Micro-weather forecasting for astronomy

astronomy, forecast-systems, gfs, python, weather-data
pip install ucast==0.1.3


Actions Status PyPI version


µcast (Unix name ucast starts with the Roman letter "u") is a toolkit for micro-weather forecast for astronomy. It provides a high-level and developer-friendly interface to combine weather data and radiative transfer models to enable accurate micro-weather forecast at telescope sites. It is built on top of other projects including am for atmospheric modeling.


µcast is a pure python package that can easily be installed by pip install ucast. However, in order to use am Atmospheric Model, the am needs to be in your $PATH. After installation, one can simply import ucast to use µcast inside python. µcast also comes with a command line tool ucast to automatically pull GFS data, create summary tables of atmospheric properties, and plot the results:

$ ucast mktab KP # create weather forecast table
$ ucast psite KP # create summary plot for one site
$ ucast pall     # create summary plot for all sites
$ ucast vis      # create a bokeh visualization

Use ucast mktab --help, ucast psite --help, etc to see the detailed usages.

Backend Tools

Weather Data and Forecast Systems

µcast supports different weather data and forecast systems.

Radaitive Transfer Backends

µcast supports multiple atmospherical model backend for radiative transfer.

  • am Atmospheric Model

    am Atmospheric Model is a tool for radiative transfer computations at microwave to submillimeter wavelengths. It is µcast's default model for radio astronomy.