Singh et al. (2026) Modelling framework for asynchronous land-atmosphere coupling using NASA GISS ModelE (NASA-GISS E2.1) and LPJ-LMfire (v1.4.0): design, application and evaluation for the 2.5 ka period

Identification

• Journal: Geoscientific model development
• Year: 2026
• Date: 2026-02-16
• Authors: Rohit K. Singh, Alexander Koch, Allegra N. LeGrande, Kostas Tsigaridis, Riovie D. Ramos, Francis Ludlow, Igor Aleinov, Reto Rüedy, Jed O. Kaplan
• DOI: 10.5194/gmd-19-1405-2026

Research Groups

• NASA Goddard Institute for Space Studies (GISS), New York, NY, USA
• Center for Climate Systems Research, Columbia University, New York, NY, USA
• Department of Earth Sciences, The University of Hong Kong, Hong Kong SAR, China
• Earth Observatory of Singapore, Singapore
• Department of History, and Trinity Centre for Environmental Humanities, School of Histories and Humanities, Trinity College, Dublin 2, Ireland
• SciSpace LLC, New York, NY, USA
• Department of Earth, Energy, and Environment, University of Calgary, Calgary AB, Canada

Short Summary

This study develops and evaluates a framework for asynchronously coupling the NASA GISS ModelE climate model with the LPJ-LMfire dynamic global vegetation model to simulate paleoclimate for the 2.5 ka period, demonstrating the critical importance of bias correction for accurate land-atmosphere feedback representation and consistency with proxy data.

Objective

• To present a generalized design for asynchronously coupling the NASA GISS ModelE2.1 climate model with the LPJ-LMfire dynamic global vegetation model to simulate climate and include biogeophysical land-atmosphere feedbacks.
• To demonstrate the utility of this asynchronous coupling framework for the 2.5 ka paleoclimatic period and evaluate the model results against independent paleoclimate reconstructions.

Study Configuration

• Spatial Scale: Global coverage. Atmosphere at 2° × 2.5° (latitude/longitude), ocean at 1° × 1.25° (latitude/longitude), and vegetation at 0.5° × 0.5° (horizontal resolution).
• Temporal Scale: 2.5 ka period (2500 years Before Present, approximately 550 BCE). Simulations were run iteratively, with each iteration spanning 100 to 1020 years until climate equilibrium was reached.

Methodology and Data

• Models used:
  • NASA GISS ModelE2.1 (NINT version) - Earth System Model
  • GISS Ocean v1 - Ocean Model
  • Ent Terrestrial Biosphere Model (Ent TBM) - Land surface component of ModelE2.1 (prescribed vegetation)
  • LPJ-LMfire (v1.4.0) - Dynamic Global Vegetation Model (DGVM)
• Data sources:
  • Forcing/Boundary Conditions: Orbital parameters (for 2500 years BP), greenhouse gas concentrations (e.g., CO2 at 271.4 ppm for 2.5 ka), natural aerosol emissions. Initial land cover derived from satellite (MODIS) data, TRY database (LAI), and GLAS (vegetation height), with PMIP4 protocol 6 ka vegetation linearly interpolated for "Green Sahara" conditions.
  • LPJ-LMfire Drivers: Monthly climate fields (temperature, precipitation, cloud cover, wind, lightning) from ModelE, static topography and soil texture maps, annual global atmospheric CO2 concentration. Interannual variability was introduced using detrended and randomized climate anomalies from the 20th Century Reanalysis. Wet days were estimated using CRU TS 4.0 data.
  • Validation Data: Multi-proxy temperature reconstructions (Kaufman et al., 2020) and speleothem-based oxygen isotope records (δ18Op) from the SISAL version 2 database (Comas-Bru et al., 2020).

Main Results

• The asynchronously coupled model system exhibits strong vegetation-albedo feedback on climate.
• The model system's performance is more sensitive to bias correction of climate model output than to internal model variability or "Green Sahara" initial conditions.
• Without bias correction, ModelE drifts towards colder conditions in high northern latitudes (e.g., >3 °C cooling over Northern Hemisphere high latitudes) due to increased ground albedo from LPJ-LMfire simulated land cover, leading to model instability.
• With bias correction, high northern latitudes show increased tree cover, decreased albedo, and a warming of 2–4 °C, aligning better with proxy data.
• Regional precipitation changes include a substantial intensification of the Summer Indian Monsoon (20–40% increase) and a drying pattern over Europe (10–25% decrease).
• Comparison with multi-proxy temperature reconstructions shows that bias-corrected simulations reproduce the sign of temperature anomalies with minimal global mean bias, while uncorrected simulations are significantly colder in the Northern Hemisphere.
• Simulated isotopic composition of precipitation (δ18Op) demonstrates excellent agreement with speleothem records, with pattern correlations greater than 0.83 globally, and particularly high skill over Europe (r_pat = 0.94, RMSE = 1.26).
• The NASA GISS ModelE is found to be particularly sensitive to the representation of shrubs, highlighting their importance as a potential climate driver.

Contributions

• Presents a novel, generalized, and computationally efficient framework for asynchronously coupling a climate model (NASA GISS ModelE2.1) with a dynamic vegetation model (LPJ-LMfire) to incorporate biogeophysical land-atmosphere feedbacks for paleoclimate simulations.
• Applies and evaluates this framework for the 2.5 ka period, addressing a less-studied but crucial interval for understanding human-environment interactions.
• Quantitatively demonstrates the critical importance of applying bias correction to climate model output in asynchronous coupling, showing its significant impact on simulated land cover and climate, especially in high northern latitudes, and its necessity for consistency with paleoclimate reconstructions.
• Highlights the specific sensitivity of the GISS ModelE to shrub representation, suggesting an area for future model development.

Funding

• National Science Foundation (Grant No. ICER-1824770)
• NASA GISS (institutional support)
• NASA High-End Computing (HEC) Program through the NASA Center for Climate Simulation (NCCS) at Goddard Space Flight Center
• Department of Earth Sciences at The University of Hong Kong
• European Research Council (grant agreement no. 951649, 4-OCEANS project)

Citation

@article{Singh2026Modelling,
  author = {Singh, Rohit K. and Koch, Alexander and LeGrande, Allegra N. and Tsigaridis, Kostas and Ramos, Riovie D. and Ludlow, Francis and Aleinov, Igor and Rüedy, Reto and Kaplan, Jed O.},
  title = {Modelling framework for asynchronous land-atmosphere coupling using NASA GISS ModelE (NASA-GISS E2.1) and LPJ-LMfire (v1.4.0): design, application and evaluation for the 2.5 ka period},
  journal = {Geoscientific model development},
  year = {2026},
  doi = {10.5194/gmd-19-1405-2026},
  url = {https://doi.org/10.5194/gmd-19-1405-2026}
}