RainMaker.jl

A repository for the atmospheric general circulation model SpeedyWeather.jl to measure and document precipitation at a given location. In particular, this repository contains functions and a documentation to show off and rank model setups that make it rain as much as possible at a certain location (e.g. UK) within a certain period. New model setups can be submitted and are automatically evaluated and ranked in the documentation to create a leader board for the most successful rainmaker.

Measuring precipitation with RainGauge

RainMaker.jl exports RainGauge which is fully described in the documentation. In short,

using SpeedyWeather, RainMaker

# create a model
spectral_grid = SpectralGrid(trunc=31, nlayers=8)
model = PrimitiveWetModel(spectral_grid)

# add the rain gauge as callback
rain_gauge = RainGauge(spectral_grid, lond=-1.25, latd=51.75)
add!(model, rain_gauge)

# run the simulation
simulation = initialize!(model)
run!(simulation, period=Day(20))

# visualise 
RainMaker.plot(rain_gauge)

which will produce a Makie.jl plot

The RainMaker challenge

New to SpeedyWeather? Read the instructions to know which knobs you could change in SpeedyWeather in order to make it rain more or less. While the most successful rainmaker probably requires a combination of several knobs this is a good starting point.

Submit to the RainMaker challenge

Create a pull request with a julia script some_file_name.jl to be placed in the folder /submissions. After execution, there needs to be in the global scope of that script the following variables defined

• a rain_gauge::RainGauge having measured precipitation of a SpeedyWeather.jl simulation
• an author::String, e.g. author = "Kermit the Frog"
• a description::String describing your model setup in a few words (<6 probably), e.g. description = "3000m mountain in the atlantic" or "300K aquaplanet".

See Submit to the RainMaker challenge and Rules in the documentation for more details.

Leader board

The RainMaker challenge is used for teaching. After every such teaching sessions we tag a new release to have a version of the documentation with a leader board corresponding to that session. Afterwards we remove the submissions in the main branch so that for the next teaching session the group does not have to compete with the previous subsmissions. Also because we may make changes to the rules. Leaderboards and corresponding versions of RainMaker.jl are

• RainMaker.jl v0.1.1: Intelligent Earth CDT, Oxford, November 2024

Installation

Install Julia. Then, RainMaker.jl is a registered package, in the Julia REPL simply do

julia> ] add SpeedyWeather, RainMaker

where ] opens the package manager (the prompt changes to (@v1.11) pkg>). Alternatively, you can do using Pkg; Pkg.add(["SpeedyWeather", "RainMaker"]) to import the package manager Pkg and then to add the RainMaker and SpeedyWeather package. You can install just RainMaker and SpeedyWeather will be installed automatically, but in most cases you need SpeedyWeather explicitly anyway.