Parser for METeorological Aerodrome Reports (METARs) and Terminal Aerodrome Forecasts (TAFs). This is a port of python-metar-taf-parser to Typescript with some additional features.
Features:
✈️ Complete METAR and TAF parsing- 🛠 Fully typed
- 🪶 Dependency free
- 🧪 Full test suite
- ✅ Runs anywhere: Web browser or Node
- 🌎 i18n: Translations
- 🌏 i18n: Handling international TAF & METAR report format differences
- 🌪 Remark parsing to human and machine readable formats
- 🗓
Forecastabstraction to easily query TAF reports byDate
pnpm i metar-taf-parser
# or
npm i --save metar-taf-parserThe parseMetar & parseTAF functions are designed to parse the raw report string into an object representation of a METAR/TAF.
If the payload begins with METAR or SPECI, that will be added as the type.
import { parseMetar } from "metar-taf-parser";
const metar = parseMetar(rawMetarString);
// -or-
// Optionally pass the date issued to add it to the report
const datedMetar = parseMetar(rawMetarString, { issued });👉 Note: One of the common use cases for TAF reports is to get relevant forecast data for a given
Date, or display the various forecast groups to the user. Check out theForecastabstraction below which may provide TAF data in a more normalized and easier to use format, depending on your use case.
import { parseTAF } from "metar-taf-parser";
const taf = parseTAF(rawTAFString);
// -or-
// Optionally pass the date issued to get the report issued and
// trend validity dates (start/end) on the report:
const datedTAF = parseTAF(rawTAFString, { issued });TAF reports are a little funky... FM, BECMG, PROB, weird validity periods, etc. You may find the higher level Forecast abstraction more helpful.
Forecast abstraction makes some assumptions in order to make it easier to consume the TAF. If you want different behavior, you may want to use the lower level parseTAF function directly (see above). Below are some of the assumptions the Forecast API makes:
-
The
validityobject found fromparseTAF'strends[]is too low level, so it is removed. Instead, you will findstartandendon the baseForecastobject. The end of aFMandBECMGgroup is derived from the start of the nextFM/BECMGtrend, or the end of the report validity if the last.Additionally, there is a property,
by, onBECMGtrends for when conditions are expected to finish transitioning. You will need to type guardtype = BECMGto access this property.const firstForecast = report.forecast[1]; if (firstForecast.type === WeatherChangeType.BECMG) { // Can now access `by` console.log(firstForecast.by); }
-
BECMGtrends are hydrated with the context of previous trends. For example, if:TAF SBBR 221500Z 2218/2318 15008KT 9999 FEW045 BECMG 2308/2310 09002KTThen the
BECMGgroup will also have visibility and clouds from previously found conditions, with updated winds.
Returns a more normalized TAF report than parseTAF. Most notably: while the parseTAF function returns initial weather conditions on the base of the returned result (and further conditions on trends[]), the parseTAFAsForecast function returns the initial weather conditions as the first element of the forecast[] property (with type = undefined), followed by subsequent trends. (For more, please see the above about the forecast abstraction.) This makes it much easier to render a UI similar to the aviationweather.gov TAF decoder.
import { parseTAFAsForecast } from "metar-taf-parser";
// You must provide an issued date to use the Forecast abstraction
const report = parseTAFAsForecast(rawTAFString, { issued: tafIssuedDate });
console.log(report.forecast);
⚠️ Warning: Experimental API
Provides all relevant weather conditions for a given timestamp. It returns an ICompositeForecast with a prevailing and supplemental component. The prevailing component is the prevailing weather condition period (type = FM, BECMG, or undefined) - and there will always be one.
The supplemental property is an array of weather condition periods valid for the given timestamp (any PROB, TEMPO and/or INTER) - conditions that are ephemeral and/or lower probability.
You will still need to write some logic to determine what data to use - for example, if supplemental[0].visibility exists, you may want to use it over prevailing.visibility, or otherwise present it to the user.
This function throws a TimestampOutOfBoundsError if the provided date is outside of the report validity period.
This example provides an array of hourly weather conditions over the duration of the TAF report.
import { eachHourOfInterval } from "date-fns";
import {
parseTAFAsForecast,
getCompositeForecastForDate,
} from "metar-taf-parser";
const report = parseTAFAsForecast(rawTAFString, { issued: tafIssuedDate });
const forecastPerHour = eachHourOfInterval({
start: report.start,
end: report.end,
}).map((hour) => ({
hour,
...getCompositeForecastForDate(hour, report),
}));The description property in the Remark is translated, if available.
import { parseMetar } from "metar-taf-parser";
import de from "metar-taf-parser/locale/de";
const rawMetarReport = "KTTN 051853Z 04011KT RMK SLP176";
const metarResult = parseMetar(rawMetarReport, {
locale: de,
});
console.log(metarReport.remarks[0].description);Remarks may be found on base TAF and METARs, along with TAF trends.
Each Remark will have a description (if translated), type and raw properties. There are additional properties for each unique remark, depending on the remark's type. We can type guard on type to access these unique properties.
If the remark is not understood, it will have type as RemarkType.Unknown, with raw containing everything until the next understood remark.
import { Remark, RemarkType } from "metar-taf-parser";
/**
* Find the sea level pressure given remarks, if defined
*/
function findSeaLevelPressure(remarks: Remark[]): number | undefined {
for (const remark of remarks) {
switch (remark.type) {
case RemarkType.SeaLevelPressure:
// can now access remark.pressure
return remark.pressure;
}
}
}Because certain abstractions such as flight category and flight ceiling can vary by country, this logic is left up to you to implement. However, if you're looking for somewhere to start, check out the example site (based on United States flight rules) in example/src/helpers/metarTaf.ts. Feel free to copy - it's MIT licensed.
Please see the example site README.md.
This project is based on python-metar-taf-parser and the parsing should be as similar to that project as possible. That being said, PRs are welcome.
This software port was made possible due to the fantastic work of @mivek in python-metar-taf-parser. If you like this project, please consider buying @mivek a coffee.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent