Progress of AI-weather models over time
Below are two charts showing the progress of global AI-weather models over time, for 500hPa geopotential (Z500) and 850hPa temperature (T850). The purpose of...
Blog
2024
2021
“If you have a hammer…” - 5 question to ask before starting a machine learning project
There is a saying that “if you have a hammer everything looks like a nail”. The lesson, of course, being that, even though hammers are an amazing tool for ha...
2020
The key challenge for machine learning parameterizations of clouds
The argument goes like this: Climate models continue to have large uncertainties → these uncertainties are primarily caused by the parameterization of sub-gr...
WeatherBench: A benchmark dataset for data-driven weather forecasting
It seems that in 2020 no domain of science is safe from the AI hype. This also includes weather and climate science where publications and conferences on AI ...
2019
Sugar, Flower, Fish or Gravel - Now a Kaggle competition
I am very happy to announce the launch of our Kaggle competition “Understanding Clouds from Satellite Images”. This competition is the culmination of literal...
Lorenz ‘96 is too easy! Machine learning research needs a more realistic toy model.
Clicking on the Binder button will open an interactive notebook, in which you can reproduce all visualizations and results in this post.
Deep Learning (for Atmospheric Scientists): How to get started
Updated version of an earlier post after version 3 of fast.ai has been released.
2017
A simple hack to make figures reproducible
In my previous posts I explained why reproducibility is an important topic that is probably underappreciated in the weather and climate sciences at the momen...
Reproducibility, Part 2 - My experiences
In my previous blog post I explained what reproducibility is, why it is important and how it could work in a practical sense. In this post, I will outline my...
Reproducibility, Part 1 - What is it and how to start?
Before I start let me thank and give credit to Sebastian Lerch of the Heidelberg Institute for Theoretical Studies (HITS), whose slides on the topics served ...
Thoughts on shallow cumulus through an airplane window
Recently, I found myself on a plane crossing the Atlantic. As I was looking out of the window I was struck by some spectacular views of the clouds beneath us...