, ::decimate] is a 2D array. All of that infrastructure is based around data that are columnar i.e. given as 1D lon, 1D lat, and 1D values. You'd probably need to do something like:

plot_lon, plot_lat = np.meshgrid(rtma_data['lon'][::decimate], rtma_data['lat'][::decimate])
stn = mpplots.StationPlot(ax, plot_lon.flatten(), plot_lat.flatten(),
                                 transform=ccrs.PlateCarree(), fontsize=12, zorder=7, clip_on=True)

stn.plot_parameter('C', rtma_data[0, ::decimate, ::decimate].flatten(), color='black', path_effects=[withStroke(linewidth=1, foreground='black')], zorder=7)

[edrewitz]

Thanks! That worked. :D