Urban impacts on climate - The Global Climate System: Patterns, Processes, and Teleconnections - page 229

or climate elements such as temperature and humidity. For example, it is
possible for Q Hu r to be negative (see Figure 7.7c ), resulting in less boundary
layer heating in cities surrounded by dry landscapes. Whether this results in a
shallower UBL depends on the relative roles of heat flux and mass conver-
gence (due to dynamic processes, roughness, and barrier effects) in producing
uplift. The Sacramento results show that choosing a rural site, which is a
reasonable analogue of pre-urban conditions, gives radically different out-
comes to that using a site characteristic of the modern agriculturally managed
landscape (Figure 7.7 ).
7.3.5 Investigating urban climates
Conducting measurements in urban areas requires all the quality assurance
and quality controls of any measurement campaign. Often, routine observa-
tions become more difficult in urban area and other issues need to be
addressed. Moreover, many cities are located in complex settings: along a
coast or shoreline, the confluence of river valleys, on a hill, in a sheltered
basin. Each of these settings carries with it an additional set of local climate
influences that complicate the study of urban areas.
A critical question in any urban climate study, intricately tied to the scale of
interest, is where does it make sense to locate instruments to obtain a
representative measurement? Not only does the height of the instruments
relative to the roughness elements (Figure 7.4 ) have to be considered, but also
the spacing of the elements around the sensor location and the orientation of
streets and buildings (which will affect radiation geometry and diurnal and
seasonal heating patterns and wind speeds) need to be taken into account. If
the typical micro-scale unit is the urban canyon (Figure 7.4 ), the 1-2m height
of instruments in a standard weather screen is less than the height of buildings
and much urban vegetation. In cities, there is often the added complication
that instruments have to be mounted higher above the surface to avoid
vandalism or to be out of the way of traffic or pedestrians. The field of
view of urban-installed sensors may be less open than what would be recom-
mended for a standard climatological/meteorological station. All of these
things will introduce bias that the user of the data must be aware of when
interpreting data.
An alternative approach to studying urban climates involves the use of
models. These range from physical hardware models to study micro- and
local-scale effects, for example wind tunnel studies of air flow or studies of
the thermal effects of parks (Spronken-Smith and Oke 1999 ), to numerical
models to simulate energy, mass, and momentum exchanges. Many larger
scale numerical models represent urban areas by just changing the thermal
materials of the ''soil,'' representing the urban area as concrete. Comparisons