248 Expanding Geology’s Horizons: Geoinformatics, Open Science, and Open Data

This session highlights the development of new methodologies for collecting, sharing, and analyzing geological datasets and the advancement of FAIR (findable, accessible, interoperable, and reusable) data practices.

Wednesday, 22 October 2025
8:00 AM - 12:00 PM
HBGCC, 301C

Predicting floral composition with climate: Testing climate models with plant distributions

Paper number: 248-2
Program: Technical Sessions
Time: 8:35 AM – 8:50 AM

Authors

Abstract

Fossil floras from high latitudes often include taxa indicative of much warmer climates, suggesting that such assemblages reflect macroclimatic conditions not present today. Earth System Models (ESMs) built on physical and chemical processes have provided paleoclimate estimates for which fossil-inferred climates can be compared. Paleobotanists typically estimate paleoclimates from plant fossils, assuming that extinct species had the same climatic requirements as their extant relatives. In this study, we proposed a method to evaluate the climate-plant correlation without assuming modern plant distributions reflect climate, relying instead on the assumption that climate has historically been a primary driver of plant distributions. We use gridded historical distributions of North American tree taxa alongside pre-industrial simulated climate from the Community ESM, version 1.3 (CESM 1.3), at three spatial resolutions, to train a XGBoost model that predicts the presence or absence of tree taxa at family, genus, and species taxonomic levels based on monthly temperature, rainfall, and solar radiation. Results show the CESM climatic variables predict North American floral composition with >90% overall accuracy, increasing at higher latitudes, where taxa seem to have broader, more continuous geographical ranges. Lower accuracy (<50%) occurred along the Rocky Mountains, where taxon ranges are spatially restricted or scattered, and climate varies abruptly with elevation. Prediction performance was strongest (>95%) at the finest grid resolution (625km2) versus the coarsest grid (40,000km2; 85-95%). Taxonomic performance was slightly better at the family and genus levels than species. We are currently extending this method to fossil pollen records to test Paleogene climates simulated by the latest generation of ESMs. This approach provides climatic context for past plant communities and evaluates the importance of temperature and moisture availability in different environments.

Geological Society of America Abstracts with Programs. Vol. 57, No. 6

doi: 10.1130/abs/2025AM-413217

© Copyright 2025 The Geological Society of America (GSA), all rights reserved. Permission is hereby granted to the author(s) of this abstract to reproduce and distribute it freely, for noncommercial purposes. Permission is hereby granted to any individual scientist to download a single copy of this electronic file and reproduce up to 20 paper copies for noncommercial purposes advancing science and education, including classroom use, providing all reproductions include the complete content shown here, including the author information. All other forms of reproduction and/or transmittal are prohibited without written permission from GSA Copyright Permissions.