Fusion of Airborne Hyperspectral and LiDAR Remote Sensing Data to Study the Thermal Characteristics of Urban Environments - Computational Approaches for Urban Environments - page 275

CASI data (up to 1,050 nm) impaired the detection of some surface materials with
spectral key features occurring only at longer wavelengths (e.g., in the short-wave
infrared domain; Heiden et al. 2007 ;Heroldetal. 2007 ). Apart from these errors, the
overall accuracy of the material map is 78.3 %, and the kappa coefficient amounts to
0.76. Therefore, and according to other authors Altman ( 1991 ) and Grouven et al.
( 2007 ), the mapping result is considered as suitable for the parameterization of
urban microclimate models like ENVI-met.

11.5.2 Air Temperature Maps

The simulation results for the southern test site are compiled in Fig. 11.5 .The
displayed day- and nighttime temperatures were calculated at a height of 1.75 m.
The mean wind direction was east-southeast. Both simulations reproduce the typical
thermal behavior of urban environments over the course of a warm summer day
and night. At daytime, streets, driveways, parking lots, and the air layers above
these and similar land use elements are heating up most intensively because they
are almost constantly exposed to direct solar radiation and their physical properties
facilitate the absorption of high amounts of thermal energy. At nighttime, higher air
temperatures prevail in the slipstream of buildings and over broad streets because
construction materials release the energy that has been stored during the day with
less efficiency than natural surfaces. For both simulations, air temperatures around
trees and over larger patches of short vegetation are relatively low due to the cooling
effect of plant transpiration. Besides these observations, it is interesting to see that
temperatures in the northernmost block of the test site are considerably raised at
night. Obviously, the number and density of buildings in this block have exceeded
a critical threshold so that the cooling effect of trees and larger patches of short
vegetation has almost no impact on air temperature. The block's proximity to the
main street in the north further contributes to the observed thermal pattern.
The simulation results for the northern test site are compiled in Fig. 11.6 .
The displayed nighttime temperatures were calculated at a height of 1.75 m
aboveground. In this hypothetical pre-post comparison, two large parking lots are
replaced by multilevel parking garages (see the yellow buildings in Fig. 11.6 b) to
study the effect of such a planning measure on nocturnal air temperatures. The
mean wind direction was again east-southeast. The modeling outputs reveal that
the residential area in the slipstream of the original parking lots would heavily
suffer from increased heat stress due to the construction of further parking garages.
According to the simulations, people living in close vicinity to the fictional parking
garages would be exposed to raises in nighttime air temperature of up to 2 °C.
Moreover, additional model runs indicate that this observation is independent of the
height of the new parking garages. This case study highlights that even small-scale
changes or modifications of land cover and land use can have severe effects on the
urban microclimate and, by implication, also on thermal comfort (Toy et al. 2007 ),