Probabilities: significant wave height¶

This notebook will provide you guidance how to explore and plot ECMWF open dataset to produce the map from the ECMWF open charts web product.
The original product can be found on this link: https://charts.ecmwf.int/products/medium-swh-probability

The full list of available Open data products can be found here, and more information can be found in the User documentation.

Access to ECMWF Open data is governed by the Creative Commons CC-BY-4.0 licence and associated Terms of Use.

_{In applying this licence, ECMWF does not waive the privileges and immunities granted to it by virtue of its status as an intergovernmental organisation nor does it submit to any jurisdiction}

To find out how to obtain the access to the full forecast dataset at higher resolution please visit our Access page.

Retrieve Data¶

This product takes in input one of these 4 parameters:

These are already calculated probabilities, so no need to download all ensemble members.

In this example, we will use:

ecmwf.opendata Client to download the data
Metview library to read, process and plot the data

First we need to install them in the current Jupyter kernel:

Note: If you are running the notebook on MyBinder or already have the libraries installed, go directly to importing the libraries.

Note: If you don't have these libraries installed, click on three dots below, uncomment the code and run the next cell.

In [1]:

#!pip install ecmwf-opendata metview metview-python

In [2]:

import metview as mv
from ecmwf.opendata import Client

In [3]:

client = Client("ecmwf", beta=True)

In [4]:

parameters = ['swhg2', 'swhg4', 'swhg6', 'swhg8']
filename = 'medium-swh-probability.grib'
filename

Out[4]:

'medium-swh-probability.grib'

Setting date to 0 will download today's data. Removing date and time altogether from the request will download the latest data.
Try commenting out date and time to download latest forecast!

In [5]:

client.retrieve(
    date=-1,
    time=0,
    step=48,
    stream="waef",
    type='ep',
    levtype="sfc",
    param=parameters,
    target=filename
)

20241223000000-240h-waef-ep.grib2:   0%|          | 0.00/759k [00:00<?, ?B/s]

Out[5]:

<ecmwf.opendata.client.Result at 0x10da64c50>

Reading and processing the data¶

Now we can use ecmwf.data to read the file.

In [6]:

data = mv.read(filename)

The describe() function will give us the overview of the dataset.

In [7]:

data.describe()

Out[7]:

parameter	typeOfLevel	date	step	number	paramId	class	stream	type	experimentVersionNumber
swhg2	meanSea	20241223	48	None	131074	od	waef	ep	0001
swhg4	meanSea	20241223	48	None	131075	od	waef	ep	0001
swhg6	meanSea	20241223	48	None	131076	od	waef	ep	0001
swhg8	meanSea	20241223	48	None	131077	od	waef	ep	0001

And an overview of one parameter, where we can see more information, such as units or type of level.

In [8]:

data.describe('swhg4')

Out[8]:

shortName	swhg4
name	Significant wave height of at least 4 m
paramId	131075
units	%
typeOfLevel	meanSea
level	0
date	20241223
time	0
step	48
number	None
class	od
stream	waef
type	ep
experimentVersionNumber	0001

We can use ls() function to list all the fields in the file we downloaded.

In [9]:

data.ls()

Out[9]:

	centre	shortName	typeOfLevel	level	dataDate	dataTime	stepRange	dataType	number	gridType
Message
0	ecmf	swhg4	meanSea	0	20241223	0	48	ep	None	regular_ll
1	ecmf	swhg6	meanSea	0	20241223	0	48	ep	None	regular_ll
2	ecmf	swhg2	meanSea	0	20241223	0	48	ep	None	regular_ll
3	ecmf	swhg8	meanSea	0	20241223	0	48	ep	None	regular_ll

The grib file contains all the parameters, and we will use the select() function to filter what we need.
Feel free to use describe() to inspect the other filtered fields.

In [10]:

swhg4 = data.select(shortName = 'swhg4')

Plotting the data¶

And finally, we can plot the data on the map.

In [11]:

# define coastlines
coast = mv.mcoast(
    map_coastline_colour="charcoal",
    map_coastline_resolution="medium",
    map_coastline_land_shade="on",
    map_coastline_land_shade_colour="cream",
    map_coastline_sea_shade="off",
    map_boundaries="on",
    map_boundaries_colour= "charcoal",
    map_boundaries_thickness = 1,
    map_disputed_boundaries = "off",
    map_grid_colour="tan",
    map_label_height=0.35,
)

# define view
view = mv.geoview(
    area_mode="name",
    area_name="europe",
    coastlines=coast
)

#define styles
swhg4_shade = mv.mcont(legend= "on",
                contour_automatics_settings = "style_name",
                contour_style_name = "sh_blup_f0t100lst")

title = mv.mtext(
    text_lines=["Probabilities: significant wave height > 4 m", 
           "START TIME: <grib_info key='base-date' format='%a %d %B %Y %H'/> ",
           "VALID TIME: <grib_info key='valid-date' format='%a %d %B %Y %H'/>, STEP: <grib_info key='step' />"],
    text_font_size=0.4,
    text_colour         = 'charcoal')

ecmwf_text = mv.mtext(    
    text_lines          = ["© European Centre for Medium-Range Weather Forecasts (ECMWF)",
                           "Source: www.ecmwf.int Licence: CC-BY-4.0 and ECMWF Terms of Use",
                            "https://apps.ecmwf.int/datasets/licences/general/"],
    text_justification  = 'center',
    text_font_size      = 0.3,
    text_mode           = "positional",
    text_box_x_position = 6.,
    text_box_y_position = -0.2,
    text_box_x_length   = 8,
    text_box_y_length   = 2,
    text_colour         = 'charcoal')

# generate plot
mv.setoutput('jupyter', plot_widget=False)
mv.plot(view, swhg4, swhg4_shade, title, ecmwf_text)

Out[11]:

To generate the png file you can run the following cell.

In [12]:

png = mv.png_output(
      output_name = "medium-swh-probability",   # specify relative or full path
      output_title = "medium-swh-probability",    # title used by a viewer
      output_width = 1000,                 # set width in pixels
)
mv.setoutput(png)
mv.plot(view, swhg4, swhg4_shade, title, ecmwf_text)

Note that plot produced using open data dataset will slightly differ from one from Open Charts. This is due to different resolution of the data.
Open data is on 0.25x0.25 resolution, while high resolution data is 0.1x0.1 grid.