A modeling study of land surface processes and surface energy budgets using the maximum entropy production theory

Abstract

Land surface model (LSM) plays an important role in numerical climate simulations. However, the existing LSMs have been found to produce inconsistent surface energy and water budgets due to the deficiencies in parameterization of land surface processes. In particular, surface heat flux parameterizations using the conventional gradient-based methods are subject to large modeling error and uncertainty. The primary goal of this study is to investigate the potential applications of an innovative approach, the Maximum Entropy Production (MEP) model of surface heat fluxes, in facilitating the understanding of land surface processes and global surface energy budgets. Specifically, two objectives, by applying the MEP model, are conducted to (1) improve model predictions of surface temperature, surface soil moisture, and near-surface air temperature for used in LSMs as well as climate models (2) reconstruct the global surface heat flux budgets. A coupled model of surface temperature, surface soil moisture, and near-surface air temperature is formulated based on the classical Force-Restore Method (FRM) incorporating the MEP model of surface heat fluxes, referred to as the FRMEP model. The FRMEP model is driven by surface net radiation and precipitation without explicitly using other meteorological variables and location specific empirical tuning parameters. The case studies suggest that the FRMEP model outperforms the classical FRMs, which are forced by observed or gradient-based parameterized surface heat fluxes. The FRMEP model well captures the diurnal and seasonal variations of surface temperature, surface soil moisture, near-surface air temperature, as well as surface heat fluxes. The results presented in this study justify the potential usefulness of the MEP model in climatic and hydrological studies. In this study, the 2001-2010 climatology of global surface heat flux budgets along with the corresponding trend and uncertainty is re-estimated using the MEP model and the input data from remote sensing observations and reanalysis data products. The MEP model generates the first dataset of global ocean surface conductive heat flux, which is not available from the existing data products. Global sublimation/deposition, sensible, and surface conductive heat fluxes over land snow-ice and sea ice covered areas are produced separately owing to the unique formulation of the MEP model. The uncertainties of MEP modeled surface heat fluxes are less than those of existing estimates and bounded by that of surface net radiation. Analysis of MEP heat fluxes suggests a global increase of land surface heat fluxes and a decrease of ocean surface heat fluxes during 2001-2010 consistent with the trends of surface radiation. The results indicate that the MEP model can be applied as an alternative approach to meet the challenge of monitoring and modeling global surface energy budgets. The new estimates of global surface heat fluxes based on the MEP model lead to a broader view of global energy and water cycles from a surface perspective.