Abstract

Accurate representation of the interplay between ecosystems and climate in Earth system models (ESMs) is important for understanding and predicting changes in carbon, water, and energy cycles. The traditional land components of ESMs often rely on the categorization of plant functional types (PFTs), a method that may overlook the ecological variations within plant groups. Using the CliMA Land model with surface hyperspectral data primarily sourced from Manuscript draft submitted to Ecosphere-SHIFT collection 2 AVIRIS-NG (Airborne Visible/Infrared Imaging Spectrometer-Next Generation) during the NASA's SHIFT (Surface Biology and Geology High-Frequency Time Series) campaign, we derive leaf biogeochemical traits, and compare a trait-based land modeling approach to the commonly used PFT one. Our findings reveal significant spatial and temporal variations in traits and other related optical and photosynthetic parameters, not only within the same PFT but also across different seasons. Validation against TROPOMI Solar-Induced Fluorescence (SIF) at 740nm shows that the trait-based approach enhanced model accuracy due to better representation of phenology and spatial heterogeneity. Our findings highlight the limitations of traditional PFTbased approaches in ESMs and underscore the importance of incorporating detailed, trait-based data for a more accurate and comprehensive understanding of ecosystem functions.