Two micrometeorological methods utilizing high-frequency sampled air temperature were tested against eddy covariance (EC) sensible heat flux (H) measurements at three sites representing agricultural, agro-forestry, and forestry systems. The two methods cover conventional and newly proposed forms of the flux-variance (FV) and surface renewal (SR) schemes of differing complexities. The sites represent measurements in surface, roughness, and roughness to surface transitional layers. Regression analyzes against EC show that the most reliable FV and SR forms estimate H with slopes within ±10% from unity and coefficient of determination R2>0.9 across all the three sites. The best performance of both FV and SR was found at the agricultural site with measurements well within the surface layer, while the worst was found for the tall forest with measurements within the roughness sublayer where its thickness needed to be additionally estimated. The main variable driving H in FV is the temperature variance, whereas in SR, it is the geometry of ramp-like structures. Since these structures are also responsible for most of the temperature variance, a novel FV-SR approach emerging from combining the methods is proposed and evaluated against EC measurements and conventional FV and SR schemes. The proposed FV-SR approach requiring only a single fast response thermocouple is potentially independent of calibration and ameliorates some of the theoretical objections that arise when combining ramp statistics with similarity arguments. The combination of methods also provides new insights into the contribution of coherent structures to the temperature variance and its dependence on atmospheric stratification. Other potential utility of the new method is to include it in multi-tool assessments of surface energy fluxes, since a convergence or divergence of the results has a high diagnostic value.