New York ACTUALLY HAS 12 SEASONS

This web page is inspired by this meme.

While the meme is simple, this page... not so much. It might even be a bit over-engineered.

How this Works

The goal of this page is to capture, not the astrological or meteorological seasons of New York, but the experiential seasons of a New Yorker. A creature who lives their life largely outside the traditional comforts of climate-controlled cars.

For instance, if it's unseasonably warm in late winter, the New Yorker's brain knows it's not spring, yet they might be tricked into donning a light jacket, only to be corrected later that evening by the city's unforgiving weather.

This is a classic case of experiencing Fool's Spring.

This web page dynamically updates based on the day of the year, the time of day, and the current temperature in NYC.

It uses both live data and 53 years of historical data from OpenWeatherMap.org.

Methodology

I mapped the meme's seasons to a two-dimensional grid, using the astrological sub-season and the relative temperature for that time of year.

This is more art than science.

Winter

Spring

Summer

Fall

Next, I needed to define these axes.

Sub-Seasons

For the early, mid, and late I divided the seasons by into three roughly equal segments using their astrological definitions.

Ranges are inclusive start and exclusive end.

Historical Temperature Distributions

Next, I established a weather baseline for NYC by analyzing temperature data based on the day of the year and the hour of the day. I collected hourly temperature data for NYC from January 1, 1979, to January 1, 2023, and calculated the median and standard deviation for every hour across the entire year.

If you're interested in exploring this data, it's available in the repo here.

import pandas as pd
weather_df = pd.read_csv("nyc_weather.csv")
weather_df["dt_datetime"] = pd.to_datetime(weather_df.dt_iso.str.replace(" UTC", ""))
weather_df["day_hour"] = weather_df.dt_datetime.dt.strftime("%j-%H")
weather_df = weather_df.groupby(by="day_hour").aggregate({"feels_like": ["mean", "std"]})
hour_temps_df = df.groupby(by="day_hour").aggregate({"feels_like": ["mean", "std"]})["feels_like"]
import json
hour_temps_json = json.dumps(hour_temps_df.to_dict("index"))

From this baseline, we define a "hot" day as any day where the temperature is 1.5 standard deviations above the mean, and a "cold" day as one that falls 1.5 standard deviations below the mean.

With this baseline for what is "normal," we can now dynamically compare the current temperature to our historical data. Every time the page loads, we call OpenWeatherMap.org to get the current temperature for Times Square. This live reading is then compared against our historical distribution to determine if it’s unusually hot or cold.

June 2023 - Fire Season Update

After the events of June 7th, 2023, I've updated the page to reflect the new reality. When the air quality index is 3 or worse in the summer/spring/fall, the season is now 'Hell's Front Porch'.

Deciding the Season

Using these features, we update the arrow based on the logic described above. For instance, if it's late winter and the temperature is 1.5 standard deviations above the average for that day and hour, we classify it as "Fool's Spring." A detailed explanation of the data used to make this decision is appended below the header.

This page was created by Devon Peticolas. If you'd like to view the source code, fork the project, or suggest a change, feel free to check out the repository on GitHub.