Evaluation of the Simple Urban Energy Balance Model Using Selected Data from 1-yr Flux Observations at Two Cities

The authors' objective was to apply the Simple Urban Energy Balance Model for Mesoscale Simulation (SUMM) to cities. Data were selected from 1-yr flux observations conducted at three sites in two cities: one site in Kugahara, Japan (Ku), and two sites in Basel, Switzerland (U1 and U2). A simple vegetation scheme was implemented in SUMM to apply the model to vegetated cities, and the surface energy balance and radiative temperature T(R) were evaluated. SUMM generally reproduced seasonal and diurnal trends of surface energy balance and T(R) at Ku and U2, whereas relatively large errors were obtained for the daytime results of sensible heat flux Q(H) and heat storage Delta Q(S) at U1. Overall, daytime underestimations of Q(H) and overestimations of Delta Q(S) and T(R) were common. These errors were partly induced by the poor parameterization of the natural logarithm of the ratio of roughness length for momentum to heat (kappa B(-1)); that is, the observed kappa B(-1) values at vegetated cities were smaller than the simulated values. The authors proposed a new equation for predicting this coefficient. This equation accounts for the existence of vegetation and improves the common errors described above. With the modified formula for kappa B(-1), simulated net all-wave radiation and T(R) agreed well with observed values, regardless of site and season. However, at U1, simulated Q(H) and Delta Q(S) were still overestimated and underestimated, respectively, relative to observed values.