A Bayesian Network Approach to Weather‐Dependent Mechanisms and Dynamic Corrections of Energy Balance Closure in Summer Maize Fields

ABSTRACT

This study evaluated daytime surface energy balance closure (EBC) in a summer maize field in Daxing, Beijing, during 2022–2024. Ordinary least squares (OLS) and standardised major axis (SMA) regression were used to assess the effects of horizontal thermal advection (A), additive damping (kB ⁻¹ ) and aerodynamic roughness length (Z _0m ) on EBC. A Bayesian network (BN) sensitivity analysis was then applied to explain interannual differences in correction performance and to identify the dominant controls and pathways under clear and cloudy/rainy conditions. The OLS and SMA results showed that the combined correction of Z _0m , kB ⁻¹ and A was the most effective, increasing EBC by more than 26% across the study period, with Z _0m contributing more strongly than the other single factors. The BN analysis further indicated that the effectiveness of EBC correction depended strongly on interannual meteorological variability. Correction performance was stronger under clear conditions than under cloudy/rainy conditions, and the dominant pathways differed markedly between the two weather regimes. These findings provide new insight into the weather‐dependent mechanisms of daytime surface energy balance closure in summer maize fields.