English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Progress in hybrid physics-machine learning (ML) climate simulations has beenlimited by the difficulty of obtaining performant coupled (i.e. online)simulations. While evaluating hundreds of ML parameterizations of subgridclosures (here of convection and radiation) offline is straightforward, onlineevaluation at the same scale is technically challenging. Our softwareautomation achieves an order-of-magnitude larger sampling of online modelingerrors than has previously been examined. Using this, we evaluate the hybridclimate model performance and define strategies to improve it. We show thatmodel online performance improves when incorporating memory, a relativehumidity input feature transformation, and additional input variables. We alsoreveal substantial variation in online error and inconsistencies betweenoffline vs. online error statistics. The implication is that hundreds ofcandidate ML models should be evaluated online to detect the effects ofparameterization design choices. This is considerably more sampling than tendsto be reported in the current literature.

Systematic Sampling and Validation of Machine Learning-Parameterizations in Climate Models