<p>Exploring surface energy exchange is critical for evaluating agricultural development and the stability of regional water resources in paddy-upland rotation areas of low-latitude plateau lakesides. Traditional eddy covariance (EC) technology struggles to capture large-scale turbulence, leading to challenges in surface energy balance closure, which is a key knowledge gap in understanding land-atmosphere interactions in this region. This study collected 2014–2015 water and heat flux data in an Erhai Lake lakeside fava bean-rice rotation area using EC and large aperture scintillometer (LAS) systems, focusing on their performance in quantifying energy components and improving closure. The results showed that surface albedo was highest in the fava bean period (0.22), lowering its net shortwave radiation (163&#xa0;W/m²) versus the rice period (181&#xa0;W/m²); rice had the highest daily evaporation (3.72&#xa0;mm) driven by latent heat flux. The weakening effect of clouds on incident solar radiation (12.5% decrease) is weaker than their enhancing effect on longwave radiation (14.2% increase). H_LAS (LAS-derived sensible heat flux) exceeded H_EC (EC-derived) by 20.9&#xa0;W/m² (fava bean period) and 14.1&#xa0;W/m² (transition period), and increased by 51% (dry season) and 82% (wet season) under stable stratification. LAS improved energy closure by capturing low-frequency turbulence (especially over fava bean and sunny conditions), while EC underestimated sensible heat flux. Differences were observed in the sensitivity of energy components to soil moisture, where the energy components during the fava bean period responded to soil moisture levels of ≥ 0.2 m³/m³ and those during the rice period responded to soil moisture levels of &lt; 0.4 m³/m³. This study highlights LAS’s novelty in enhancing flux quantification in rotation systems, providing insights into land-atmosphere interactions in low-latitude plateau lakeside regions and a methodological reference for heterogeneous agricultural energy balance research.</p>