Urban impacts on climate - The Global Climate System: Patterns, Processes, and Teleconnections - page 244

of temperature vary in relation to the surface energy balance. However, the use
of remote sensing has also created important complications to the determination
and interpretation of UHI.
The remote sensor is not measuring the urban-rural contrast in air temperature
(AUHI), but the ''upwelling thermal radiance'' (Voogt and Oke 2003 ) from the
surface structures at ground level (SUHI). Thus the satellite ''sees'' a different
temperature regime than a ground-based sensor, and integration between the two
can be difficult. Also, SUHI is greatest during the day, while under ideal
conditions (clear skies, high pressure, light winds) AUHI is greatest at night.
During the day, land-cover distribution is very important to SUHI. At night the
importance of land cover diminishes in the development of AUHI (Roth et al.
1989 ), and surface themal release becomes more important. For example, for
Huntsville (Alabama, USA) irradiance from surface land uses during the
day ranged from 70Wm 2 for the commercial district to 55Wm 2 for water
and rural areas, based on 5-meter resolution from ATLAS. At night, except
for the water surface, irradiance from all surfaces ranged from 43 to 47Wm 2
(Lo et al. 1997 ).
A major part of the reason for this difference is the complexity of heat transfer
between a solid surface and a fluid atmosphere. Radiative and convective heat
transfers occur initially from a surface layer into a volume containing the atmo-
sphere. Under strongly convective conditions, the volume will be very large, and
the heat flux will mix and dilute rapidly, with little resultant impact on atmo-
spheric temperature. Under stable atmospheric conditions at night, with clear
skies and very light winds, the volume can be much smaller, but heat transfer
will still be diluted. If there is a near-surface inversion, the UCL layer can be
decoupled from the UBL above (Figure 7.5 ), and mixing to the upper layers may
be prevented.
The thermal sensor does not see the entire urban surface. Rather, depending on
surface structure and morphology, the angle of the Sun, and the sensor viewing
geometry, the thermal emissions detected can be spatially limited. Depending on
the three-dimensional structure of the urban landscape, the sensor may only see
rooftops, treetops, flat parking lots, etc. While the measurements from the
satellite sensor should be more representative of horizontal surfaces than the
vertical sides of buildings (Nichol 1996 ), the thermal results are likely to be
unrepresentative of the full urban surface.
Table 7.4 lists many of the important considerations that must be addressed
for accurate use of remote sensing data. Corrections for the atmospheric inter-
ferences and surface structure are essential or considerable interpretation error
may result. Matching the scale of the thermal observations with that of the
surface features provides much more representative results. For example,
AVHRR satellite imagery has a coarse resolution of about 1.1 km. Studies trying
to explain SUHI and AUHI distributions using AVHRR have not been very
successful. However, Nichol ( 1996 ) used LANDSAT's thematic mapper, with a