Overlaying Contours on Color-Filled Shading in UXarray/Holoviews/Matplotlib

[falkojudt]

Is there a way to overlay contours on a color-filled shading, similar to this example where the background represents wind speed and the black contours indicate sea level pressure?

[philipc2]

Hi @falkojudt

There was some preliminary work shown in #366 for contours. We are planning to add .contour() and .contourf() methods that wrap hvPlot

https://hvplot.holoviz.org/reference/xarray/contour.html#

https://hvplot.holoviz.org/reference/xarray/contourf.html

We could consider adding pure matplotlib functionality too if that would of be interest.

[philipc2]

Here's a basic example, without the in-line contour labels. Currently it doesn't seem to be possible to add the labels, as discussed in holoviz/holoviews#4494

import holoviews as hv
from holoviews import opts

hv.extension("bokeh")

# radial subset
sub = uxds['relhum_200hPa'][0].subset.bounding_circle((0.0, 0.0), r=30)

# generate color plot
color_plot = sub.plot()

# generate contour plot
contour_plot = hv.operation.contours(color_plot, levels=3).opts(line_width=2)

# overlay them together
(color_plot * contour_plot).opts(frame_width=500, frame_height=500, title="Relative Humidity (200 hPa) with Contours")

[philipc2]

Would you be willing to share the data used in this example? I'd love to try some things locally and see if I can replicate the plot you shared.

[falkojudt]

Hi @philipc2. Thanks for taking a look at this.

What I'm trying to plot is a contour field over a color-shaded field, but for a different variable.

Here’s my notebook with links to the data on Glade, which you should be able to access.

import warnings

import cartopy.crs as ccrs
import geoviews.feature as gf
import holoviews as hv

import pandas as pd
import numpy as np
import xarray as xr
import uxarray as ux

from metpy.calc import wind_speed

# import geocat.datafiles as geodf

plot_opts = {"width": 700, "height": 350}
hv.extension("bokeh")
warnings.filterwarnings("ignore")

#--- Load tropical cyclone track data
track_file = f"/glade/campaign/mmm/wmr/fjudt/projects/dyamond_1/3.75km/te/tc_tracks.txt"
df = pd.read_csv(track_file, header=1, names=['track_id', 'year', 'month', 'day', 'hour', 'i', 
                                                 'lon', 'lat', 'slp', 'wind', 'ter'])
df['datetime'] = pd.to_datetime(df[['year', 'month', 'day', 'hour']])

# Find the index of the row where wind speed has its maximum value
index_of_max_wind = df['wind'].idxmax()

# Get the latitude and longitude values from that row
lat_at_max_wind = df.loc[index_of_max_wind, 'lat']
lon_at_max_wind = df.loc[index_of_max_wind, 'lon']
datetime_at_max_wind = df.loc[index_of_max_wind, 'datetime']

# Now let's plot a map of the cyclone's wind speed field

# Load MPAS data
datafiles = [
    "/glade/campaign/mmm/wmr/fjudt/projects/dyamond_1/3.75km/x1.static.nc",
    "/glade/campaign/mmm/wmr/fjudt/projects/dyamond_1/3.75km/diag.2016-08-25_06.00.00.nc",
]

uxds = ux.open_dataset(datafiles[0], datafiles[1])

#--- add wind speed variable as a uxdatarray 
uxds["windspeed"] = ux.UxDataArray(wind_speed(uxds["u10"], uxds["v10"]))

#--- subset the data (1.5 degrees around the storm center) and plot
center_coord = [lon_at_max_wind, lat_at_max_wind]
r = 1.5

features = gf.coastline(
    projection=ccrs.PlateCarree(), line_width=1, scale="50m"
) * gf.states(projection=ccrs.PlateCarree(), line_width=1, scale="50m")

sub_wspd = uxds["windspeed"][0].subset.bounding_circle(center_coord, r)
sub_mslp = 0.01*uxds["mslp"][0].subset.bounding_circle(center_coord, r)

color_plot = sub_wspd.plot.rasterize(
    method="polygon",
    pixel_ratio=3.0,
    fig_size=200,
    xaxis=None,
    yaxis=None,
    backend="matplotlib",
) * features

color_plot

#--- How to overlay contours of sub_mslp?

You need these tow files:

/glade/campaign/mmm/wmr/fjudt/projects/dyamond_1/3.75km/x1.static.nc
/glade/campaign/mmm/wmr/fjudt/projects/dyamond_1/3.75km/diag.2016-08-25_06.00.00.nc

and these two values for subsetting:
at_at_max_wind: 30.036345
lon_at_max_wind: 281.248091