Next Steps

Thanks for reading out the tutorial!

After you finish, try experimenting with other tools in the DASH toolbox. Some possible ideas include:

  • Run the demos with a different localization radius

  • Run the demos with inflation instead of localization

  • Compare the time to update the ensemble mean vs. mean and deviations

  • LGM Demo: Use PSM.estimate to generate validation proxies from the updated ensemble

  • LGM Demo: Incorporate proxy seasonality into the PSM inputs (Hint: See the checkSeasonality command from the BaySPLINE code)

  • NTREND Demo: Compute a temperature index from the updated field and compare it with the assimilated index

  • NTREND Demo: Withhold some proxies from the assimilation

  • Use the ensemble.useMembers command to rerun an assimilation with different ensemble members

  • Create an evolving prior

  • Make a framework for online assimilation

  • Or anything else that raises your interest

You can also read in depth about the commands in the DASH toolbox in the reference guide at:

dash.doc

We recommend reading the documentation of the different classes, as they contain outlines for working with their commands. A suggested reading order:

  • gridMetadata

  • gridfile

  • stateVector

  • ensemble

  • ensembleMetadata

  • PSM

  • kalmanFilter

  • particleFilter

  • optimalSensor

Finally, we welcome contributions and feedback. If you have any questions, find a command confusing, or have ideas for new features - don’t hesitate to reach out! You can contact the developers by sending an email to DASH.toolbox@gmail.com, or by submitting a pull request to the DASH repository.