The urban street canyon has been widely recognized as a basic surface unit in urban micrometeorological studies. Urban canopy models (UCMs), which quantify the exchange of energy and momentum between the urban surface and the overlying atmosphere, often adopt this type of street canyon representation as the fundamental surface element. Since UCMs can be coupled to regional-scale weather and climate models such as the Weather Forecast and Research Model (WRF), parametrizations of the surface momentum and scalar fluxes in UCM are of paramount importance. However, many current single-layer UCMs rely on empirical relations that were obtained over 80 years ago and often invoke the exponential wind profile derived from the existing literature for vegetation canopy. In this study, we conducted wallmodeled large-eddy simulations (LES) to study the forced (very weak buoyancy) convective heat transfer over idealized two-dimensional street canyons. It shows that the transfer efficiency computed following commonly applied resistance formulations can be one order of magnitude lower than LES results. The main reasons for the deviation include inaccurate wind speed parameterization and the use of a log-law based formulation for turbulent heat exchange between canyon air and the flow above.