Aucun titre de diapositive

1
APPLICATION OF THE URBAN VERSION OF MM5 FOR
HOUSTON
Sylvain Dupont Collaborators Steve Burian, Jason
Ching E-mail dupont_at_hpcc.epa.gov
University Corporation for Atmospheric Research
2
Urbanized version of MM5
DA-SM2-U
Soil Model for Sub-Meso scales - Urbanized version
Drag-Force Approach

• Simulation of meteorological fields within and
  above rural and urban canopies.
• Neighborhood scale 1-km horizontal grid spacing
  and few meter vertical grid spacing inside the
  canopy
• Canopy elements are not explicitly defined but
  spatially averaged

3
General Objective
Modeling air-quality for estimating human
exposure to air pollution in urban areas by using
CMAQ.
Specific Objectives of this presentation

• Computation of the Houston morphological
  parameters for DA-SM2-U.
• Influence of the Houston representation on the
  urban boundary layer structure
• What degree of urban representation detail do we
  need at neighborhood scales ?

4
DA-SM2-U
Introduced inside the Gayno-Seaman PBL scheme
5
Momentum equation forcing terms (modification
of vertical turbulent transport term) momentum
sources due to building horizontal surfaces
(friction force) momentum sources due to the
pressure and viscous drag forces induced by the
vegetation and the building vertical surfaces
6
Heat equation forcing terms (modification of
vertical turbulent transport term) sensible
heat sources from surfaces anthropogenic heat
sources (Taha, 1999)
Humidity equation forcing terms (modification
of vertical turbulent transport term) humidity
sources from surfaces anthropogenic humidity
sources (not considered)
7
TKE equation forcing terms (modification of
vertical turbulent transport) shear production
by building horizontal surfaces buoyant
production from the surface sensible heat fluxes
wake production due to the presence of
vegetation and buildings dissipation due to
the accelerated cascade of TKE from large to
small scales due to the canopy elements
8
Turbulent Length Scale

• Parameterization of the turbulent length scale of
  Bougeault and Lacarrère (1989) inside the
  Gayno-Seaman PBL model.
• Addition of a turbulent length scale in the
  dissipation rate of TKE to consider the size of
  the wake eddies inside the canopy (following
  Martilli et al. (2002) for building canopy).

9
SM2-U(3D)

• SM2-U(3D) is a multi-layers rural and urban
  canopy model derived from the one-layer canopy
  model SM2-U.
• The model estimates the sensible and latent heat
  fluxes at each level within the canopy.

Dupont et al. 2003a, Parameterisation of the
Urban Water Budget by Using SM2-U model.
(Submitted to Journal of the Applied
Meteorology) Dupont et al. 2003b,
Parameterisation of the Urban Energy Budget with
the SM2-U model for the Urban Boundary Layer
Simulation. (Submitted to Boundary-Layer
Meteorology)
10
Philadelphia case (July 14th, 1995)
Dupont et al. 2003c, Simulation of
Meteorological Fields within and above Urban and
Rural Canopies with a Mesoscale Model
(MM5). (Submitted to Boundary-Layer Meteorology)

• DA-SM2-U is capable of simulating the important
  features observed in the urban and rural
  roughness sub-layer.
• Comparison with measurements showed that the
  surface air temperature simulation above rural
  and urban areas is improved with DA-SM2-U
  compared to the standard version of MM5.

11
Meteorological fields inside the canopy at 2 m
above the ground
12
Houston case

• August 25 September 1, 2000 (portion of the
  Texas 2000 Air Quality Study field program).
• MM5 has been run by Nielsen-Gammon in a one-way
  nested configuration 108-, 36-, 12-, and 4-km
  horizontal grid spacing.
• DA-SM2-U is used for a 1-km horizontal grid
  spacing domain (141 x 133 x 48).
• Canopy morphological parameters computed by Steve
  Burian

13
Morphological parameter domain
MM5 1-km domain
14
Land Use / Land Cover USGS level II (38
categories)
15
Airborne LIDAR dataset from TerraPoint LLC

• For the all Harris county (compressed data is 70
  GB, uncompressed 300 GB)
• Give the earth elevation and the elevation of the
  top of canopy elements.
• 1-m and 5-m horizontal grid spacing,
• Horizontal accuracy of 15 to 20 cm RMSE, Vertical
  accuracy of 5 to 10 cm RMSE

16
Example of Airborne LIDAR
17
Land Use / Land Cover
Airborne LIDAR data

High-resolution aerial photos (Harris A)
Building footprint dataset (Harris A)
ArcView map calculator
Morphological parameters for Harris A (1-km2
horizontal resolution and 1-m vertical resolution)
Correlation between the morphological parameter
values and the Land Use
Morphological parameters for the all
computational domain
18

• Mean building and vegetation height
• Building plan area density
• Vegetation plan area density
• Building rooftop area density
• Vegetation top area density
• Building frontal area density for 4 wind
  directions
• Vegetation frontal area density
• Wall-to-plan area ratio
• Building height-to-width ratio
• Surface fraction of vegetation, roads, rooftops,
  and water
• Sky view factor at ground level and as a function
  of height
• Aerodynamic roughness length and displacement
  height (Raupach, Macdonald, Bottema)
• Mean orientation of streets
• Approximations for impervious area, directly
  connected impervious area, and building material

For DA-SM2-U
19
Influence of the city representation
Detailed city specific morphological parameters
are deduced for each grid cell of Harris A,
outside they are deduced following the Land Use
from their correlations in Harris A. Average
city morphological parameters are deduced
following the Land Use for the entire domain.
20
Detailed city
Roof fraction
Average city
21
Detailed city
Height-to-width ratio
Average city
22
Detailed city
Average city
23
Detailed city
Average city
4 p.m.
Detailed city
Average city
12 a.m.
24
Detailed city
Average city
TKE
Detailed city
Average city
PBL height
25
Conclusions

• A neighborhood scale version of MM5 (DA-SM2-U)
  has been developed and tested successfully on
  Philadelphia.
• A huge morphological database has been
  constructed on Houston for DA-SM2-U
• The choice of the representation of the city of
  Houston (detailed or average city) seems to have
  an impact on the UBL structure, especially during
  unstable conditions.
• This study needs to be continued with different
  average representations of the city.

26
Future plans

• Comparison of simulated and observed surface
  meteorological fields (25 surface observation
  stations).
• The first results seem to indicate an improvement
  of the wind speed at 10 m by comparison to the
  results of the standard version of MM5.
  However, the see breeze seems to be too weak
  toward the city.

• CMAQ simulation on Houston by using
  meteorological fields from MM5-DA-SM2-U