When different LSMs drive the same dynamic phenology module, which better simulates surface-to-atmosphere fluxes?
Rosero, The Jackson School of Geosciences at the University of Texas at Austin, firstname.lastname@example.org
Gulden, The Jackson School of Geosciences at the University of Texas at Austin, email@example.com
Yang, The Jackson School of Geosciences at the University of Texas at Austin, firstname.lastname@example.org
Niu, The Jackson School of Geosciences at the University of Texas at Austin, email@example.com
As part of the LBA-ECO Model Intercomparison Project, we evaluate the energy partitioning and carbon fluxes simulated by the Noah Land Surface Model (LSM) and the Community Land Model (CLM). We present fluxes simulated using 3 - 4 years of point-scale meteorological forcing for the three Santarem sites. Unique vegetation covers each of the three locations: the first site sits within moist tropical forest, the second in pasture-agriculture landscape, and the third in a selectively logged moist forest. Both Noah and CLM are augmented with the dynamic phenology module of Dickinson et al. (1998), which allows biomass to vary with changing environmental conditions by partitioning assimilated carbon into stems, leaves, and roots. Looking only at the Santarem 83 site, we use the model evaluation framework of Gulden et al. (2007) by minimizing parameter uncertainty between models before comparing model performance and by using a Monte Carlo approach to assess model robustness and performance within parameter space. We analyze differences between optimal values of parameters for the dynamic penology module and for other physically similar parameters and draw inferences about the dependence of the fluxes simulated by the dynamic phenology module on the underlying structure of the host LSM.