Earthquakes and the JMA

import pandas as pd
import datetime
import matplotlib.pyplot as plt

# The json url
quakes_json_url = "https://www.jma.go.jp/bosai/quake/data/list.json"

# Load it into a dataframe
quakes = pd.read_json(quakes_json_url)

# Get rid of entries where the magnitude is unknown
mask = quakes.mag != "Ｍ不明"
quakes = quakes.loc[mask]

# Get rid of distant stuff (overseas)
mask = quakes.en_ttl != "Distant Earthquake Information"
quakes = quakes.loc[mask]

# Drop anything where ttl (Title) isn't "震源・震度情報" (Epicenter information)
# Some of the entries are more like "information" and I believe these are not necessary
mask = quakes.ttl == "震源・震度情報"
quakes = quakes.loc[mask]

# Extract the latitude, longitude and depth from 'cod' column
# We won't include the +/- for lat/long because Japan is all +
pattern = "[+-](\d{0,3}\.\d)[+-](\d{0,3}\.\d)[+-](\d*)/"
quakes[["latitude", "longitude", "depth"]] = quakes.cod.str.extract(pattern)

# Convert some columns to float/int
quakes.mag = pd.to_numeric(quakes.mag)
quakes.acd = quakes.acd.astype("int")
quakes.latitude = pd.to_numeric(quakes.latitude)
quakes.longitude = pd.to_numeric(quakes.longitude)
quakes.depth = (pd.to_numeric(quakes.depth) / 1000).astype("int") # convert to km first

# Convert 'at' to datetime but first will drop the offset
quakes["at"] = quakes["at"].apply(lambda x: x.replace("+09:00", "").replace("T", " "))
quakes["at"] = pd.to_datetime(quakes["at"])

# One other issue is that there are duplicates. I believe this is resolved by keeping the "ser" 2 
# There was one row, for example, that showed the big earthquake at M7.4, and another at M7.6
# But on the website the M7.6 is what is shown
# The eid is duplicated, and the first duplicate is the "ser" 1
quakes = quakes.drop_duplicates("eid")

# Drop columns we don't need
# ctt ?, rdt (Report Date Time), ttl (Title), en_ttl (English Title), ift ?, ser (Serial), cod, int, eid (Event ID)
# 'int' is intensity information for stations, but the json file is more detailed
# 'at' is utc datetime
quakes = quakes.drop(
    ["ttl", "en_ttl", "ctt", "rdt", "ift", "cod", "int", "ser"],
    axis=1
)

# Set index to eid
quakes = quakes.set_index("eid")

# Rename columns
quakes = quakes.rename(
    columns={
        "anm": "area_name", 
        "en_anm": "area_name_en", 
        "maxi": "max_intensity", 
        "at": "date_time",
        "acd": "area_code",
    })


quakes.head()

# Number of quakes in 390 in a 24 hour time frame

# First, specify some dates/times
# 4pm, new years
start = datetime.datetime.fromisoformat("2024-01-01 16:00") 
# 24hrs later
end = datetime.datetime.fromisoformat("2024-01-02 16:00") 

# Make a date/time mask
dt_mask = (quakes.date_time > start) & (quakes.date_time < end)

# Make a location mask for Ishikawa prefecture
location_mask = quakes.area_code == 390

# Apply the mask and see how many there are
quakes[location_mask & dt_mask].shape[0]

# result
116

ishikawa_quakes = quakes[location_mask & dt_mask]

# sort in reverse, and reset the index
ishikawa_quakes = ishikawa_quakes.sort_values("date_time", ascending=True).reset_index().drop("eid", axis=1)

# plot it
ishikawa_quakes.mag.plot(
    title="Quake Magnitude, Ishikawa", 
    xlabel="Date/Time",
    ylabel="Magnitude",
    figsize=(30,10),
    kind="bar",
)