idaweb

idaweb is an R package that provides a programmatic interface to the MeteoSwiss Open Data API for ground-based meteorological measurements. It lets you search for stations, parameters, and time ranges, and downloads the actual data directly into R.

NOTE: This package has been created to help short-cutting the (for my opinion rather cumbersome) way of accessing data through the Open Data Explorer provided by MeteoSwiss. So far it has been only me using it. Feel free to use it, but don't expect things to work the way you are using R (and don't expect any documentation including this README to be well-written (indeed any documentation is now AI-generated, lol)). If you like to use this package and have any improvement suggestions or feature ideas, open an issue or contribute to the package by opening a PR. I prefer code which has been thoroughly thought about (even if it is ugly-looking) over quick-and-dirty-ai coolness by no measures! Any code from me is (and will remain) AI-free.

The package is designed around a simple three-stage workflow:

1. Search for the data you need with met_search().
2. Inspect the resulting metadata with stations(), parameters(), and datainventory().
3. Download the data with get_data().

All data files are downloaded once and cached locally, so repeated analyses are fast.

---

Installation

Install the latest development version from GitHub with:

# install.packages("remotes")
remotes::install_github("ChHaeni/idaweb")

For compiled versions or a specific release, see the latest GitHub releases.

---

Package overview

Built-in metadata

idaweb ships with a pre-packaged metadata catalogue (metadata) covering all standard ground-based collections:

Collection	Short Name	Content
Automatic weather stations	smn	Temperature, Precipitation, Wind, Sunshine, Humidity, Radiation and Pressure
Automatic precipitation stations	smn-precip	Precipitation
Automatic tower stations	smn-tower	Temperature, Wind, Sunshine, Humidity and Radiation
Manual precipitation stations	nime	Precipitation and Snow
Totaliser precipitation stations	tot	Precipitation
Pollen stations	pollen	Pollen Concentration
Meteorological visual observations	obs	Visibility, Current and Past Weather, Ground Conditions and Clouds
Phenological observations	phenology	Phenophases of 26 Plant Species

You can inspect the metadata directly:

library(idaweb)

metadata
names(metadata)

# Access individual tables
stations(metadata)
parameters(metadata)
datainventory(metadata)

Quick example: Daily air temperature at Zollikofen

library(idaweb)

# 1. Search
mtemp <- met_search(
from = "12.08.2014 to 02.02.2026",
granularity = "D", # daily averages
lon = "7.43..7.49", lat = "46.96..47.12",
group = "Temperature"
)

mtemp
parameters(mtemp)

# 2. Narrow down to 2 m mean temperature
parameters(mtemp[[1]])[, "parameter_description_en"]

mfinal <- met_search(
description = "2mmean",
meta_data = mtemp
)

# 3. Download
zol_temp <- get_data(
meta_data = mfinal,
outclass = "data.table"
)

# The result is a nested list (collection -> station/granularity)
zol_temp[[1]][[1]]

Searching

By location — coordinates, altitude, station name, or canton:

# Coordinate box (WGS84)
meta <- met_search(lon = "7.4..7.5", lat = "46.9..47.3")

# Exact station abbreviation
meta <- met_search(abbr = "BER")

# Fuzzy name matching
meta <- met_search(name = "Zurich")

# By canton
meta <- met_search(canton = "BE")

Swiss coordinates (LV03 / LV95) are also accepted and automatically converted to WGS84 when the sf package is installed.

By parameter — group, description, short name, or unit:

# Temperature and precipitation at daily resolution
meta <- met_search(
group = c("temperature", "precipitation"),
granularity = "D"
)

# By exact parameter short name
meta <- met_search(shortname = "tre200s0")

# By description (fuzzy matching)
meta <- met_search(description = "2mmean")

By date and time — many formats are accepted:

meta <- met_search(from = "01.01.2020", to = "31.12.2020")
meta <- met_search(from = "2020-01-01", to = "2020-12-31")
meta <- met_search(from = "2020")

Downloading data

get_data() turns a met_metadata object into actual observations.

Output options:

• outclass: data.frame (default), data.table, or ibts
• outstruc: split-all (default), by-station, by-granularity, or cbind-all
• single_timestamp: TRUE gives a single time column; FALSE gives st/et
• tzone: Convert timestamps from UTC to another timezone

By default, files are cached in tempdir(). For persistence across R sessions, specify a dedicated directory:

dat <- get_data(meta, cache_dir = "C:/meteoswiss_cache")

Every returned data object carries metadata attributes:

meta <- met_search(from = '2020', shortname = 'tre200s0', name = 'Zol')
dat <- get_data(meta, outstruc = "cbind-all")
stations(dat)
# -> (station) name has been fuzzy matched |Z||o||l|likofen and |Z|ürich Aff|o||l|tern
parameters(dat)

---

More examples

10-minute wind data in the Canton of Jura

meta_jura <- met_search(
from = "12.08.2014", to = "12.08.2014",
canton = "JU",
granularity = "T",
group = "wind"
)

wind_jura <- get_data(meta_jura, outclass = "data.table")

Yearly precipitation in Adelboden

meta_adelb <- met_search(
from = "1999", to = "2025",
abbr = "ADEL",
granularity = "Y",
group = "precipitation"
)

adel_precip <- get_data(
meta_adelb[[1]],
outstruc = "cbind-all",
outclass = "data.table"
)

---

References

• MeteoSwiss Open Data – Ground−based measurements
• MeteoSwiss Open Data – Data download documentation

---

Contributing

This package was originally created to streamline personal access to MeteoSwiss data. If you use it and have suggestions, feature ideas, or bug reports, please open an issue or submit a pull request.