Skip to main content

TAF (Terminal Aerodrome Forecast) parser and decoder

Project description

pytaf is python module for parsing and decoding aviation weather forecasts


TAF stands for “Terminal Aerodrome Forecast”. It’s the weather forecast reporting format used in aviation.

Unlike “normal” weather forecasts for everyday use, TAF is issued for no more than next 24-30 hours, and includes information critical for flight safety, such as exact clouds type and ceiling.

This is what a TAF report from the United States may look like:

  AMD KMKE 172034Z 1721/1824 14013G19KT P6SM SCT028 BKN035 BKN250
 FM180100 17008KT P6SM SCT035 BKN120
 FM181000 17007KT P6SM VCSH BKN040 OVC080
  TEMPO 1811/1815 6SM -TSRA BR BKN030CB
 FM181500 18009KT P6SM VCSH BKN050
 FM182100 16012KT P6SM SCT050 BKN150

What it means:

"TAF AMD KMKE": TAF amended for General Mitchell International Airport,

"172034Z": Issued at 20:34 Zulu time (UTC) on [September the] 17th

"1721/1824": Valid from 21:00 UTC on 17th to 24:00 UTC on 18th

"14013G19KT": Wind from 140 degrees at 13 knots, gusting to 19 knots

"P6SM": Visibility more than 6 statute miles

"SCT028 BKN035 BKN250": Clouds scattered at 2800 (28*100) feet, broken at 3500 feet, broken at 25000 feet

"FM180100 ...": From 01:00 UTC on 18th [wind, visibility, clouds]

"FM181000 17007KT P6SM VCSH BKN040 OVC080": From 10:00 UTC on 18th [wind, visibility],
      showers in the vicinity, clouds broken at 4000, overcast at 8000

"TEMPO 1811/1815 6SM -TSRA BR BKN030CB": Temporarily between 11:00 UTC and 15:00 UTC on 18th
       visibility 6 statute miles, light thunderstorms and rain, mist, broken cumulonimbus clouds at 3000


Note: despite the terse format, TAF reports are written by humans, and are supposed to be interpreted by humans too.

This means the format (from computer point of view) is loosely standardized and is not guaranteed to be machine readable.

Moreover, TAF format is not exactly the same in all countries. Not different enough for people to have problems understanding it, but different enough to make implementing a universal parser very hard, if not impossible.

United Stated civil airports produce the most machine friendly and standardized reports and those are the most likely to be interpreted correctly. Effort was made to interpret European Union civil airport reports properly, but they exhibit more regional variations, so the interpretation may be incomplete. Remember that the interpretation is provided for information purposes only and should not be used for flight planning (at least not without inspecting the original undecoded report).

Information about the US TAF format can be found at NOAA website:

You can get raw and interpreted reports from there too:


pytaf contains two base classes: pytaf.TAF and pytaf.Decoder

The constructor of the TAF class takes a string, takes it apart, and stores raw values in a TAF object.

The Decoder is initialized with a TAF object and provides decode() method that returns a string that contains a human-readable interpretation.

import pytaf

taf = pytaf.TAF("<my TAF string>")
decoder = pytaf.Decoder(taf)

This is what a decoded string may look like:

TAF for KSFO issued 05:46 UTC on the 20th, valid from 06:00 UTC on the 20th to 12:00 UTC on the 21st
Wind: variable at 04 knots
Visibility: more than 6 statute miles
Sky conditions: few clouds at 1000 feet, scattered clouds at 1500 feet, broken clouds at 20000 feet


If you want to change the decoder output format (e.g. output to HTML), inherit from pytaf.Decoder and overload the _decode_taf() method. That method contains nothing but calls to other methods and output string formatting.

The Decoder class provides methods for decoding every type of weather information independently, so you can easily combine them to wrap that information in any desirable format.

If you want to redefine the interpretation, e.g. produce numeric values for display in a widget rather than plain english descriptions, you may want to use the TAF object directly. All its methods return dicts with pretty straightforward key names.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pytaf-1.1.1.tar.gz (9.7 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page