cRacle: R tools for estimating climate from vegetation.

Area of Science:

• Paleoclimatology
• Ecological Informatics
• Computational Biology

Background:

• Estimating past and present climate conditions is crucial for understanding ecological and evolutionary processes.
• Traditional methods often rely on limited proxy data or complex modeling techniques.
• Biodiversity data repositories offer a vast, underutilized resource for climate reconstruction.

Purpose of the Study:

• To introduce and validate the 'cRacle' R package for climate estimation from vegetation data.
• To provide a user-friendly and robust tool for paleoclimate and climate reconstruction.
• To enhance the accessibility of climate estimation methods for a broader scientific audience.

Main Methods:

• Implementation of the Climate Reconstruction Analysis using Coexistence Likelihood Estimation (CRACLE) method within an R package.
• Utilizing plant community composition data from extant and fossil vegetation.
• Employing data access, aggregation, and non-parametric modeling techniques.
• Incorporating Generalized Boosted Regression (GBR) for model correction and bias reduction.

Main Results:

• The 'cRacle' package demonstrates superior performance compared to alternative methods in climate estimation.
• CRACLE estimates of mean annual temperature typically fall within 1°C of actual values under optimal parameters.
• GBR model correction significantly improved CRACLE's accuracy by reducing estimation bias.

Conclusions:

• CRACLE provides highly accurate climate estimates derived from modern plant community composition.
• The combination of non-parametric CRACLE modeling and GBR correction yields state-of-the-art results.
• The 'cRacle' R package democratizes climate reconstruction from plant data, facilitating wider research applications.