Ranit De (রণিত দে | रणित दे)

Parametric uncertainties are among the largest epistemic uncertainties in land surface models (LSM) simulating carbon fluxes, such as gross primary production (GPP). A persistent challenge is that inter–annual variability (IAV) is currently not well-represented in models, which can be attributed to temporally fixed parameters. We parameterized an optimality-based and a light use efficiency (LUE) model (1) per site–year, (2) per site, (3) per site with an additional component in the cost function representing IAV, (4) per plant functional type (PFT), and (5) globally by pooling data from all sites, using hourly eddy-covariance data from 198 sites. Furthermore, parameters representing a given environmental factor in models were grouped to investigate site-year variability of groups of parameters that contributed to improved simulation of IAV of GPP. Temporally varying parameters associated with soil-water availability and drought stress produced a better annual model performance for the optimality-based (median normalized Nash–Sutcliffe efficiency, viz. NNSE: 0.13) and LUE-based (median NNSE: 0.4) models, respectively. We found both high between-PFT and within-PFT variation of parameters, especially for parameters related to hydrology, for example, median available water capacities, which was generally high in broadleaf forests compared to savannas for both models. Though we found that temporally variable parameters can improve annual model performance, the temporal variability of parameters was not as high as their spatial variability. Our study suggests that discovering relations between the spatiotemporal variability of model parameters and their controlling factors can improve the representation of IAV in LSMs.

A long-standing challenge in studying the global carbon cycle has been understanding the factors controlling inter–annual variation (IAV) of carbon fluxes, and improving their representations in existing biogeochemical models. Here, we compared an optimality-based model and a semi-empirical light use efficiency model to understand how current models can be improved to simulate IAV of gross primary production (GPP). Both models simulated hourly GPP and were parameterized for (1) each site–year, (2) each site with an additional constraint on IAV (Cost^IAV), (3) each site, (4) each plant–functional type, and (5) globally. This was followed by forward runs using calibrated parameters, and model evaluations using Nash–Sutcliffe efficiency (NSE) as a model-fitness measure at different temporal scales across 198 eddy-covariance sites representing diverse climate–vegetation types. Both models simulated hourly GPP better (median normalized NSE: 0.83 and 0.85) than annual GPP (median normalized NSE: 0.54 and 0.63) for most sites. Specifically, the optimality-based model substantially improved from NSE of -1.39 to 0.92 when drought stress was explicitly included. Most of the variability in model performances was due to model types and parameterization strategies. The semi-empirical model produced statistically better hourly simulations than the optimality-based model, and site–year parameterization yielded better annual model performance. Annual model performance did not improve even when parameterized using Cost^IAV. Furthermore, both models underestimated the peaks of diurnal GPP, suggesting that improving predictions of peaks could produce better annual model performance. Our findings reveal current modelling deficiencies in representing IAV of carbon fluxes and guide improvements in further model development.