/mf_modelling

Multifidelity Climate Modelling (Practical for "ML and the Physical World" module)

Primary LanguagePython

Multifidelity Climate Modelling / Climate Downscaling

Open In Colab

Data Challenge

To provide probabilistic predictions of temperature on the high-fidelity domain (Panel b) from low-fidelity wind and temperature data (Panel a). Elevation data for the high-fidelity domain is also avaiable to help learn the temperature change with height.

Data

You can find the data here

Other potentially useful datasets

  • Climate Data Guide - Large-scale climate phenomena and oscillations that may relate to local climate variability, e.g., Nino 3.4
  • Weather stations - local point-based measurements which provide highly localised timeseries of the atmospheric conditions.

Schematic of data challege