Ensemble bias correction of climate simulations: preserving internal variability

Area of Science:

• Climate Science
• Environmental Modeling

Background:

• Climate simulations require bias correction for impact studies.
• Standard methods fail to distinguish bias from climate simulation uncertainties like scenario, model, and internal variability.
• Correcting multi-run ensembles individually conflates model bias with internal variability.

Purpose of the Study:

• Propose two novel ensemble bias correction (EnsBC) methods.
• Evaluate EnsBC approaches against standard individual member correction.
• Assess the preservation of internal variability and climate change signals.

Main Methods:

• Developed two EnsBC approaches designed to preserve ensemble internal variability.
• Applied and compared EnsBC methods to precipitation and temperature data from a North American regional climate ensemble.
• Evaluated internal variability preservation using monthly means and hourly quantiles, including in a changing climate context.

Main Results:

• Proposed EnsBC methods successfully preserve internal variability, unlike standard approaches.
• Internal variability is maintained even under changing climate conditions.
• Both EnsBC approaches conserve the original ensemble's climate change signal.

Conclusions:

• EnsBC methods offer a significant improvement over traditional bias correction for multi-run climate ensembles.
• These approaches ensure more reliable climate impact assessments by preserving crucial internal variability.
• The findings support the use of EnsBC for robust climate change research.