3 IBTrACS

IBTrACS (Knapp et al. 2010) is the global archive of historical storm tracks. We load the track points within the ZMA’s bounds from the team database and plot them as a check.

Code

from src.datasources import ibtracs, zma
import pandas as pd
import geopandas as gpd
from shapely.geometry import LineString
from matplotlib import pyplot as plt

Code

gdf_zma = zma.load_zma()

Code

total_bounds = gdf_zma.total_bounds

Code

df_all = ibtracs.load_ibtracs_in_bounds(*total_bounds)

Code

df_all

	point_id	sid	valid_time	latitude	longitude	wind_speed	gust_speed	pressure	max_wind_radius	last_closed_isobar_radius	last_closed_isobar_pressure	basin	nature	provider	quadrant_radius_34	quadrant_radius_50	quadrant_radius_64	created_at
0	fdc4f638-eb3d-4044-912c-a80e2921f92f	2023152N26274	2023-06-03 18:00:00.000039	23.900000	-85.000000	25.0	NaN	1001.0	40.0	140.0	1005.0	NA	DS	nhc_working_bt	[nan, nan, nan, nan]	[nan, nan, nan, nan]	[nan, nan, nan, nan]	2025-05-13 21:22:52.749544
1	8cd22c3c-b67f-4c58-947e-1375b54f0d46	2023152N26274	2023-06-03 21:00:00.000039	23.677523	-84.609840	25.0	NaN	1001.0	40.0	140.0	1005.0	NA	DS	None	[nan, nan, nan, nan]	[nan, nan, nan, nan]	[nan, nan, nan, nan]	2025-05-13 21:22:52.749544
2	030c5694-3836-436d-a5ef-e5b83fa46a81	2023152N26274	2023-06-04 00:00:00.000039	23.500000	-84.099998	25.0	NaN	1002.0	40.0	140.0	1005.0	NA	DS	nhc_working_bt	[nan, nan, nan, nan]	[nan, nan, nan, nan]	[nan, nan, nan, nan]	2025-05-13 21:22:52.749544
3	ebb42dc3-fd00-4381-acd4-8f4f874ad1a6	2023152N26274	2023-06-04 03:00:00.000039	23.382824	-83.555038	25.0	NaN	1002.0	35.0	140.0	1005.0	NA	DS	None	[nan, nan, nan, nan]	[nan, nan, nan, nan]	[nan, nan, nan, nan]	2025-05-13 21:22:52.749544
4	037fb1c8-0fb4-4455-ac15-3687be8880d1	2023152N26274	2023-06-04 06:00:00.000039	23.299999	-83.000000	25.0	NaN	1002.0	30.0	140.0	1005.0	NA	DS	nhc_working_bt	[nan, nan, nan, nan]	[nan, nan, nan, nan]	[nan, nan, nan, nan]	2025-05-13 21:22:52.749544
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
3058	4ff60e09-8b82-497c-bdf7-57fdaff92070	2022266N12294	2022-09-27 00:00:00.000039	20.799999	-83.300003	85.0	105.0	965.0	20.0	130.0	1006.0	NA	TS	hurdat_atl	[100.0, 90.0, 60.0, 90.0]	[50.0, 50.0, 20.0, 30.0]	[30.0, 25.0, nan, 20.0]	2025-05-13 21:22:52.749544
3059	89a90801-a7f1-492e-8c69-75ea047d8a2a	2022266N12294	2022-09-27 06:00:00.000039	21.799999	-83.599998	100.0	120.0	956.0	15.0	200.0	1008.0	NA	TS	hurdat_atl	[100.0, 100.0, 70.0, 90.0]	[50.0, 50.0, 30.0, 40.0]	[30.0, 25.0, 20.0, 20.0]	2025-05-13 21:22:52.749544
3060	4bc322f1-1600-41a0-b8a9-18b2005d9843	2022266N12294	2022-09-27 08:30:00.000066	22.200001	-83.699997	110.0	NaN	947.0	15.0	200.0	1008.0	NA	TS	hurdat_atl	[100.0, 100.0, 70.0, 90.0]	[50.0, 50.0, 30.0, 40.0]	[30.0, 25.0, 20.0, 20.0]	2025-05-13 21:22:52.749544
3061	44c6b365-52e7-4945-9669-4843f6b68c81	2022266N12294	2022-09-27 12:00:00.000039	22.600000	-83.599998	100.0	135.0	963.0	15.0	270.0	1009.0	NA	TS	hurdat_atl	[120.0, 120.0, 100.0, 120.0]	[60.0, 60.0, 50.0, 60.0]	[30.0, 25.0, 20.0, 20.0]	2025-05-13 21:22:52.749544
3062	f856ebc9-966b-4337-8efd-a60a998bcec3	2022266N12294	2022-09-27 18:00:00.000039	23.500000	-83.300003	105.0	130.0	951.0	15.0	270.0	1009.0	NA	TS	hurdat_atl	[120.0, 120.0, 100.0, 120.0]	[60.0, 60.0, 50.0, 60.0]	[35.0, 35.0, 20.0, 20.0]	2025-05-13 21:22:52.749544

3063 rows × 18 columns

Code

def df_to_track_lines(df: pd.DataFrame) -> gpd.GeoDataFrame:
    """
    Convert a DataFrame of points to a GeoDataFrame of LineStrings,
    grouped by 'sid'.

    Parameters:
        df: DataFrame with 'sid', 'latitude', 'longitude' columns

    Returns:
        GeoDataFrame with one LineString per 'sid'
    """
    # Ensure sorted order if needed (e.g., by time)
    
    df = df.sort_values(["sid"])  # optionally add 'time' or similar

    # Group and build LineStrings
    lines = (
        df.groupby("sid")
        .filter(lambda group: len(group) > 1)  # Keep only groups with more than one point
        .groupby("sid")
        .apply(
            lambda group: LineString(
                zip(group["longitude"], group["latitude"])
            )
        )
        .reset_index(name="geometry")
    )

    return gpd.GeoDataFrame(lines, geometry="geometry", crs="EPSG:4326")

Code

gdf_lines = df_to_track_lines(df_all)

Code

gdf_lines

	sid	geometry
0	1851228N13313	LINESTRING (-79 21.2, -72.6 18.9, -74.3 19.4, ...
1	1852232N21293	LINESTRING (-74 21.9, -74.9 22.1, -76.1 22.2, ...
2	1852264N13309	LINESTRING (-73 21.4, -73.6 21.9, -74 22.5, -7...
3	1852278N14293	LINESTRING (-83.1 18.7, -86.4 21.2, -85.5 20.4...
4	1855236N12304	LINESTRING (-72.3 19.5, -73.6 19.7)
...	...	...
410	2022154N21273	LINESTRING (-86.3 22.8, -85.1 23.9)
411	2022266N12294	LINESTRING (-82.4 18.7, -83 19.7, -83.3 20.8, ...
412	2023152N26274	LINESTRING (-85 23.9, -84.60984 23.67752, -84....
413	2023239N21274	LINESTRING (-86.1 20.8, -86.14497 21.11002, -8...
414	2023321N15278	LINESTRING (-77.84354 18.57282, -77.2 19.2)

415 rows × 2 columns

Code

fig, ax = plt.subplots()
max_lines = 100
plotted_lines = 0
for sid, row in gdf_lines.iterrows():
    plotted_lines += 1
    if plotted_lines > max_lines:
        break
    x, y = row.geometry.xy
    ax.plot(x, y, label=str(row["sid"]))  # matplotlib auto-assigns color

gdf_zma.boundary.plot(ax=ax, color="k")

All the tracks fall inside the ZMA, so the load looks right.