Geoscience Reference
In-Depth Information
Table 11.1
Overview of the ENVI-met surface types used to represent the classes in the material
map
Roofing materials
Ground surfaces
Vegetation
Map class
Tiles
Tar-Bitumen
Concrete
Metal
Asphalt road
Bare soil
Wat er
Trees
Grass
C2
b
AL ST SD WW B7/07
c
FG
d
R1
Roofing: Tile,
BI
Bitumen,
C2
Concrete (lightweight),
AL
Aluminum,
ST
Asphalt road,
SD
Sandy soil,
WW
Deep water,
B7
Silver birch,
07
Norway maple,
FG
Grass
a
Class and its physical properties taken from Heldens (
2010
)
b
Used to model concrete roofs and walls
c
Used to model small (
BI
a
Model ID
R1
6 m) and large trees (>6 m), respectively
d
Plant height reduced to 15 cm
site comprised 170
100
30 grid cells. To parameterize the models, information on
the location and height of all land cover elements have to be provided. In addition,
the type of vegetation, wall and roofing materials of buildings, as well as the
surface material and soil type of non-built-up areas needs to be known (Table
11.1
).
Except for wall materials and soil types, which were defaulted to thin concrete
walls respectively sandy soil, all necessary information were available through the
material map and the LiDAR nDSM. After transferring these information from the
remote sensing products to ENVI-met, each model area was slightly simplified to
exclude smaller misclassifications in the material map from modeling.
As part of the second stage, some general settings and the meteorological
framework for each model run were specified. With regard to the general settings,
the spatial resolution of the model building blocks (2.5 m) and the geographic
location of the test sites were provided. Furthermore, the dates of the day- (6 am-6
pm) and nighttime (6 pm-6 am) simulations were set to June 22 and 23, 2012 (i.e.,
the acquisition dates of the CASI and LiDAR data), and the number of nesting
grids, which are used to minimize modeling uncertainties at the border of each
test site, was defined as three. Considering the meteorological framework, ENVI-
met requires hourly values of air temperature and humidity as well as the mean
wind direction for each modeling date. These information were taken from data of
a nearby climate station (John Dunn Helistop, Houston, Texas; Lott et al.
2001
).
After model preparation, the third and final stage of the approach comprised the
actual simulation task. While day- and nighttime simulations of air temperature
were carried out for the southern test site, two nighttime simulations were run for
the northern test site (what-if scenario).
11.5
Results and Discussion
The results of this study are the urban surface material map and the microclimate
modeling outputs. In the following, both results are described and discussed.
Search WWH ::
Custom Search