Time Series

Meteostat provides access to time series data for thousands of weather stations worldwide. These time series are provided in different granularities and can be consumed through a single interface.

🚀 Example

Hourly

Let's fetch some hourly data for Atlanta, USA from January 1 to December 31, 2018:

# Import Meteostat library and dependencies
from datetime import datetime
import meteostat as ms

# Set time period
start = datetime(2018, 1, 1)
end = datetime(2018, 12, 31, 23, 59)

# Get hourly data
ts = ms.hourly(ms.Station(id='72219'), start, end)
df = ts.fetch()

# Print DataFrame
print(df)

Daily

Let's fetch some daily temperature data for Frankfurt, Germany in 2018:

# Import Meteostat library and dependencies
from datetime import date
import meteostat as ms

# Set time period
start = date(2018, 1, 1)
end = date(2018, 12, 31)

# Get daily data
ts = ms.daily(ms.Station(id='10637'), start, end)
df = ts.fetch()

# Print DataFrame
print(df)

Monthly

Let's fetch some monthly data for Frankfurt, Germany from 2000 to 2018:

# Import Meteostat library and dependencies
from datetime import date
import meteostat as ms

# Set time period
start = date(2000, 1, 1)
end = date(2018, 12, 31)

# Get monthly data
ts = ms.monthly(ms.Station(id='10637'), start, end)
df = ts.fetch()

# Print DataFrame
print(df)

👀 Learn More

📄️ Hourly Weather

If you want to investigate the weather on a particular day or a short period of time, hourly time series data is a great choice.

📄️ Daily Weather

Daily time series data is ideal for analyzing weather patterns over longer periods, such as weeks, months, or years.

📄️ Monthly Weather

Monthly time series data is perfect for examining long-term climate trends and seasonal patterns over extended periods, such as years or decades.

📄 Source Code

The source code of the TimeSeries class is available on GitHub.

🔍 API

Interface

Through one of the following:

• meteostat.hourly
• meteostat.daily
• meteostat.monthly
• meteostat.normals

Parameters

Please refer to the respective interface documentation for a list of accepted parameters.

Properties

granularity

The time series's granularity.

Data Type

Granularity

---

stations

Included weather stations.

Data Type

DataFrame

---

start

Start date of the requested period.

Data Type

datetime

---

end

End date of the requested period.

Data Type

datetime

---

timezone

Time zone of the period and records (only hourly granularity).

Data Type

str

---

parameters

Included meteorological parameters.

Data Type

List[Parameter]

---

freq

The time series's frequency (e.g. 1h in case of hourly granularity).

Data Type

str

---

empty

Is the time series empty?

Data Type

bool

---

providers

Included data providers.

Data Type

List[Provider]

---

licenses

Applicable licenses.

Data Type

List[License]

---

attribution

The attribution/copyright string.

Data Type

str

---

commercial

Can data be used for commercial purposes?

Data Type

bool

---

valid

Does the time series pass all quality checks?

Data Type

bool

---

lapse_rate

The temperature's lapse rate, if applicable (at least two included weather stations with sufficient data).

Data Type

float or None

Methods

validate

Check if the time series passes all quality checks.

Attributes

This method does not accept any attributes.

Return Value

bool

Example

from datetime import date
import meteostat as ms

# Specify time range
START = date(2018, 1, 1)
END = date(2018, 12, 31)

# Get daily data
ts = ms.daily('10637', START, END)
is_valid = ts.validate()

print(f'Time series valid: {is_valid}')

---

fetch

Fetch the actual weather/climate data.

Attributes

Attribute	Description	Type	Default
squash	Squash data from different sources	bool	True
fill	Fill missing records	bool	False
sources	Include source columns?	bool	False
location	Add location-related columns (latitude, longitude, elevation)	bool	False
clean	Remove inaccurate data	bool	True
humanize	Humanize data values for wind direction and condition code	bool	False
units	Unit system for data values (e.g., metric or imperial)	UnitSystem	UnitSystem.METRIC

Return Value

DataFrame

Example

from datetime import date
import meteostat as ms

# Specify time range
START = date(2018, 1, 1)
END = date(2018, 12, 31)

# Get daily data
ts = ms.daily('10637', START, END)
df = ts.fetch()

# Print the DataFrame
print(df)

---

count

Get the number of rows in the whole time series or by parameter. NaN values are excluded.

Attributes

Attribute	Description	Type	Default
parameter	The parameter which should be counted, if None the whole time series is considered	Parameter or str	None

Return Value

int

Example

from datetime import date
import meteostat as ms

# Specify time range
START = date(2018, 1, 1)
END = date(2018, 12, 31)

# Get daily data
ts = ms.daily('10637', START, END)
count = ts.count(ms.Parameter.PRCP)

# Print the DataFrame
print(f'{count} rows with precipitation data')

---

completeness

The share of non-NaN values of the time series's full length.

Attributes

Attribute	Description	Type	Default
parameter	The parameter of interest, if None the whole time series is considered	Parameter, str or None	None

Return Value

float

• 0 means all data is missing
• 1 means all data is available

Example

from datetime import date
import meteostat as ms

# Specify time range
START = date(2018, 1, 1)
END = date(2018, 12, 31)

# Get daily data
ts = ms.daily('10637', START, END)
completeness = ts.completeness(ms.Parameter.PRCP)

# Print the DataFrame
print(f'Precipitation data completeness: {completeness * 100}%')