Verification and Case Studies for Urban Effects in HIRLAM Numerical Weather Forecasting - Meteorological and Air Quality Models for Urban Areas - page 132

Fig. 14.2 Sensitivity tests to urban features with the DMI-HIRLAM high resolution model are
shown as the difference fields (runs without vs. with modifications, i.e. urban roughness and anthro-
pogenic heat flux) for the (a) 10 m wind velocity and (b) 2 m temperature over the Copenhagen
and Malmö metropolitan areas on 18 Jun 2005, 18 UTC (See also Colour Plate 13 on page 177)
Malmo (Sweden) urban area even reaching a higher value. Anthropogenic heat flux
increased the temperature above the urban cells with higher values at night and in
evening hours and a minimum at noon. For both urbanised areas, this increase is, on
average, up to 1-1.5 ºC but with a large variance.
The diurnal variability of wind direction was modelled in all runs with prac-
tically no differences between the control and modified runs. When only anthro-
pogenic heat flux is included the wind velocity between 07-19 UTC is closer to the
observed local maximum compared with urban roughness run. Combining rough-
ness and anthropogenic heat flux should provide better results. For temperature,
when the anthropogenic heat flux is included, the fit to observational data is better
for urban compared with suburban stations. Moreover, additional inclusion of a stor-
age heat flux (using for example, the objective hysteresis model Grimmond et al.,
1991) can allow adjustments in the time shift of the temperature field observed on a
diurnal cycle, and especially during the transitional morning and evening periods.
It is important to remark that, in comparison with the original (non-urbanised)
model, the computational time for this urbanised version of the model is almost the
same this variant of the urban parameterisations in NWP models is computationally
very cheap.
14.3.2 Modified Urban Building Effects
The difference plots for wind velocity at 10 m and air temperature at 2 m (on exam-
ple of forecasts on 1st August 2004 at 06 UTC) are shown in Fig. 14.3 for both types
of urbanization.
For wind velocity, the effect of BEP urbanization is also well pronounced over the
metropolitan area showing a maximum of 1.5 m s − 1 in the morning, although it has
a smaller extension of the area where it is observed compared with anthropogenic
heat plus roughness urbanization. Throughout the day, the highest differences in