New MJO reference data retrieval capability

Workshops, competitions and events AI Weather Quest
1

Dear all,

In addition to downloading training datasets, you can now directly download Wheeler and Hendon (2004) MJO reference data using:

retrieve_training_data.retrieve_MJO_projection_data(password)

This function retrieves:

  • Wheeler and Hendon (2004) combined EOFs (WH04_combinedEOFs.nc) for U200, U850 and OLR.
  • The observed RMM1 and RMM2 standard deviations (WH04_RMM_stddevs.nc).

These files provide the observational reference data required to compute real-time Multivariate MJO (RMM) indices from ERA5 or forecast data using the Wheeler and Hendon (2004) methodology. The function returns the local paths to both downloaded files and can optionally save them to a user-specified directory.

Further details are available in python package documentation. To enable download you will need the python package version > 3.28.

This new capability should support those developing MJO phase probabilities forecasts.

Thanks all,

Josh