This project aims to conduct an analysis of heatwave patterns across the Chicago over the past lustrum (2020.10.1-2025.10.1). The research will focus on identifying spatial variations and intensity patterns of heatwave events using CyberGISX. The ultimate goal is to create interactive visualizations that effectively communicate heat risk patterns to support urban planning and public health preparedness.¶
The primary data will be sourced from the National Oceanic and Atmospheric Administration (NOAA) through its Global Historical Climatology Network (GHCN). I use daily maximum temperature collected by different stations as the main variable to analyze.¶
Due to the restrictions of the data resources on the size of the dataset, I will make multiple requests for data and combine the datasets from the same observation station into one before conducting further analysis. I will use the geospatial tools including folium, matplotlib, pandas and geopandas to conduct further analysis and comparison of different observation stations, and finally visualize the patterns and intensities of heat waves.¶
For now, I just request the data once from NOAA, and I can get the data of 2 years, from 2023.9.1 to 2025.9.1. I select one station to give an example of what I would do. I use 90°F as a heatwave threshold and draw a graph of daily maximum temperature. Also, I draw the map of Chicago using Folium.¶
In [1]:
import pandas as pd
import matplotlib.pyplot as plt
import folium
In [2]:
df = pd.read_csv('weather_data.csv')
In [3]:
df_clean = df[df.STATION == "USC00111577"].reset_index()
In [4]:
plt.figure(figsize=(14, 6))
Out[4]:
In [5]:
plt.plot(df_clean.index, df_clean['TMAX'], 'b-', linewidth=1.0, alpha=0.8, label='Daily Max Temperature (°F)')
plt.axhline(y=90, color='red', linestyle='--', linewidth=2, alpha=0.8, label='Heatwave Threshold (90°F)')
plt.xlabel('Days from 2023.9.1')
plt.ylabel('Temperature (°F)')
plt.title('Maximum Temperature (TMAX) with Heatwave Threshold in Chicago')
plt.legend()
plt.grid(True, alpha=0.3)
In [55]:
def plot_chicago():
chicago_center = [41.8781, -87.6298]
m = folium.Map(location=chicago_center, zoom_start=10, tiles='OpenStreetMap')
folium.Marker(
chicago_center,
popup='Chicago Downtown',
tooltip='Click for more info',
icon=folium.Icon(color='red', icon='info-sign')
).add_to(m)
return m
In [56]:
plot_chicago()
Out[56]:
After deciding the variables and scope to focus on, now it is time to calculate the heatwave frequency. As I mentioned, I choose 90°F as a heatwave threshold. If the maximum temperature exceed 90°F for more 2 consecutive days, these days will be considered as having a heatwave. Since I can only get the data of maximuim temperature for 2 years at a time, I respectively request the data of 2020-2021, 2021-2023, and 2023-2025. Then I write a function to calculate the number of heatwave days of the each stations with effective data and record the result, presenting below.¶
In [38]:
df_01 = pd.read_csv('Tmax01.csv', na_values = ["NaN"])
df_01_clean = df_01.dropna()
df_01_clean.reset_index()
df_01_clean.NAME.unique()
df_01_s = df_01_clean.sort_values(["STATION", "DATE"]).reset_index(drop=True)
def count_hot_days(sub_df):
temps = sub_df["TMAX"]
mask = temps > 90
groups = (mask != mask.shift()).cumsum()
streak_lengths = mask.groupby(groups).sum()
valid_groups = streak_lengths[streak_lengths >= 2].index
total_days = mask.groupby(groups).sum().loc[valid_groups].sum()
return total_days
result_01 = df_01_s.groupby("STATION").apply(count_hot_days).reset_index(name="NumDays")
print(result_01)
In [37]:
df_13 = pd.read_csv('Tmax13.csv', na_values = ["NaN"])
df_13_clean = df_13.dropna()
df_13_clean.reset_index()
df_13_clean.NAME.unique()
df_13_s = df_13_clean.sort_values(["STATION", "DATE"]).reset_index(drop=True)
result_13 = df_13_s.groupby("STATION").apply(count_hot_days).reset_index(name="NumDays")
print(result_13)
In [33]:
df_35 = pd.read_csv('Tmax35.csv', na_values = ["NaN"])
df_35_clean = df_35.dropna()
df_35_clean.reset_index()
df_35_clean.NAME.unique()
df_35_s = df_35_clean.sort_values(["STATION", "DATE"]).reset_index(drop=True)
result_35 = df_35_s.groupby("STATION").apply(count_hot_days).reset_index(name="NumDays")
print(result_35)
After getting the result of 8 stations of 2020-2021, 2021-2023, and 2023-2025, I add them together to get the total heatwave days of these stations in 5 years.¶
In [40]:
combined = pd.concat([result_01, result_13, result_35])
combined = combined[~combined["STATION"].str.contains("110")]
final = combined.groupby("STATION", as_index=False)["NumDays"].sum()
print(final)
After that, I use the original dataframe to find the latitude and longitude information for these effective stations, and merge the dataset.¶
In [49]:
coords = df[["STATION", "LATITUDE", "LONGITUDE"]].drop_duplicates().reset_index()
final_coords = coords[coords["STATION"].isin(final["STATION"])].reset_index()
final_coords
Out[49]:
The heatwave index is simply calculated by dividing the number of heatwave days (heatwave frequency) by 1826, the total days from 2020.10.1 to 2025.10.1.¶
In [53]:
final_with_coords = final.merge(final_coords[["STATION", "LATITUDE", "LONGITUDE"]], on="STATION", how="left")
final_with_coords["heatwave_index"] = final_with_coords["NumDays"] / 1826
final_with_coords
Out[53]:
At last, I connect the result with folium, and present the data on folium. I record the location of the stations on the map as seperate points, and when clicking the point, we can get the heatwave frequency and heatwave index for these stations for 2020-2025.¶
In [54]:
import folium
def plot_chicago(final_with_coords):
chicago_center = [41.8781, -87.6298]
m = folium.Map(location=chicago_center, zoom_start=10, tiles='OpenStreetMap')
folium.Marker(
chicago_center,
popup='Chicago Downtown',
tooltip='Click for more info',
icon=folium.Icon(color='red', icon='info-sign')
).add_to(m)
for _, row in final_with_coords.iterrows():
folium.CircleMarker(
location=[row["LATITUDE"], row["LONGITUDE"]],
radius=8,
color='blue',
fill=True,
fill_color='orange',
fill_opacity=0.7,
popup=(
f"<b>Station:</b> {row['STATION']}<br>"
f"<b>Heatwave Index:</b> {row['heatwave_index']:.4f}<br>"
f"<b>NumDays:</b> {row['NumDays']}"
),
tooltip=f"{row['STATION']}"
).add_to(m)
return m
m = plot_chicago(final_with_coords)
m
Out[54]: