METAR GeoParquet Interactive Visualization

Overview¶

Within this notebook, we will create interactive visualizations of METAR data at a single hour. We will use the following libraries for our visualizations:

Geopandas

Prerequisites¶

Concepts	Importance	Notes
Cartopy Intro	Required	Projections and Features
Pandas	Required	Tabular Datasets

Time to learn: 20 minutes

Imports¶

import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
from datetime import datetime,timedelta
import geopandas as gpd
from lonboard import viz, Map, ScatterplotLayer, HeatmapLayer
import duckdb
import folium

By default, set the date and time to one hour prior to the current time. Or, specify a past date and hour.¶

# Use the current time, or set your own for a past time.
# Set current to False if you want to specify a past time.

nowTime = datetime.now()

current = True
#current = False
if (current):
    time_1 = datetime.now()
    offset = timedelta(hours = 1) # Worldwide METARS typically don't come in until ~15 minutes past the hour
    time_1 = time_1 - offset
else:
    year = 2026
    month = 1
    day = 1
    hour = 0
    time_1 = datetime(year, month, day, hour)

The timestamps of many hourly METAR reports are 5-10 minutes prior to the top of the hour. When we make our DuckDB query, we will specify a beginning and end time, one hour apart.

time_0 = time_1 - timedelta(hours=1)
YYYY_0 = time_0.strftime("%Y")
YYYY_1 = time_1.strftime("%Y")
date_string = time_1.strftime("%Y-%m-%d %H00 UTC")
print(time_0, time_1)
# Handle edge case when the two hours straddle the end/beginning of a yearw
if (YYYY_0 == YYYY_1):
    URLs = [f'https://data.source.coop/dynamical/asos-parquet/year={YYYY_1}/data.parquet']
else:
    URLs = [f'https://data.source.coop/dynamical/asos-parquet/year={YYYY_0}/data.parquet',
            f'https://data.source.coop/dynamical/asos-parquet/year={YYYY_1}/data.parquet']

2026-10-23 02:24:23.385799 2026-10-23 03:24:23.385799

date_string

'2026-10-23 0300 UTC'

URLs

['https://data.source.coop/dynamical/asos-parquet/year=2026/data.parquet']

Open the GeoParquet file for the specific year.¶

df = duckdb.execute("""
    SELECT *
    FROM read_parquet($1, hive_partitioning=true)
    WHERE 
---      country = 'FR' AND
---    station = 'ENGM' AND
    valid BETWEEN $2 AND $3
    ORDER BY country
""", [URLs, time_0, time_1]).fetchdf()

Loading...

/tmp/ipykernel_4489/3583836804.py:9: Pandas4Warning: The inplace keyword in DataFrame.drop is deprecated and will be removed in a future version (PDEP-8).
See https://pandas.pydata.org/docs/dev/whatsnew/v3.1.0.html#deprecation-inplace for more details.
  """, [URLs, time_0, time_1]).fetchdf()

df

Loading...

gdf = gpd.GeoDataFrame(df,geometry=gpd.points_from_xy(df.longitude,df.latitude))

gdf.explore()

Loading...

gdf.set_crs(epsg=4326, inplace=True, allow_override=True)

Loading...

gdf.explore()

Loading...

Now, let’s choose a specific variable ... in this case, tmpc (2-meter temperature in °C)

gdf.explore(column='tmpc')

Loading...

Plot hourly precipitation

gdf.explore(column='p01m')

Loading...

gdf.loc[gdf['p01m'] < 0.01, 'p01m'] = np.nan

gdf = gdf.dropna(subset=['p01m']).reset_index(drop=True)

m = gdf.explore(column='p01m')

Add a title to the graphic¶

title_string = f"Hourly Precip (mm), {date_string}"

title_html = f"""
<h3 align="center" style="font-size:20px">
    <b>{title_string}</b>
</h3>
"""

m.get_root().html.add_child(folium.Element(title_html))
m

Loading...

References¶

What’s next?¶

We will use Lonboard as an additional interactive visualization tool.