Pacheco‐Labrador et al. (2025) BOSSE v1.0: the Biodiversity Observing System Simulation Experiment

Identification

• Journal: Geoscientific model development
• Year: 2025
• Date: 2025-11-11
• Authors: Javier Pacheco‐Labrador, Ulisse Gomarasca, Daniel E. Pabon‐Moreno, Wantong Li, Mirco Migliavacca, Martin Jung, Grégory Duveiller
• DOI: 10.5194/gmd-18-8401-2025

Research Groups

• Environmental Remote Sensing and Spectroscopy Laboratory (SpecLab), Spanish National Research Council (CSIC), Madrid, Spain
• Max Planck Institute for Biogeochemistry, Jena, Germany
• Department of Environmental Science Policy and Management, UC Berkeley, Berkeley, CA, USA
• European Commission, Joint Research Centre, Ispra, Italy

Short Summary

This paper introduces BOSSE v1.0, a novel Observing System Simulation Experiment (OSSE) designed to simulate synthetic landscapes featuring diverse vegetation, associated remote sensing signals, and ecosystem functions. BOSSE aims to address the critical lack of consistent global ground diversity measurements, enabling the benchmarking and development of remote sensing methodologies for estimating plant functional diversity (PFD) and biodiversity-ecosystem function (BEF) relationships.

Objective

• To develop and present BOSSE v1.0, a modeling tool that simulates spatially explicit vegetation scenes where plant traits and ecosystem functions respond to meteorology, and simultaneously generates physically and physiologically connected remote sensing imagery (hyperspectral reflectance, sun-induced chlorophyll fluorescence, and land surface temperature).
• To provide a controlled environment for benchmarking and improving methods dedicated to estimating plant functional diversity (PFD) from remote sensing and exploring biodiversity-ecosystem function (BEF) relationships, facilitating advances in this research area.

Study Configuration

• Spatial Scale: Spatially explicit scenes (e.g., 60 pixels by 60 pixels) representing vegetation-populated landscapes, with individual plant phenotypes assigned to each pixel. Remote sensing imagery can be simulated at various spatial resolutions, from 100% (plant-to-pixel size ratio of 1:1) down to 10%.
• Temporal Scale: Hourly steps for simulations of remote sensing imagery and ecosystem functions. Phenological models run on 30-day averaged meteorological values. Meteorological data time series cover 2 years (2020–2022).

Methodology and Data

• Models used:
  • BOSSE v1.0 (Biodiversity Observing System Simulation Experiment)
  • SCOPE (Soil Canopy Observation, Photochemistry and Energy fluxes) - emulated for remote sensing imagery (hyperspectral reflectance factors, sun-induced chlorophyll fluorescence, land surface temperature) and most ecosystem functions (gross primary production, latent heat flux, sensible heat flux, net radiation, soil heat flux, light-use efficiency).
  • Growing Season Index (GSI) phenological model.
  • Gaussian Mixture Model (GMM) - for generating correlated plant traits and realistic meteorological conditions.
  • Neutral Landscape model - for spatial distribution of species.
  • Semi-empirical model (Migliavacca et al., 2011) - for ecosystem respiration.
  • Physical model (Wallace and Verhoef, 2000) - for friction velocity.
  • 2D interpolator - for soil resistance for evaporation.
  • Gaussian Point Spread Function (PSF) model - for degrading remote sensing imagery resolution.
• Data sources:
  • ERA5-Land hourly meteorological data (2020–2022).
  • European Space Agency’s Land Cover Climate Change Initiative (ESA LC-CCI) Global Plant Functional Types Dataset (v.2.08).
  • Köppen Climate Classification System maps (Rubel et al., 2017).
  • NACP MsTMIP simulations (Global 0.5-degree Model Outputs in Standard Format, Version 2.0) - for C3/C4 grass leaf area fraction.
  • Combined and gap-filled spectral libraries and TRY database (Kattge et al., 2020) - for training GMM for plant traits.
  • Bibliographic references for model parameters (e.g., Croft et al., 2017; Luo et al., 2019; Miner et al., 2016; Asner et al., 2003; Jones, 1998; Forkel et al., 2014; Gerten et al., 2004; Leuning et al., 2008).

Main Results

• BOSSE v1.0 successfully simulates spatially explicit scenes with varying species distributions ("clustered", "intermediate", "even") and plant functional types across different climatic zones.
• It generates realistic time series of plant functional traits (e.g., leaf area index, leaf chlorophyll content) exhibiting intra- and inter-specific variability and expected phenological responses to meteorology.
• The model produces physically and physiologically consistent remote sensing imagery (hyperspectral reflectance factors, sun-induced chlorophyll fluorescence, land surface temperature) and derived products (e.g., Normalized Difference Vegetation Index, Near-infrared reflectance of vegetation, retrieved optical traits) that align with vegetation properties and taxonomy.
• Degradation of spatial resolution significantly impacts the computation of functional diversity metrics (e.g., Rao’s quadratic entropy, fraction of α-diversity), showing complex biases (initial decrease then increase) as resolution coarsens.
• BOSSE simulates time series of various ecosystem functions (e.g., gross primary production, ecosystem respiration, net ecosystem productivity, light-use efficiency, latent heat flux, sensible heat flux) that follow expected seasonal behaviors.
• Emulators for remote sensing and ecosystem functions demonstrate performance consistent with expected uncertainties of real-world products and observations, without introducing significant bias in functional diversity metric computations.

Contributions

• BOSSE is the first Observing System Simulation Experiment (OSSE) specifically dedicated to studying plant diversity and biodiversity-ecosystem function (BEF) relationships from remote sensing.
• It provides a comprehensive, spatially explicit, and temporally dynamic modeling tool that addresses the critical lack of consistent, global, and spatially matched ground diversity measurements needed for benchmarking and developing remote sensing methodologies.
• The model simulates multiple spectral signals (hyperspectral reflectance, sun-induced chlorophyll fluorescence, land surface temperature) and ecosystem functions, establishing a systematic and physically-based link between plant traits, remote sensing signals, and ecosystem processes.
• BOSSE enables robust benchmarking and development of remote sensing methodologies for PFD estimation and BEF analysis in a controlled environment, allowing for flexible configuration (e.g., degrading sensor resolutions, testing different spatial patterns and climatic conditions).
• It supports the analysis and interpretation of real-world measurements and guides the design of future field campaigns by clarifying fundamental methodological questions and assessing the robustness of approaches to various confounding factors (e.g., spatial resolution, phenology, sensor characteristics).

Funding

• Living Planet Fellowship “Integrated Remote Sensing for Biodiversity-Ecosystems Function” IRS4BEF (ESA Contract no. 4000140028/22/I-DT-lr) of the European Space Agency.
• “Integrated Observing Systems and Simulation Experiments to Analyze Biodiversity-Ecosystem Function Relationships in Savanna Ecosystems” research project (PID2023-151046NB-I00 funded by MCI-U/AEI/10.13039/501100011033/FEDER, UE) funded by Ministerio de Ciencia, Innovación y Universidades.
• European Research Council (ERC) Synergy Grant “Understanding and modeling the Earth System with Machine Learning (USMILE)” under the EU Horizon 2020 research and innovation program (grant agreement no. 855187).

Citation

@article{PachecoLabrador2025BOSSE,
  author = {Pacheco‐Labrador, Javier and Gomarasca, Ulisse and Pabon‐Moreno, Daniel E. and Li, Wantong and Migliavacca, Mirco and Jung, Martin and Duveiller, Grégory},
  title = {BOSSE v1.0: the Biodiversity Observing System Simulation Experiment},
  journal = {Geoscientific model development},
  year = {2025},
  doi = {10.5194/gmd-18-8401-2025},
  url = {https://doi.org/10.5194/gmd-18-8401-2025}
}