MEGAPOLI Scientific Report 10-04 Hierarchy of Urban Canopy

MEGAPOLI Scientific Report 10-04

Hierarchy of Urban Canopy Parameterisations for Different Scale Models MEGAPOLI Deliverable 2.2 Alexander Mahura, Alexander Baklanov (Eds.) Contributing Authors Baklanov A., Martilli A., Grimmond CSB., Mahura A., Ching J., Calmet I., Clark P., Esau I., Dandou A., Zilitinkevich S., Best M., Mestayer P., Santiago J.L., Tombrou M., Petersen C., Porson A., Salamanca F., Amstrup B.

PeiraiasNOA

Marousi

Spata

Penteli

Elliniko

Zografou

Copenhagen, 2010

FP7 EC MEGAPOLI Project

www.megapoli.info 2

Colophon Serial title: MEGAPOLI Project Scientific Report 10-04

Title: Hierarchy of Urban Canopy Parameterisations for Different Scale Models Subtitle: MEGAPOLI Deliverable D2.2

Editor(s): Alexander Mahura, Alexander Baklanov Contributing Author(s): Isabelle Calmet, Patrice Mestayer Laboratoire de Mécanique des Fluides, Ecole Centrale de Nantes (ECN), France Aggeliki Dandou, Maria Tombrou Department of Environmental Physics & Meteorology, National & Kapodistrian University of Athens, Greece CSB Grimmond King’s College London (KCL), Strand, London, UK Alberto Martilli, Santiago J. L., Salamanca F. Environ. Department, Atmos. Pollution Modelling Division, Ministry of Science & Innovation, Madrid, Spain Igor Esau Nansen Environmental and Remote Sensing Center and Bjerknes Centre for Climate Research, Oslo, Norway Sergej Zilitinkevich Division of Atmospheric Sciences, University of Helsinki, Finland Jason Ching Atmospheric Modeling and Analysis Division, US Environmental Protection Agency (EPA), NC, USA Aurore Porson Department of Meteorology, University of Reading, Reading, UK Peter Clark Met Office, Joint Centre for Mesoscale Meteorology (JCMM), Reading, UK Martin Best Met Office, Hadley Centre for Climate Prediction and Research, Wallingford, UK Alexander Baklanov, Alexander Mahura, Claus Petersen, Bjarne Amstrup Danish Meteorological Institute (DMI), Copenhagen, Denmark

Responsible institution(s): Research Department, Danish Meteorological Institute, DMI Lyngbyvej 100, Copenhagen, DK-2100, Denmark Contact e-mail: [email protected] Language: English Keywords: urban meteorology and air pollution, urban sub-layer, roughness and land-use classification, street-, city-, regional- and global-scales modelling, parameterizations of urban canopy, surface energy balance, urbaniza-tion of different types of models Url: http://megapoli.dmi.dk/publ/MEGAPOLI_sr10-04.pdf Digital ISBN: 978-87-992924-7-9

MEGAPOLI: MEGAPOLI-07-REP-2010-03 Website: www.megapoli.info Copyright: FP7 EC MEGAPOLI Project

Part I – also contribution to WG1 of the COST Action 728 Part II – also contribution to WG2 of the COST Action 728


www.megapoli.info 3

Content: Background on Deliverable .................................................................................................................4 PART I - Urbanization of Meteorological Models ..........................................................................5 Abstract ................................................................................................................................................5 1 Introduction.......................................................................................................................................6 2. Methodologies for Urbanization of Meteorological Models ...........................................................7

2.1 Increased grid resolution and nesting of models........................................................................8 2.2 Urban land-use classification and algorithms for roughness parameters...................................9 2.3 Urban fluxes and sublayer parameterisation............................................................................10 2.4 Approach based on improved urban roughness and fluxes......................................................11 2.5 Effect of urban canopy roughness............................................................................................12 2.6 Effective roughness over inhomogeneous terrain....................................................................13 2.7 Surface energy budget in urban areas ......................................................................................13

3 Testing with Different Urbanizations .............................................................................................16 3.1 Simple modification of land surface schemes..........................................................................16 3.2 Medium-Range Forecast Urban Scheme (MRF-Urban)..........................................................17 3.3 Building Effect Parameterization (BEP)..................................................................................18 3.4 Soil Model for Sub-Meso scales Urbanised version (SM2-U) ................................................19 3.5 UM Surface Exchange Scheme (MOSES)...............................................................................20 3.6 Urbanized Large-Eddy Simulation Model PALM...................................................................22 3.7 Evolution Lines of Urban Canopy Parameterizations..............................................................24

References..........................................................................................................................................26 PART II - Model Urbanization Strategy .......................................................................................33 Abstract ..............................................................................................................................................33 1. Introduction....................................................................................................................................34 2. Model Urbanization Strategy: Summaries, Recommendations and Requirements .......................35

2.1 “Fitness-for-purpose” guidance ...............................................................................................35 2.2 Strategy to urbanize different types of models ........................................................................37 2.3 Overview of major applications...............................................................................................38

2.3.1 Numerical weather prediction and meso-meteorological models.....................................38 2.3.2 Urban air pollution and emergency response models .......................................................38 2.3.3 Multiscale atmospheric environment modeling................................................................39 2.3.4 Urban pollution and climate integrated modeling.............................................................40

2.4 Database and evaluation aspects of urbanized models ............................................................40 2.4.1 Database requirements ......................................................................................................41 2.4.2 Evaluation .........................................................................................................................41

2.5 Potential community activities.................................................................................................42 References..........................................................................................................................................44 Acknowledgements............................................................................................................................47 Previous MEGAPOLI reports............................................................................................................48


4

Background on Deliverable The MEGAPOLI WP2 “Megacity Environments: Features, Processes and Effects” is focusing on the megacity features (e.g. morphology), along with processes taking place in the urban canopy and boundary layer, which are responsible for the airborne transport and transformation of pollutants and urban climate effects. This WP is also aimed at the testing of existing and developing of sub-grid parameterisations of urban layer processes for megacity, regional and global scale models. Challenging sub-grid features in the WP tasks include: spatial and temporal distribution of emission source activities; flow modification by the urban canopy structure; flow modification by the urban surface heat balance; enhancement/damping of turbulent fluxes in the urban boundary layer due to surface and emission heterogeneity; chemical modification of pollutants in the dispersion process. Because megacities are localized, heterogeneous and variable sources of the anthropogenic impact on air quality and ultimately on climate, the major difficulty in megacity forcing in simulations arises from the sub-grid scale features. They are typically unresolved in climate models and barely resolved in regional scale models. Thus, models rely on parameterizations of megacity features aggregated within the model grid cell. Aggregation is not straightforward given surface heterogene-ity and strong non-linearity of the turbulent transport in the urban atmospheric boundary layer (UABL). The latter prohibits the application of direct averaging to obtain the large-scale forcing. The aggregation problems are still largely ignored in existing urban parameterizations. A more sophisticated approach which accounts for emission at different levels and for the surface thermal and drag heterogeneity is needed. Recent progress in street- and urban-scale turbulence-resolving simulations has opened the way for the development of a new generation of effective urban parame-terizations. The models require databases of emissions and surface characteristics as initial and boundary conditions. Feature analysis helps assessment of the megacity climate. It also relaxes the stability constraints on the megacity forcing in large-scale models. This deliverable (Del 2.2 - Hierarchy of Urban Canopy Parameterisations for Different Scale Models) is connected with the studying of the flow deformation by urban canopy in the urban sub-layer in order to test existing and develop sub-grid parameterisations of urban layer processes for megacity, regional and global scale models. Through systematic study of small-scale features of urban canopy effects on air flow the parameterizations of flow deformation and inter-canopy transport processes will be analysed and improved. A variety of single and multi-layer canopy approaches will be used. Vegetation will be considered included as roughness elements with more bluff bodies (buildings). The work is aggregating urban canopy properties to identify a hierarchy of approaches relevant to different urban and meteorological scales. See also WP2 relevant Deliverables:

• Grimmond CSB., M. Blackett, M.J. Best, et al. (2010): Urban Energy Balance Models Comparison. Deliverable D2.3, MEGAPOLI Scientific Report 10-07, MEGAPOLI-10-REP-2010-03, 72p, ISBN: 978-87-993898-0-3

• Esau I. (2010): Urbanized Turbulence-Resolving Model and Evaluation for Paris. Deliverable D2.4.1, MEGAPOLI Scientific Report 10-06, MEGAPOLI-09-REP-2010-03, 20p, ISBN: 978-87-992924-9-3

• Sievinen P., Hellsten A., Praks J., Koskinen K., J. Kukkonen (2010): Urban Morphological Database for Paris, France. Deliverable D2.1, MEGAPOLI Scientific Report 10-02, MEGAPOLI-05-REP-2010-03, 13p, ISBN: 978-87-992924-5-5


5

PART I - Urbanization of Meteorological Models

Baklanov A., Mahura A., Calmet I., Clark P., Esau I., Dandou A., Martilli A., Zilitinkevich S., Best M., Mestayer P., Santiago J.L., Tombrou M., Petersen C., Porson A., Salamanca F., Amstrup B.

Abstract The increased resolution of meteorological models, including numerical weather prediction models,

allows nowadays addressing more specifically urban meteorology forecasts and hence, improving

air pollution modeling. This has triggered new interest in modelling and describing experimentally

the specific features and processes of urban areas. Outcomes of developments and results are shown

here. Several approaches and modules for the meteorological models urbanization, including the

effective roughness and flux modifications, source and sink terms in the momentum, energy and

turbulent kinetic energy equations due building effects, urban soil model, etc. are considered in

different meteorological models and compared. Issues of optimum resolution, parameterising urban

roughness sublayer and surface exchange fluxes and the role of the urban soil layers are addressed

with different meteorological models.


6

1 Introduction During the last decade, substantial progresses in both meso-meteorological and numerical weather prediction (NWP) modelling and the description of urban atmospheric processes have been achieved. For instance, state-of-the-art nested NWP models can use land-use databases down to 1 km resolution and finer, enabling to provide high quality urban meteorological data. Thus, models are now approaching the necessary horizontal and vertical resolution to provide weather forecasts for the urban scale. Note that many urban features can influence the atmospheric flow, its turbulence regime, the micro-climate, and, accordingly modify the transport, dispersion, and deposition of atmospheric pollutants within urban areas, namely:

• Local-scale non-homogeneities, such as sharp changes of roughness and heat fluxes; • Sheltering effects of buildings on the wind-velocity; • Redistribution of eddies, from large to small, due to buildings; • Trapping of radiation in street canyons; • Effect of urban soil structure, • Different diffusivities of heat and water vapour in the canopy layer; • Anthropogenic heat fluxes, including the so-called urban heat island; • Urban internal boundary layers and the urban mixing height; • Effects of pollutants (including aerosols) on urban meteorology and climate; • Urban effects on clouds and precipitation. • Despite the increased resolution and various improvements, current meteorological models

still have several shortcomings with respect to urban areas, such as: • Urban areas are mostly described by similar sub-surface, surface, and boundary layer formu-

lations as for rural areas. • These formulations do not account for specific urban dynamics and energetics or for their

impacts on the simulation of the atmospheric urban boundary layer (UBL) and its intrinsic characteristics (e.g. internal boundary layers, urban heat islands, precipitation patterns).

Nevertheless, in recent years, a number of parameterisation schemes have been developed to esti-mate the components of the surface energy balance (net radiation, sensible heat flux) and other UBL parameters. For instance, COST-715 Action (Fisher et al. 2005ab; Piringer & Joffre, 2005) re-viewed several approaches for specific treatment of UBL features and surface energy budget (SEB). A palette of urban SEB schemes and models are now available (e.g., Oke et al., 1999; Grimmond and Oke, 1999a; Masson, 2000; Dupont, 2001; Martilli et al., 2002), but they have not all been validated to the same degree. They range from simple transformations of some key coefficients in exchange schemes developed for natural surfaces to detailed modules computing quasi-explicitly the radiative and turbulent energy exchanges of each built element category, e.g., the ground surface, walls and roofs, treated in group by type. Furthermore, even more detailed models and software are available to compute the thermo-radiative budgets of, or interactions with, elemental building surfaces. These tools may be used to analyze experimental data from validation campaigns, to run numerical experiments for urban areas, or to perform sensitivity analysis studies. Some of the SEB developments were derived from, e.g., the SOLENE (Groleau et al., 2003), POV RAY (Lagouarde et al., 2002), and DART (Gastellu-Etchegorry et al., 2004) studies. The development and validation of these SEB models brought to light and helped to quantify several specificities of the urban canopy energetics:

• Net radiation varies in time at the local scale with solar orientation and in space with district morphology, which is not much different from its rural counterpart on average;

• The diurnal cycle of the turbulent sensible heat flux is large but highly variable, strongly de-pendent on district structure, and often positive at night. In the dense city centres, this flux is limited by a strong aerodynamic resistance, favouring heat storage;


7

• A large heat storage in building materials, rather than in the ground, as a function of build-ing density and morphology;

• A low but highly variable latent heat flux; • A hysteresis in the diurnal cycles, with phase lags between the energy budget components

due to heat being diverted from the budget and provisionally stored in the building materials in the morning at the expenses of the sensible heat, while the stored heat is released in the evening and at night.

Therefore, to improve meteorological forecasts for urban areas and to provide the high-resolution meteorological fields needed, for example, by urban air quality (UAQ) models, it is important to implement specific urban surface layer and surface energy balance parameterizations into meteoro-logical models, or so to speak to ‘urbanise’ these models. The improvement of UBL formulations and parameterizations using urban physiographic data classifications in meteorological models together with the evaluation of the induced improved simulation of urban meteorology for NWP and UAQ forecasting was one of the main aims of the FP6 EC FUMAPEX project (Integrated Systems for Forecasting Urban Meteorology, Air Pollution and Population Exposure; Baklanov et al., 2005a). Before that the COST Action 715 deeply ana-lyzed and made recommendations on urban meteorology modeling and parameterizations applied to urban air pollution problems (Fisher et al., 2005; Piringer & Joffre, 2005). Several members of the currently finalized COST Action 728 had been actively working to improve and apply different UBL formulations and parameterisations for meso-meteorological and atmospheric pollution models. Here, reviews and examinations of the approaches and results are presented with respect to: (i) Finer spatial grid resolution and model downscaling; (ii) Detailed physiographic data and land-use classification; (iii) Calculation of effective urban roughness; and (iv) Estimation of urban heat fluxes; (v) Urban canopy and soil sub-models.

2. Methodologies for Urbanization of Meteorological Models Recently many multi-scale, and especially meso-scale, meteorological models in Europe have been going through changes to incorporate influence of urban features on the meteorological fields. Some of the models that have been adapted and incorporate different “urbanisation” schemes include the Local Model (LM), MM5, RAMS, Topographic Vorticity-Model (TVM), Finite Volume Model (FVM), HIRLAM, SUBMESO (derived from ARPS), UM, WRF, and others. The general strategy followed to improve performance of the meteorological models includes the following aspects for the urbanisation of relevant submodels or processes:

• Model down-scaling, including increasing vertical and horizontal resolution and nesting techniques (one- and two-way nesting);

• Modified high-resolution urban land-use classifications, parameterizations and algorithms for roughness parameters in urban areas based on the morphologic method;

• Specific parameterization of the urban fluxes in meso-scale models; • Modelling/parameterization of meteorological fields in the urban sublayer; • Calculation of the urban mixing height based on prognostic approaches.

Improved urban meteorological forecasts could provide information to city management regarding additional hazardous or stressing urban climate (e.g. urban runoff and flooding, icing and snow accumulation, high urban winds or gusts, heat or cold stress in growing cities and/or a warming climate). Moreover, the availability of reliable urban scale weather forecasts could be a relevant support for emergency management of fires, accidental toxic emissions, potential terrorist actions, etc.


8

2.1 Increased grid resolution and nesting of models

Increased computer power and the implementation of grid nesting techniques have allowed modern meteorological models to approach the resolution necessary for the urban and city scales. For example, the operational Danish modelling system (Sass et al., 2002; Unden et al., 2002) is running with HIRLAM model at several horizontal resolutions (ranging from 15, 9, 5 and 3 km) and vertical resolution of 40 levels (with a possibility to increase up to 60 levels). Moreover, for the Copenha-gen metropolitan area the model can be run even at finer resolution of 1.4 km (Mahura et al., 2005a). The German DWD Local Model LM (Doms & Schättler, 1999) is operated as a nest within the Global Model for Europe. LM has resolutions of 7 and 2.8 km for the Central and Western Europe. Both the horizontal and vertical resolution of the model can be increased to 1.1 km and 43 layers, respectively, as required by the 1-way self-nesting version of LM (Fay & Neunhaeuserer, 2005). In Norway, the non-hydrostatic MM5 model (Grell et al., 1994) is nested with the HIRLAM NWP model (Berge et al., 2002). The latter model is operated on a 10 km horizontal resolution for North-Western Europe. A domain with a finer resolution of 3 km is used for the Oslo metropolitan area in which MM5 is one-way nested with HIRLAM. A two-way nesting is used between the 3 and 1 km resolutions area covering Oslo. The MM5 output is input to the Air Quality Model of the Norwe-gian Institute for Air Research. The UM model at present uses the finest horizontal resolution of 1 km. The UM surface exchange scheme (MOSES II; Essery et al., 2001, 2003) uses a tiled approach to surface heterogeneity. UM output is used in the off-line NAME transport and dispersion model, and further work is needed to improve the internal representation of the canopy.

15

20

25

30

35

0 2 4 6 8 10 12 14 16 18 20 22 24

time (h) LT

Tem

pera

ture

(oC

)

measurementsMM5-urbanMM5

(a) (b)

Figure 2.1: (a) Diurnal variation of average wind velocity at 10 m between observational data and three Danish HIRLAM-S05, -U01, and -D05 model versions for 00 UTC forecasts during May 2005; and (b) Time series of air temperature at 2 m during 14 Sep 1994 at urban station in Athens, as calculated by the MM5 at

different resolutions (urbanized vs. non-urbanized) vs. measurements. Verification and sensitivity studies with numerical weather prediction models vs. measurement data have been reported for several episodes in different European cities: Helsinki, Oslo, Bologna, Valencia, Copenhagen (Neunhaeuserer et al., 2004; Fay et al., 2005; Baklanov et al., 2005b). Results for the verification of the LM model using different resolutions are discussed by Fay & Neunhaeuserer (2005) and for the Norwegian urban nested MM5-HIRLAM system by Berge et al. (2002). Verifications for high-resolution versions of the Danish HIRLAM modelling system were carried out for Copenhagen by Mahura et al., (2005a) as shown in Figure 2.1a: the better perform-ance/ skill of the HIRLAM model is observed at finer resolution runs compared with lower ones. Another example (Figure 2.1b) of the affect of the urban characteristics at finer scales on the air temperature at the Athens’s urban station during 14 Sep 1994, as calculated by the MM5 model at


9

different resolutions (urbanized at 2 km vs. non-urbanized at 6 km) (Dandou et al., 2005) and compared with measurements. The changes in the air temperature, proved to be favourable through the whole diurnal cycle, resulting in decreasing the temperature amplitude wave, which is in accor-dance with Oke (1978).

2.2 Urban land-use classification and algorithms for roughness parameters Surface characteristics such as albedo, thermal properties, roughness, or moisture availability significantly control the surface energy balance partitioning of any type surface. Urban landscapes show a much larger range and variability of surface characteristics compared with natural surfaces. However, many of the meteorological models still do not consider any urban class at all, or include only one urban class for all types of urban surfaces. In view of a wide range of urban surface types, it is not possible to single out one set of universal urban surface values, which would be valid for all types of urban neighbourhoods worldwide. Therefore, much more detailed surface information than in existing meteorological models is needed. Typical surface characteristics can be attributed to distinct categories of urban neighbourhoods. Such a classification can be performed based on land use maps or aerial photos. Digital land use classification (LUC) datasets can help to define different urban classes and are a source of increas-ing importance. Unfortunately, most LUCs are classified by functional aspects (residential, indus-trial) and not by surface morphometry or surface cover. Focussing on meteorological aspects, Ellefsen (1991) classified North American cities into 17 Urban Terrain Zones (UTZ) according to building contiguity, construction, and materials. Fehrenbach et al. (2001) have automated the classification of urban climatological neighbourhoods from satellite image analysis. However, no universal classification scheme exists. Historical development results in a huge variety of urban neighbourhood types worldwide. The more complete a description scheme is, the more it is re-stricted to a specific (historical) region, e.g., UTZs are difficult to apply to European cities because typical morphometry and building materials are different. There are also other datatsets such as CORINE (Europe), CEH (UK), USGS (US), and others. An appropriately chosen set of surface parameters can be related to specific physical processes. For example, it is not surprising that the area covered by vegetation drives the magnitude of the latent heat flux, or that morphometric parameters help to describe the roughness and turbulence characteristics over a particular urban surface. Therefore, the following three most important char-acteristics can be outlined (cf. Piringer & Joffre, 2005). (i) Urban cover: Two dimensional plan aspect ratios (“plan area fractions”) describe the 2D surface

fraction of a particular surface type per total plan area (as viewed from above), e.g. the plan area ratios of buildings, vegetation, impervious surfaces. It can also include dominant street directions in a grid cell.

(ii) Three dimensional structure: 3D morphometric parameters describe the configuration of urban buildings (it can include vegetation as well). The most important morphometric parameters to be used in urban meteorology models include: the mean building height, frontal aspect ratio, surface enlargement, normalized building volume, characteristic inter-element spacing, can-yon width, building breadth, etc. For many cities, authorities provide digital 3D building data sets, which are a powerful tool for the analysis of urban surface forms. Such high resolution data can provide detailed measures of 3D parameters, and additionally vertical profiles, e.g., of building volume density and sky view factors.

(iii) Urban materials: This information (e.g., construction materials of buildings roofs and walls) is of great importance especially for the estimation of radiative properties (e.g. surface albedo) and the determination of storage heat flux densities. Detailed analysis of aerial photos or field surveys can provide the necessary information.


10

Proceeding from the urban LUC, the calculation of the main aerodynamic characteristics of urban areas such as the roughness length and displacement height, can be performed based on the mor-phometric or morphologic methods. With morphometric methods, ranking these aerodynamic characteristics depends on the model intrinsic requirements for input data. Bottema & Mestayer (1998) and Grimmond & Oke (1999b) reviewed methods to deduce aerodynamic properties from a set of morphometric parameters. Tests of the models against individual datasets showed poor performances. The simplified model of Bottema (1997) gave relatively better results and additionally can be applied across the full range of building density parameters. Considering its relatively low input requirements, it is an efficient alternative. More simple models cannot be recommended, especially due to their limited range of applicability. As to the “recommended rule of thumb” of Grimmond & Oke (1999b), one should keep in mind that it does not include any building density dependency of roughness and, therefore, will overestimate roughness for low and high densities, and underestimate it at medium densities. With morphologic methods a more empirical and pragmatic approach can be considered, based on the visual observation of the physical structure of the urban canopy (e.g., from aerial photography). From survey of experimental data, Grimmond & Oke (1999b) offered a first-order evaluation of the roughness parameters of urban zones, separated into only 4 categories. These categories are associ-ated with 4 flow regimes: (1) Low height and density – isolated flow; (2) Medium height and density – wake interference flow; (3) Tall and high density – skimming flow; (4) High rise – chaotic or mixed flow. Ellefsen (1991) designed a scheme to identify 17 types of urban terrain zones that are defined by a written description and model photography. Furthermore, Grimmond & Oke (1999b) adapted this scheme to their proposed 4 urban roughness categories, offering physical description, matrix of typical photographs, and table of the most probable non-dimensional roughness parameters. For the above mentioned categories the classical Davenport classification of effective terrain roughness was revised by including explicitly the urban terrains (Davenport et al., 2000; Mestayer & Bottema, 2002).

2.3 Urban fluxes and sublayer parameterisation

Simulating urban canopy effects in meteorological models at different scales can be considered with the following main approaches. The 1st approach: modifying the existing non-urban approaches (e.g., the Monin-Obukhov similar-ity theory MOST) for urban areas by finding proper values for the effective roughness lengths, displacement height, and heat fluxes (adding the anthropogenic heat flux (AHF), heat storage capacity and albedo change). In this case, the lowest model level is close to the top of the urban canopy (displacement height), and a new analytical model is suggested for the urban roughness sublayer which is the critical region where pollutants are emitted and where people live (Zilitinke-vich & Baklanov, 2005). The 2nd approach, alternatively, source and sink terms are added in the momentum, energy and turbulent kinetic energy equation to represent the effects of buildings. Different parameterizations (Masson, 2000; Kusaka et al., 2001; Martilli et al., 2002) have been developed to estimate the radiation balance (shading and trapping effect of the buildings), the heat, the momentum and the turbulent fluxes inside the urban canopy, considering a simple geometry of buildings and streets (3 surface types: roof, wall and road). The 3rd approach, a quite different approach is taken by Gryning & Batchvarova (1999), originally developed for the aggregation of fluxes of momentum and heat over a mixed forest agricultural areas and later applied over the forested sub-arctic area (Batchvarova et al., 2001). It was applied for urban conditions in Batchvarova & Gryning (2001). It is considered that a typical urban area can be subdivided into a large number of relatively homogeneous parts (neighbourhoods) reflecting the


11

development of the town, such as the central inner, residential, recreation and industrial parts. The description of the ceaseless adjustments of the flow in the urban area is simplified by introducing the neighbourhoods that are complexly interacting with the flow and forming internal boundary layers. On the level of street canyons, the roughness sub-layer, the flow varies in space and time. At a level of 3-5 times the average building height (Batchvarova & Gryning, 2006) the flow is in equilibrium with the underlying surface, known as inertial sub-layer. Higher up the differences in meteorological fields introduced by surface characteristics of different neighbourhoods are blended and the boundary layer is forced by area aggregated features. Additionally to the above mentioned approaches, the flux aggregation technique of Hasager et al. (2003) can be used for non-homogeneous surfaces and needs to be tested for urban areas. In some cases the urban module architecture could be built independently of the type of meteoro-logical models to allow a simple implementation into different models. Note that it is not always possible to build it as a completely independent module; so, the urban modules need to be modified substantially in order to satisfy the main requirements and formats of the used meteorological models. There is also a freedom on how to implement such module: either to incorporate it inside the model code or to call the separate module from the meteorological model code. The algorithm to call the urban module by the model should also take into account constant urban characteristics and parame-ters during initialisation stage (i.e. called only once when the meteorological model is initialised for simulations). Then the urban module is called on every time step during the simulations when the cell contains, at least, a fraction of the urban class. The urban canopy modules can be also built as an interface/post-processor module separated from the meteorological model. In such case, the urban sub-layer model will be run separately (using previously simulated meteorological data as a fist approximation) and will improve the meteoro-logical fields in an area close to and inside the urban canopy with higher resolution. Although such a way is less promising, because it does not yield any improvement of the meteorological forecast in urban areas and cannot allow feedbacks.

2.4 Approach based on improved urban roughness and fluxes Approach based on improved urban roughness and fluxes is usually adopted to meet a set of the following two requirements. The 1st requirement is to be relatively cheap computationally and as close as possible to the parameterisations of the surface/ boundary layer in the meteorological model. The 2nd requirement is to split the surface layer over urban areas into two/ three sub-layers (Fisher et al., 2005b). Such a split distinguishes: (i) the roughness layer (including logarithmic layers), where MOST can be used with correction to the urban roughness, and (ii) the urban canopy layer, where MOST does not work and new analytical parameterisations for the wind and eddy profiles have to be considered. In such a module algorithms are required for calculating the following urban parameters for the meteorological model and steps for each model grid having urban features:

• Land-use classification, including, at least, one urban class and several urban sub-classes; • Displacement height for the urban (and forest) canopies; • Urban and effective roughness (and flux aggregation); • Stability-dependent urban roughness lengths for momentum; • Urban anthropogenic heat fluxes, • Urban storage heat fluxes by the Objective Hysteresis Model (OHM, Grimmond et al.,

1991) or specific roughness lengths for heat and moisture; • Albedo correction for urbanised surfaces; • Prognostic mixing height parameterisations; • Parameterisation of wind and eddy profiles within the canopy layer.


12

It is reasonable to use such approach for relatively cheap simulations and for models having low vertical resolution (i.e. the first vertical level is higher than 20 m), when other more complex modules of the urban sub-layer would not much affect results or would be too expensive for opera-tional forecasting applications.

2.5 Effect of urban canopy roughness

The classical MOST theory with a modified calculation of the urban roughness cannot give a satisfactory solution for the urbanisation of meteorological model. To avoid or minimise this problem, it is suggested to consider the MOST profiles in such models only above an elevated level of the order of the displacement height. Therefore, the roughness for urban areas is characterised by, at least, two parameters: the roughness length and the displacement height. Theoretical aspects of such an approach were discussed by Rotach (1994, 1999), Belcher & Coceal (2002), Belcher et al. (2003), Zilitinkevich et al. (2005b), Batchvarova & Gryning (2006) and the COST-715 Action (Fisher et al., 2005a). Roughness parameters for urban areas are calculated by the modified algorithm based on the mor-phological methods. The displacement height is calculated only for grid-cells tagged as urban class following Fisher et al. (2005a). The roughness length is calculated for each grid-cell in the follow-ing way: (i) constant values in each urban sub-class are tabulated for the urban class; (ii) effective roughness is calculated based on values and percentages of each land-use class and urban sub-classes in the cell; (iii) at each time step, the roughness value is recalculated due to effect of tem-perature stratification. In the general case of very inhomogeneous surfaces, such as urban areas, in order to include mutual effects of neighbouring cells it would be reasonable to simulate the effective roughness fields for grid-cells of a given city separately for different situations (e.g., for different seasons, wind direc-tions) and to build a kind of effective roughness maps library. Nevertheless, most of meso-meteorological models, consider the roughness length as a constant for each grid cell. Experimental data (Arya, 1975; Joffre, 1982; Wood & Mason, 1991) showed that it can depend on temperature stratification. This effect can be considerable especially for very rough surfaces, like the urban canopy. Therefore, the algorithm for recalculation of the effective roughness separately for stable or unstable stability, based on a new stability-dependent parameterisation of the urban roughness length for momentum, is suggested (Zilitinkevich et al., 2008). The interpolation formulae for the effective roughness length are the following: for neutral, moderately stable and very stable stratifi-cation regimes:

νν //1 0000

00

∗− ++

=uzCLzC

zzmmS

meffectivem ,

where ν0C ≈10, ≈= −10 10 κuS CC 50 are empirical constants; and for unstable stratification:

m

effectivem

zz

0

0 − = [ ]|)|/exp()1( 0100 LzCCC m−−− ,

where 0C and 1C are empirical constants. The theoretical background for these formulations, their verification versus experimental data and choice of the constants are discussed in details by Zilit-inkevich et al. (2005a). However, just modifying the current rural MOST approaches for urban areas with specific values for the effective roughness lengths and displacement height, still does not solve the main problem, i.e., how to describe the vertical structure of meteorological parameters inside the urban canopy? A simple heuristic model of Zilitinkevich & Baklanov (2005) for the vertical profiles of the momen-tum flux and the mean wind velocity within the urban canopy can be applied. It considers the vertical wind profile inside the canopy (below the displacement height) as an analytical function of the average building height, size and density, as well as of some meteorological parameters.


13

It is noteworthy that the suggested improvements based on the canopy profile model and displace-ment height do not require substantial modification of the meteorological model itself, because the first computational model level is usually above the canopy.

2.6 Effective roughness over inhomogeneous terrain Based on the concept of dividing the urban area into neighbourhoods, the large scale aerodynamic roughness, covering several neighbourhoods, can be derived by use of the drag coefficient. When dealing with inhomogeneous conditions the regional (effective or aggregated) roughness length for momentum can be defined as the parameter that gives the correct surface stress for the area as a whole when used in connection with a wind profile relationship

⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

⎟⎠⎞

⎜⎝⎛−⎟

⎟⎠

⎞⎜⎜⎝

⎛=

−effm

effectivem

eff

Lz

zzu

zU ψκ 0

* ln)( blz > ,

where κ is the von Karman constant, effu* is the effective friction velocity, effectivemz −0 is the effective roughness length for the inhomogeneous area that will yield the correct surface stress from the wind profile and remains to be determined from the surface characteristics, mψ is the wind profile stabil-ity correction function and effL is the Monin-Obukhov length incorporating the effective momentum and sensible heat fluxes (Wood & Mason, 1991). For a homogeneous area the drag coefficient DC is given as the following (Stull, 1988):

2

0

2

ln⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

⎟⎠⎞

⎜⎝⎛−⎟

⎟⎠

⎞⎜⎜⎝

⎛=

−effm

effective

D

Lz

zz

C

ψ

κ .

At the blending height the wind speed is nearly horizontally constant. Mason (1988) proposes that the drag coefficient at the blending height for the entire area should be estimated as the drag coeffi-cients for the various sub areas weighted in proportion to their fraction of the area. Therefore:

2

0

2

0

lnln−−

−∑

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎠⎞

⎜⎝⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛=

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎠⎞

⎜⎝⎛−⎟

⎟⎠

⎞⎜⎜⎝

⎛

ii

bmi

bieff

bm

effectivem

b

Ll

zl

fLl

zl

ψψ ,

where if is the fraction of the total area covered by the ith surface having the local momentum roughness length iz0 and stability iL The effective roughness length, effectivez −0 obtained from this equation represents the surface stress that originates from surface roughness and does not include the effect of major obstacles such as topography, isolated buildings, houses and forest edges. The relationship shows that the regional surface roughness is a function of stability, indeed is depends on the distribution of the stability and surface roughness of the surfaces that makes up the grid. A consequence is that the regional surface roughness is large as compared to the average of the local surface roughnesses as also found by Gryning et al. (2001). For practical purposes the stability dependence of effectivemz −0 is often neglected as Wood & Mason (1991) found it to be small, but whether this is permissible for the urban area remains to be investigated.

2.7 Surface energy budget in urban areas

In general, the Surface Energy Budget (SEB) in urban areas can be written in the following way

(Piringer & Joffre, 2005):


14

Q* = K↓ - K↑ + L↓ - L↑ = H + LE + QAs ,

where Q* [W/m2] - net all-wave radiation; K↓ - incoming shortwave radiation; K↑ = α0.K↓ - outgoing, reflected shortwave radiation where α0 - surface albedo; L↓ - incoming longwave radia-tion from the sky and surrounding environment ‘seen’ from the point; L↑ = ε0σT0

4 + (1-ε0); L↓ - outgoing longwave including both that emitted from the surface consistent with its emissivity ε0 and absolute surface temperature T0, and the reflected incoming longwave; H - turbulent sensible heat flux; LE - turbulent latent heat flux (L is the latent heat of vaporisation); QAs - specific urban an-thropogenic surface heat flux. Thus, the urban formulation differs from the non-urban one only by the QAs term. This formulation is suitable for detailed urban canopy models, when the surface is just millimetres above ground and the canopy layers are within the simulation domain. For meteorological models in which the surface may be high above the urban canopy (average roughness level or displacement height), the SEB can be rewritten in the following form:

Q* = K↓ - K↑ + L↓ - L↑ = H + LE + QA + ΔQS

where QA is the anthropogenic heat flux from sources within the urban canopy and ΔQS is an imbal-ance term, which includes the storage heat flux in the urban canopy elements, the ground and the air layer, extending from the surface to a level where the vertical heat exchange divergence is negligi-ble (i.e., the constant flux layer).

Correspondingly, in the model most of the terms of the above equation are simulated for urban grid cells as usual with corresponding urban characteristics, but we need to define and parameterise two new urban terms: QA and ΔQS as well as the albedo for urban areas.

Urban anthropogenic heat flux Following estimations of the average anthropogenic heat fluxes (AHFs) for cities in different climatic zones (Oke, 1978), reference values for a full urban area (100% of urban class; e.g., city centre or high building district) are in the range from 60 to 200 W/m2, depending on the city size. Information on the spatial distribution of AHFs over a city is not available from monitoring data and is difficult to obtain from measurements (e.g., Pigeon et al. (2005) showed for Toulouse that QA estimates are very uncertain and consequently can display negative values during summer months). Therefore, several methods are suggested for the urban AHF based on an assumed de-pendency on (e.g., proportionality to) other relevant urban characteristics, which are available in the models, e.g.:

1. Population density using maps with a high resolution in urban areas; 2. Nocturnal radiation emissions (brightness) over urban areas based on high resolution satellite

images; 3. Land-use classification as a percentage of urban subclasses (central part, urban, sub-urban,

industrial, etc.); 4. Emission inventory for specific pollutants typical of urban areas (e.g., NOx from traffic emis-

sions, etc.); 5. Monitoring or simulation fields of air pollution concentration for such specific pollutants (see

above #4). The first method for AHFs as a function of the population density distribution in urban areas is the one most frequently used. For the second method based on the nocturnal brightness of urban areas, it is suggested to use the simple dependence: QA = Iln QAmax, where Iln is the normalised light intensity (max value is 1), and QAmax is a scale (max) value of the AHFs for 100% of urban surface (from 50 W/m2 for small/medium cities and up to 200 W/m2 for large mega-cities in industrially developed countries). However, it is important to notice that the brightness of urban areas is different for industrial vs. developing countries, and hence the method should be corrected accordingly.


15

The third method using land-use classification as a percentage of urban classes was tested for the Copenhagen and Krakow study in COST-715 (see Piringer & Joffre, 2005). The anthropogenic part to the urban surface fluxes was approximated according to a coarse urban LUC: (i) 75 W/m2 for the city centre, (ii) 40 W/m2 for city periphery areas, and (iii) 20 W/m2 for other urban-suburban areas. The last two methods based on urban emissions or air pollution, can be easily used in atmospheric pollution forecasting models, because such information is usually available in the simulation. Urban storage heat fluxes Storage heat fluxes in the urban canopy are considered in our system by two different approaches. First, the heat storage capacity effect can be calculated using specific parameterisations for the temperature and moisture roughness lengths of urban areas. Most meteorological models consider for their surface layer profiles that the scalar roughness length, z0t, is equal to the roughness for momentum, z0m. However, for urban areas, they are generally very different (up to 2-3 orders of magnitude). Theoretical studies (Zilitinkevich, 1970; Brutsaert, 1975) suggest that the ratio z0t /z0m is a function of the roughness Reynolds number. Thus, the formulation of Brutsaert & Sugita (1996) for example can be suggested for urban areas. Including the modification of Joffre (1989) of Brutsaert assumption concerning the level under which the log-profile is not valid by using the Reichardt’s profile, the following formulation can be recommended for various bluff types of roughness over a wide range of the roughness Reynolds number (0.1 ≤ Re* ≤100):

[ ],)Re3.7exp(20 ½25.0*00 Scazz cmt κ−= for Re* > 0.15,

where ac is the inverse turbulent Schmidt number (= KH/KM for z0t or KE/KM for z0q) and Sc the Schmidt number (=ν/Dc, Dc is the molecular diffusivity of the particular property, i.e., heat, mois-ture but also gaseous compounds). The original Brutsaert’s formula had a coefficient 7.4 instead of 20 in the first term of the right-hand side and was valid for Re*>2 (rough case only). Equation for

aerodynamically smooth case (at Re*=0.15) is the following:

[ ]3/20 6.13exp*)/(30 Scauz ct κν −= for Re*<0.15.

However, this and other existing formulations are very uncertain, rarely verified and cannot con-sider all the mechanisms of the urban heat storage. Therefore, the heat storage in the urban fabrics/buildings, including hysteresis, can be most easily parameterised from the radiation and surface cover information using the empirical objective hysteresis model (OHM) of Grimmond et al. (1991):

ΔQS = i=1

n

∑ (λi α1i) Q* + i=1

n

∑ (λi α2i) ∂Q*/∂t + i=1

n

∑ (λi α3i) [W/m2]

where the λi are the plan fractions of each of the n surface types in the area of interest and the α1-3i are the corresponding empirical coefficients. These α coefficients have been deduced from a re-analysis of the Multi-city Urban Hydro-meteorological Database obtained from ten sites in seven North American cities Grimmond & Oke (1999a). Urban albedo effects Radiative properties (such as albedo and emissivity) of building and ground-covering materials are very different from those of natural grounds and vegetation, while the vertical structure of spaces between buildings provides shade and radiation trapping. In addition, they have not only horizontal but also vertical and/or slanted orientations, which strongly alter the radiative transfers and energy budget. The heat flux to or from the ground changes with surface material: concrete, tarmac, soil, etc. Anthropogenic energy use can be a noticeable fraction of the annual solar input and thus, influences the local air stability.


16

3 Testing with Different Urbanizations

3.1 Simple modification of land surface schemes It is known that the boundary layer in the urban areas has a complex structure due to multiple contributions of different parameters, including variability in roughness, albedo, and fluxes, etc. All these effects can be included to some extend into models. The simple urbanization of can include modifications of anthropogenic heat flux, roughness, and albedo for the land surface scheme. For example, the HIRLAM model uses the so-called the Interaction Soil Biosphere Atmosphere (ISBA) scheme originally based on Noilhan & Planton (1989). The changes of the ISBA scheme include modifications of the set of parameters in each grid cells of modelling domain where the urban class is presented. These modifications include the urban roughness, anthropogenic heat flux and albedo. The urban roughness can be changed up to a maximum of 2 m for grid cells, where the urban class will reach up to 100%. The anthropogenic heat flux (from 10 to 200 W/m2) can be modified simi-larly to the roughness. Albedo could also vary during the summer vs. winter period, i.e. when the snow is covering the surface).

(a) (b)

Figure 3.1: (a) Sensitivity tests to urban features with the HIRLAM high resolution model shown as the difference plots (runs without vs. with modifications) for the air temperature at 2 m

(a) with modification of AHF+R over the Copenhagen (Denmark) metropolitan area on 8 Aug 2004, 06 UTC; and (b) with modification of AHF+R+A on 29 Jan 2009, 00 UTC for the St.Petersburg (Russia)

metropolitan area.

Sensitivity tests and verification of this approach of urbanisation are useful to apply for model runs with high resolution. Results of sensitivity tests to urban features using the HIRLAM model with modifications (through roughness, anthropogenic heat flux and albedo) of the ISBA land surface scheme are shown on examples of studies for the Copenhagen (Denmark) metropolitan area (Ma-hura et al., 2009a) and St. Petersburg (Russia) metropolitan area (Gavrilova et al., 2009; 2010) (Figure 3.1ab). The difference fields are shown for the meteorological model runs (i.e. runs without vs. with modifications of the ISBA scheme). These studies showed that the inclusion of urban related parameters can improve the forecasted meteorological fields for urban areas, and results of several combined urban effects have underlined significant role of non-linear effects. It was found that, for St.Petersburg, during winter at low wind conditions the differences between the control vs. urbanized runs over the metropolitan area were for the wind velocity at 10 m - up to 2 m/s (with a maximum of 2.9 m/s at nighttime), and for the air temperature at 2 m - more than 1ºC (with a maximum of 2.7ºC at nighttime). For Copenhagen, during summer at low wind conditions the differences between the control vs. urbanized runs over the metropolitan area were for the wind velocity at 10 m – more than 1 m/s (with a maximum of 3 m/s at nighttime), for the air temperature at 2 m – more than 0.5ºC (with a maximum of 1.5ºC at nighttime), and for the relative humidity at 2 m – more than 4% (max up to 7, at midday). Moreover, long-term runs at high resolution with urbanized HIRLAM model showed a slight improvement for


17

the overall model performance, and this improvement is more considerable over the urbanized areas. The simple modification of the land surface schemes of the meteorological models with changing roughness, anthropogenic heat flux, and albedo associated with urban areas shows a possibility to improve the results of meso-, urban, and street-scale (probably, regional scale) models. Note, that the computational time is almost the same if compared with non-urbanised version of the model; and hence, this variant of the urban parameterisation is computationally very cheap.

3.2 Medium-Range Forecast Urban Scheme (MRF-Urban) The urbanization can be done through the dynamical and thermal parts of a meteorological model. For example, by using the non-local Medium-Range Forecast (MRF) Planetary Boundary Layer (PBL) parameterisation scheme (Troen & Mahrt, 1986) modified with the MRF-urban scheme (Dandou & Tombrou, 2009), whereby urban features are considered. Such approach was applied for the greater area of Athens (Greece) with the numerical simulations and sensitivity tests performed with the PSU/NCAR Mesoscale Model (MM5) The MM5 meteorological model modifications were carried out in two directions. At first, with respect to the thermal properties of an urban surface, the surface energy balance was modified by taking into account the anthropogenic heat released in urban areas and the urban heat storage term to account for urban/building mass effects, including hysteresis. In particular, the anthropogenic heat was calculated as a function of the diurnal spatial variation of the NO/CO emission inventories and the heat storage term was calculated by the objective hysteresis model OHM (Grimmond et al., 1991). At second, with respect to the dynamical properties of an urban environment, the surface stress and fluxes of heat and momentum were modified following recent advances in the atmos-pheric boundary layer over rough surfaces under unstable conditions (Akylas et al., 2001; Akylas et al., 2003; Akylas & Tombrou, 2005). The diffusion coefficients were also modified for stable conditions, according to King et al. (2001). The whole process was supplemented by detailed information on land use cover (30 m), derived from a Landsat 5 thematic mapper satellite image analysis, considering 7 urban land use categories. The detailed information was used with an aggre-gation method in combination with an area-weighting scheme, in order to construct new fields for the roughness length and the semi-empirical coefficients for the heat storage flux within the urban limits of the city.

PeiraiasNOA

Marousi

Spata

Penteli

Elliniko

Zografou

ΔΤ (oC)

-1.2 to -0.8 -0.8 to -0.4 -0.4 to -0.2 -0.2 to 0.0 0.0 to 0.2 0.2 to 0.4 0.4 to 0.6 0.6 to 0.8 0.8 to 1.0 1.0 to 1.2 1.2 to 1.4 1.4 to 1.6 1.6 to 1.8 1.8 to 2.0 2.0 to 2.2 2.2 to 2.4 2.4 to 2.6 2.6 to 3.0

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 2 4 6 8 10 12 14 16 18 20 22 24

fric

tion

velo

city

(m/s)

measurementsMRF-urbanMRF

(a) (b)

Figure 3.2: (a) Influence of the city of Athens (Greece) on formation of the air temperature field at 2 m on 14 Sep 1994, 03 LST shown as difference between the MRF (urban vs. non-urban) schemes; and

(b) Time series of sonic anemometer measurements of the friction velocity at the Athens’s urban station vs. results of the MRF urban/non-urban runs for the same date.

Dandou & Tombrou (2009) found that modifications, both in the dynamical and thermal parts of the model, seem to play an important role and improve the model’s results (Figure 3.2ab) as it was shown through comparison with sonic anemometer measurements of turbulence and routine


18

meteorological data. In addition, a delay in the sea breeze front was found, and a reasonable frictional retard concerning its penetration, as well as an inland displacement of the heat island, as the air moved over the Athens metropolitan area. In particular, the MRF-urban scheme calculated a decrease in the air temperature amplitude wave, and has a better agreement with the measurements. During the day, the decrease in air temperature, diffusion coefficients and sensible heat flux is mainly attributed to modifications in the thermal part and, in particular, to the heat storage flux, plus the increase in the roughness length. The modifications in the dynamical part are significant in the calculated decrease of the friction velocity. The MRF-urban scheme calculated a slowing of sea breeze front and reasonable frictional retard concerning its penetration during the day. The MRF-Urban parameterization shows improvement of the meteorological model performance. It for the meso- and urban scale (and even street-scale) models through consideration of urban areas effects. Although, the parameterisation need to be further tested in other terrain, geographical and climatic areas.

3.3 Building Effect Parameterization (BEP) The aim of the urban sub-layer parameterisation (Martilli et al. 2002) is to simulate the effect of buildings on a meso-scale atmospheric flow. It takes into account the main characteristics of the urban environment: (i) vertical and horizontal surfaces (wall, canyon floor and roofs), (ii) shadow-ing and radiative trapping effects of the buildings, (iii) anthropogenic heat fluxes through the buildings wall and roof. In this parameterisation, the city is represented as a combination of several urban classes. Each class is characterised by an array of buildings of the same width located at the same distance from each other (canyon width), but with different heights (with a certain probability to have a building with height). To simplify the formulation it is assumed that the length of the street canyons is equal to the horizontal grid size. The vertical urban structure is defined on a numerical grid. The contributions of every urban surface type (canyon floor, roofs and walls) on the momentum, heat and turbulent kinetic energy equation are computed separately. First, the contributions of the horizontal surfaces (canyon floor and roofs) are calculated using the formulation of Louis (1979) based on the MOST. The roughness lengths used for this calculation are representative for the local roughness of the specific surface types (roofs or canyon floor) and not for the entire city. Second, the exchange of momentum and turbulent kinetic energy on the vertical surfaces (walls) are parameterised as the effect of pressure and drag forces induced by the buildings. The temperature fluxes from the walls are a function of the difference between the air temperature and the wall temperatures. They are parameterised using the formulation of Clarke (1985) proposed by Arnfield & Grimmond (1998) in their urban energy budget model. The energy budget is computed for every mentioned surface (canyon floor, roofs and walls). Initially, the direct and infrared radiations at the surfaces are calculated to take into account the shadowing and radiative trapping effects of the buildings. Then, the surface temperatures of roofs, walls and canyon floor are solved by heat diffu-sion equation in several layers in the material (concrete or asphalt). In order to use the BEP parameterisation in a meteorological model a number of input parameters have to be evaluated. These parameters characterise the urban environment and they can be classi-fied in three different groups. The first group is consisted of parameters characterising the buildings and streets geometry (street width and direction, building width and height). The second group includes parameters characterising the building and street materials (heat capacity and diffusivity, albedo and emissivity, street and roof roughness length). And the third group has parameters char-acterising the energy produced inside the buildings (indoor temperature). These different input parameters producing various effects on the momentum and the energy fluxes, they influence the wind and the temperature. For example, this parameterisation has been tested on the city of Basel (Switzerland) and verified vs. the BUBBLE experiment (Basel Urban Boundary Layer Experiment: Rotach et al., 2005). Results


19

of simulations obtained with BEP are very close to the measurements in the city centre (stations Ue1 and Up11) while it slightly overestimates the temperature at the city boundary as shown in Figure 3.3a. The urban effects were also estimated on example of the city of Copenhagen (Den-mark) for selected specific case studies as well as over a 2 month period of summer 2004 (Mahura et al., 2009ac) showing improved performance of the HIRLAM model over the metropolitan area (Figure 3.3b).

(a) (b)

Figure 3.3: (a) Basel metropolitan area: Air temperature at the ground level on 26 Jun 2002, 12 UTC recalculated with BEP module /black line indicates the city boundaries; squares show the measured tem-

perature at sites/, and (b) Copenhagen metropolitan area: Difference plots (between outputs of the control vs. urbanized (using

BEP module) runs for the wind velocity at 10 m on 1st Aug 2004, 06 UTC. The BEP parameterization shows a clear capacity to improve the results of meso- and urban scale (and even street-scale) models by taking into account the effects of urban areas. However, the parameterisation should be further tested using cities situated in different terrain, geographical regions, and climatic regimes as well as over longer periods.

3.4 Soil Model for Sub-Meso scales Urbanised version (SM2-U) The physical basis of the urban canopy model – the Soil Model for Sub-Meso scales Urbanized version (SM2-U; Dupont et al., 2006ab) was developed from the ISBA rural soil model of Noilhan & Planton (1989) with the inclusion of urban surfaces and the influence of buildings and sparse vegetation, while keeping the force-restore soil model approach. The objective was twofold: to simulate the urban micro-climatology, and to evaluate the heat and humidity fluxes at the urban canopy-atmosphere interface with sub-mesoscale atmospheric models. SM2-U has the advantage of a unique model for both rural and urban soils that allows simulating continuously all districts of an urbanised area. The physical processes inside the urban canopy, such as heat exchanges, heat storage, radiative trapping, water interception or surface water runoff, are integrated in a simple way. The only horizontal exchanges inside the urban canopy are radiation reflections and water runoff from saturated surfaces; the wind advection within the canopy layer is not considered. Under the surface, the sub-grid scale transfers are ensured in the two underlying continuous soil layers. While for a natural soil partly covered with vegetation ISBA computes the budgets for the whole ground-vegetation system, SM2-U separates in each computational cell eight surface types: the bare soil without vegetation, the soil located between vegetation elements, and the vegetation cover; the building roofs, the paved surfaces without vegetation, the vegetation elements over a paved surface (e.g., road side trees), paved surface under the vegetation; and water surfaces. SM2-U computes the water budget in three soil layers. See more details on parameterization in Dupont et al., 2006ab. For example, this parameterisation has been validated against the experimental data from the


20

HAPEX-MOBILHY and EFEDA campaigns. The urban hydrological components of the model have been validated against the experimental data obtained during 10 years at the Rezé suburban site in the Nantes (France) urban area (Berthier et al., 2001), with tests both on an annual scale and for stormy events (Dupont et al., 2006b). Finally, the energy budget urban parameterisations have been validated for a densely built city centre against the measurements of Grimmond et al. (2004) at the Marseilles (France) central site during the campaign UBL-ESCOMPTE (Mestayer et al., 2005), in a forced mode without soil-atmosphere feedback (Dupont & Mestayer, 2004), showing on average a very good agreement. This module has been also applied for the Copenhagen (Denmark) metropolitan area (Mahura et al., 2006; 2008b). For example, due to the SM2-U parameterization, the city of Marseille (France) influence on formation of the low level fields of potential temperature and wind during summer month is shown in Figure 3.4a; and the city of Copenhagen (Denmark) influence of the ground surface temperature for typical meteorological conditions during winter month is shown in Figure 3.4b (Calmet & Mestayer, 2009).

845000 850000 855000

X

107500

110000

112500

115000

117500

120000

122500

Y

TP303302301300299298297296295294

5 m/s

845000 850000 855000

X

107500

110000

112500

115000

117500

120000

122500

Y

TP303302301300299298297296295294

5 m/s

845000 850000 855000

X

107500

110000

112500

115000

117500

120000

122500

Y

TP303302301300299298297296295294

5 m/s

(a) (b)

Figure 3.4: (a) Influence on the city of Marseille (France) on formation of the low level fields of potential temperature (colours) and wind (vectors) on 25 Jun 2001, 13 UTC at 7.5 m above ground level; and

Influence of the city of Copenhagen (Denmark) on formation of the averaged (based on 7 types of surfaces) surface temperatures for the typical meteorological conditions observed in January at the morning hours.

The SM2-U parameterization shows a good possibility to improve the results of meso-, urban, and street-scale models by considering the thermal and water budgets within the urban areas. However, the parameterisation should be further tested. Note that for this module the CPU computational expenses are rather high (for example, for the SUBMESO model urbanised with SM2-U, one diurnal cycle run is equal approximately to 4.2 hrs on the NEC-SX6 supercomputer (on one proces-sor).

3.5 UM Surface Exchange Scheme (MOSES) The UM surface exchange scheme (MOSES II; Essery, et al. 2001; 2003) uses a tiled approach to surface heterogeneity. This assumes independent, 1D vertical fluxes from different surfaces. In practice, this means that patches of different surfaces are, at least, 100-200 m across; so the ap-proach should deal reasonably well with parkland, but not, necessarily, urban gardens etc. The current surface exchange scheme is designed to determine the effect of urban areas on the atmos-phere (and so on the evolution of flow), and not vice versa. The implication of this is that the details of the urban canopy are regarded as unimportant and only the surface-layer fluxes are computed. This obviously has limitations. At present, the finest resolution used in the UM is 1 km.


21

The tiled approach relies on the concept of a flux blending height, which should, strictly, be derived iteratively. An iterative solution has been implemented but, in practice, provided the first model layer is within the surface layer. The current tile scheme uses 9 tiles, only one of which is urban (see in Figure 3.5a the fraction of urban areas in the modelling domain of Greater London area and surroundings; Clark et al., 2009). The characteristics of the urban tile can vary from point to point in principle, though in practice, at coarse resolution, fixed urban characteristics tend to be used.

(a) (b)

Figure 3.5: (a) Fraction of urban class at 1 km resolution over SE England (centred on London) derived from CEH data; and (b) Impact on forecasts of ‘screen’ temperature (deg C) over 21 cases (during 3 Mar

2004 – 30 Jan 2005) f including anthropogenic heat source in the operational 4 km UM at a dense urban site (London Weather Centre) /red - without, blue - with the heat source; RMS error/.

The urban tile was originally implemented in simply through modification of tile properties such as albedo, roughness, drainage, canopy capacity, etc. This might be termed the ‘rough concrete’ approach, in that no explicit account is taken of the morphology of urban areas A significant impact was found by including a ‘thermal canopy’ in the surface energy budget which mimics, in a very simple way, the impact of phase lags introduced by storage within building materials. This is a small change structurally but has a significant beneficial impact (Best, 2005). This was first implemented within the ‘Site Specific Forecast Model’, a 1D form of the UM (Clark, 1998) and has been evaluated in the operational 12 km forecast model. It is now implemented in the operational 4 km model with the addition of an anthropogenic heat source. Even without this, it does a reasonable job of predicting urban heat islands (substantially better than the simple “rough concrete” approach). However, extensive testing against surface data from various cities has revealed limitations which are addressed below. The ‘rough-concrete’ approach, with or without canopy, does not properly account for the fact that various parts of the building environment have different surface energy budgets due to different radiation balance, turbulent exchange, materials, etc. Use of a single surface temperature brings problems. A detailed, multi-faceted approach such as that of Masson addresses this problem. Har-man et al. (2004ab) developed a similar scheme for a 2D street canyon system, and also demon-strated two important simplifications. The first is that, to a very good approximation, the walls and floor of the canyon can be assumed to have the same temperatures. This means that the canyon can be treated as one surface with very reasonable results. Secondly, measurements of exchange coeffi-cients with the various facets showed that the roof and canyon are not directly coupled. This means that the assumption is valid in spite of their proximity (though the same effective roughness for momentum must apply to each). The two approximation together lead to a ‘two-tile’ approach the tiles being roof and canyon. A simple version of this was evaluated against surface flux data from various cities (Best et al., 2006). The impact on forecasted temperature based on 21 cases (selected during 3 Mar 2004 – 30 Jan 2005 period) including anthropogenic heat source in the operational 4 km UM model at a dense urban site (London Weather Centre) is shown in Figure 3.5b.


22

This approach appears an appropriate level of sophistication for computing fluxes on the meso- and urban scales, but requires further diagnostic development to produce useful within-canopy profiles.

3.6 Urbanized Large-Eddy Simulation Model PALM Meteorological models require modifications of the surface layer parameterization to account for urban effects. Only the simplest modification introduced through

0z – surface roughness, 0d –

displacement height and bh – the height of the urban canopy (Britter & Hanna, 2003) is discussed

here. In a near-neutral flow, the velocity profiles )(zu at heights bhz > reads ( )]/)ln[( 00

1 zdzuu −= −∗κ (1)

where ∗u is the friction velocity, κ = 0.4 is the von Karman constant. Free parameters

0z and 0d

characterize the aerodynamic surface roughness on sub-grid scales and the aggregated displacement height of the grid cell. Those parameters must be fitted either to measured or the simulated in much finer resolution model profiles. The latter approach is the subject of this study. The Ekman boundary layer (EBL) over Paris morphology is simulated with the urbanized PALM large-eddy simulation code. Fine resolution model data are aggregated on scales 0.5 km, 1.0 km and 2.0 km and parameters’ dependence on the urban canopy height is investigated. The Urbanized City-Scale Turbulence-Resolving Model - the parallelized code PALM was urbanized as described in Letzel et al. (2008). The method introduces block-cubic surface where boundary conditions, including Monin-Obukhov surface layer, can be prescribed on each of 5 facets. The CRAY-XT4 computer of the Bergen University Parallab was used to compute the experiments over the entire central Paris domain as defined by the MEGAPOLI Paris morphology database. In this study, the run resolution was 50 m x 50 m x 25 m, which require ~1 hour x 32 CPU to complete the run. Figure 3.6a shows the composite surface elevation used for this study. The model was run for 11 hours over the rural surface (Paris without buildings), then the surface map was changed by urban one and the model run for another 3 hours. With the prescribed westerly geostrophic wind of 5 m/s, it means ~50 km of an average air particle path over urban surface, which correspond to the megapolis extension. The vertical cross-section of the instant wind speed in the urban simulations with PALM is given in Figure 3.6b. The wind profiles from the PALM run were aggregated (averaged) within 0.5 km (100 profiles), 1.0 km (400 profiles) and 2.0 km (1600 profiles) squares covering the experiment domain. Each averaged profile was used to fit theoretical profile (1) using the least square minimization in the 2D (

0z ,0d ) parameter space. The best fit maps at resolution 1 km are given for

0z in Figure 3.7a and for

0d in Figure 3.7b. The average 0z over the entire domain is 1.7 m and

0d = 47 m whereas bh =

60 m. By comparison, Garratt (1992) relationship bhd 3/20 = (2) predicts

0d = 40 m.

(a) (b) Figure 3.6: (a) Composite Paris urban surface elevation: the Digital Elevation Model and the building

height data on 10 m resolution were aggregated and combined to produce this surface /height - in meters/; and (b) Vertical cross-section of the instant wind speed of in the urban Paris simulations.

The obtained LES data are compared with empirical relationship (2) given in Garratt (1992) and often utilized in model urbanization. Figure 3.8a reveals that urban data do not support (2) as such. It is general shape of the large-scale surface morphology, obtained in “rural” simulations without


23

buildings, which fits (2) while with some non-negligible offset. Simulations with buildings on the top of the digital elevation model morphology actually deteriorate the agreement suggesting that the surface roughness becomes less dependent on

bh . This is also supported by increasing number of squares with no dependence on the building height whatsoever. Their relative fraction increases for small aggregation scales. Generally, the proportionality coefficient of unity seems to be satisfactory for both datasets and all scales of aggregation. One should recognize, the building height itself (about 50 m) is only a small fraction of the surface height differences (about 180 m), it is obvious from Figure 3.6a.

(a) (b)

Figure 3.7: The best fit (a) surface roughness and (b) displacement height in the log-law (1).

(a) (b)

Figure 3.8: (a) Dependence between the displacement height and the averaged surface height within the aggregation area. Small dots – data aggregated over 1 km scale; large dots – data aggregated over 2 km

scale. Gray dots – rural surafce experiment (without buildings); green dots – urban surafce experiment. Red line – Garratt empirical relation (2); black and green lines – correspondingly the best LSM fits for rural and

urban data; dashed line – 1:1 fit; and (b) Anomalies of the wind speed at 3 model levels in the high resolution PALM run on 10 m mesh. The grid

points under the surface are white areas.

Preliminary conclusions show that the available computer resources allow for simulations of the central Paris area with resolution of 10 m and ever finer. The result aggregation on different scales can provide the optimal values (

0z ,0d ) for urbanization of the coarse resolution models. Although

the values of 0d linearly correspond to the surface elevation model, the obtained proportionality

coefficient (~1) is different from that one (2/3) suggested earlier. Moreover, significant fraction of the area does not show any dependence between the urban canopy height and the displacement height. Preliminary we can conclude that it is the large-scale elevation variations, but not the urban building height variations, that have the largest effect on the urban surface layer at least in the central Paris area. The roughness length does have expected dependence )( 010 dhz b −= γ ,

1γ = 0.2-0.4 (Garratt, 1992) (not shown) albeit with somewhat large scatter.

0z does not show obvious dependence on the surface morphology but it looks to have some dependence on surface steepness


24

on large-scales. The future work will progress in two general directions: (A) completing and analysis of the fine resolution PALM experiments; introduction of the urban heat flux and the atmospheric stability; and (B) urbanization of LESNIC code (Esau, 2004) using another, more flexible and accurate, method from Pourquie et al. (2009). Figure 3.8b gives a flavour of fine resolution PALM run (10 m mesh), which has been now completed for rural type surface using 5000 CPU hours.

3.7 Evolution Lines of Urban Canopy Parameterizations

The behaviour of the atmospheric Urban Canopy Layer (UCL) is the result of the interactions between atmospheric structures induced by the urban heterogeneities (at streets scale, e. g. 1-10 meters), and larger structures that form in the urban Planetary Boundary Layer (PBL), and that respond to mesoscale forcing (10-100 km). To simulate these interactions one approach is mesoscale modelling with Urban Canopy Parameterization (UCP). The nature of the UCP depends on the purpose of the simulation, available computer power, and scientific knowledge. Since all these factors are continuously evolving, UCPs are also evolving. In this contribution, UCP evolution’s lines followed at CIEMAT are presented.

Idealized vs. Real Urban Morphologies - At the basis of most of the UCPs there is an idealization of the urban morphology needed, for example, for the calculation of radiation trapping and shadowing in the street canyons. It is assumed that the real morphology is equivalent to a 2D or 3D regular array of buildings, with uniform building materials (idealized morphology). The criteria adopted for the equivalence, is again dependent on the purpose of model use. Focussing on air quality and urban climate, we decided that urban morphologies are equivalent if they have the same vertical profiles of the spatially averaged mean variables, momentum and heat fluxes. We considered that a necessary condition for this equivalence is that roads, walls, and roofs surfaces (respectively) must be the same in the real and the idealized morphologies. Moreover, sky-view factors (an important element in the estimation of the surface energy budget) must also be the same in the two morphologies. Building material used in the simplified morphology must also respond to solar forcing in a similar way to the weighted average of the building materials of the real morphology. Based on these assumptions, a series of expressions can be derived to estimate the basic morphological parameters (Martilli, 2009), and thermal properties (Salamanca et al., 2009a) of the idealized morphologies based on values of real morphologies and building materials. This work is the natural complement of the effort done to obtain detailed databases of real urban morphologies (NUDAPT; Ching et al., 2009).

Microscale Models - Urban heterogeneities generate atmospheric structures (turbulent and not) at scales smaller than the typical grid size of a mesoscale model (e.g. 1 km). This has two consequences: a) point measurements in the UCL are not representative of spatially averaged variables, and cannot be directly used to validate mesoscale models with UCPs, and b) the small scale structures affect the spatially averaged variables, and their effect need to be parameterized. To shed light on this problem, we used a microscale CFD-RANS (Reynolds Averaged Navier Stokes) model that runs on domains smaller than the mesoscale (hundreds of meters), but with resolutions high enough to explicitly resolve the building induced non-turbulent structures. RANS was selected as the best compromise between accuracy and CPU time, after a validation against wind tunnel and Direct Numerical Simulations (DNS) data (Santiago et al., 2007; Martilli & Santiago, 2007; Santiago et al., 2008). The advantage of microscale CFD is that spatial averages can be performed and compared to the outputs of the UCPs. In this way it was possible, for example, to estimate the dependency of the drag coefficient and the turbulent length scales with the packing density. Once the values of the parameters derived with the CFD models were introduced in a UCP based on BEP (Building Effect Parameterization, Martilli et al. 2002), and run in 1-D, vertical profiles of the mean wind were in good agreement with the spatially averaged profiles (Figure 3.9a). This technique can also be used to find the most appropriate idealized urban morphology that represents


25

a real one.

Building Energy Models - following the same philosophy of idealize the urban structure, a simplified Building Energy Model (BEM) has been built to estimate the energy exchanges between the buildings’ interior and the atmosphere. BEM accounts for heat generation inside the buildings due to occupants and equipments, solar radiation through windows, heat diffusion through walls and air conditioning systems (Salamanca et al., 2009b) and it allows estimating the energy consumption due to air conditioning. BEM has been linked to BEP, and tested against the BUBBLE dataset (Basel Urban Boundary Layer Experiment), showing an improvement compared to the previous version (Salamanca & Martilli, 2009). BEP+BEM have been implemented in the Weather and Research Forecast model (WRF) and run over Houston (Figure 3.9b). The energy consumption estimated for two days in August 2000 is comparable with values obtained using bottom-up and top-down approaches (Salamanca et al., 2010).

(a) (b)

Figure 3.9: Vertical profiles of the horizontal mean wind computed with the parameterization (solid lines), and spatially averaged from the CFD runs (squares), for packing densities of 0.0625 (black), 0.25 (red), and 0.44 (green); and (b) Sensible heat flux due to the air conditioning systems modelled with WRF+BEP+BEM

at 1500LST of the August 31st, over Houston. Maximum is 40Wm.

This work is a step toward a tool to evaluate the impact of a change on urban structure on air quality, urban climate, and meteorologically related (air conditioning, heating) energy consumptions. This tool can be helpful in evaluate urban development scenarios.


26

References Akylas E. and M. Tombrou (2005): Reconsidering and generalized interpolation between Kansas-

type formulae and free convection forms, Boundary-Layer Meteorol., 115, 381-398.

Akylas E., M. Tombrou, D. Lalas and S. Zilitinkevich (2001): Surface fluxes under shear-free convection, Q. J. R. Meteorol. Soc., 121, 225-270, 2001.

Akylas E., Y. Tsakos, M Tombrou and D. P. Lalas (2003): Considerations on minimum friction velocity, Q. J. R. Meteorol. Soc., 129, 1929-1943.

Arnfield, A. J., Grimmond, C. S. B. (1998): An urban canyon energy budget model and its applica-tion to urban storage heat flux modelling’, Energy and Buildings, 27, 61-68.

Arya S.P.S. (1975): Buoyancy effects in a horizontal flat-plate boundary layer, J. Fluid Mechs, 68, 321-343.

Baklanov, A., Hänninen, O., Slørdal, L.H., Kukkonen, J., Sørensen, J.H., Bjergene, N., Fay, B., Finardi, S., Hoe, S.C., Jantunen, M., Karppinen, A., Rasmussen, A., Skouloudis, A., Sokhi, R.S. (2005b): Integrated systems for forecasting urban meteorology, air pollution and population exposure. Atmospheric Chemistry and Physics,

Baklanov, A., P. Mestayer, A. Clappier, S. Zilitinkevich, S. Joffre, A. Mahura, (2005a): On parame-terizations of urban atmosphere sublayer in meteorological models. Atmospheric Chemistry and Physics Discuss., 5, 12119-12176. (FUMAPEX WP4 paper)

Baklanov, A., S. Joffre, and S. Galmarini (Eds.) (2005): "Urban Meteorology and Atmospheric Pollution". Special Issue of Atmospheric Chemistry and Physics Journal. http://www.atmos-chem-phys.net/special_issue24.html

Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou (Eds) (2009): Urbanization of Meteoro-logical and Air Quality Models. Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Batchvarova, E. and Gryning, S.-E. (2006): ‘Progress in Urban Dispersion Studies’, Theor. Appl. Climatol. 84, 57-67.

Batchvarova, E., Gryning, S.-E., Hasager, C.B. (2001): ‘Regional fluxes of momentum and sensible heat over a sub-arctic landscape during late winter’, Boundary-Layer Meteorol. 99, 489-507.

Batchvarova, E.; Gryning, S.-E. (2001): Regional-scale fluxes and mixed-layer height - towards a meteorological pre-processor for inhomogeneous terrain. In: Proceedings. 7. International conference on harmonisation within atmospheric dispersion modelling for regulatory pur-poses, Belgirate (IT), 28-31 May 2001. (Joint Research Centre, Environment Institute, Ispra, 2001) p. 380-384

Belcher, S.E. and O. Coceal (2002): Scaling the urban boundary layer. In: Rotach M., Fisher B., Piringer M. (Eds.), COST Action 715 Workshop on Urban Boundary Layer Parameterisations (Zurich, 24-25 May 2001). Office for Official Publications of the European Communities, EUR 20355, pp. 7-16, 2002.

Belcher, S.E., Jerram, N., Hunt, J.C.R. (2003): Adjustment of the turbulent boundary layer to a canopy of roughness elements. Under consid. For publication in J.Fluid Mech., 30 pp., 2003.

Berge, E., Walker, S-E., Sorteberg, A., Lenkopane, M., Eastwood, S., Jablonska, H.I. and Køltzow, M.Ø. (2002): A real-time operational forecast model for meteorology and air quality during peak air pollution episodes in Oslo, Norway. Water, Air and Soil Pollution Focus 2, 745-757, 2002.


27

Berthier, E., S. Dupont, H. Andrieu, P.G. Mestayer (2001): Comparison of evapotranspiration in an urban area, as evaluated by both an atmospheric model and a hydrological model, Interna-tional Symposium on Environmental Hydraulics, Tempe, Arizona, 5-8 Dec. 2001, Proceedings CD, 6 pp., 2001.

Best, M.J., (2005): Representing urban areas within operational numerical weather prediction models. Boundary-Layer Meteorol., 114, 91-109

Best, M.J., Grimmond, C.S.B., Villani, Maria Gabrialla (2006): Boundary-Layer Meteorol., 118, 503-525

Bottema, M. and P.G. Mestayer (1998): Urban roughness mapping - validation techniques and some first results, J. Wind Engineering & Industrial Aerodynamics, 74-76, 163-173, 1998.

Bottema, M. (1997): Urban roughness modelling in relation to pollutant dispersion. Atmos. Env., 31, 3059-3075, 1997.

Britter, R.E., and Hanna, S.R., (2003): Flow and dispersion in urban areas, Annu. Rev. Fluid Mech., 35, 469–496.

Brutsaert, W., and Sugita, M. (1996): Sensible heat transfer parameterisation for surfaces with anisothermal dense vegetation. J. Atmos. Sci., 53: 209-216, 1996.

Brutsaert, W. (1975): The roughness length for water vapor, sensible heat, and other scalars. J. Atmos. Sci. 32, 2028-2031, 1975.

Calmet I., P. Mestayer (2009): Urban Soil-Canopy-Atmosphere Exchanges at Submesoscales: Learning from Model Development, Evaluation, and Coupling with LES. In Urbanization of Meteorological and Air Quality Models; Baklanov A., S. Grimmond, A. Mahura, M. Athanas-siadou (Eds), Springer Publishers, ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Ching J., M. Brown, S. Burian, F. Chen, R. Cionco, A. Hanna, T. Hultgren, T. McPherson, D. Sailor, H. Taha, D. Williams, 2009: National Urban Database and Access Portal Tool, NUDAPT, Bulletin American Meteor.Society. 90, 1157-1168.

Clark P., Best M., A. Porson (2009): Evolution of Urban Surface Exchange in the UK Met Office’s Unified Model. In Urbanization of Meteorological and Air Quality Models; Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou (Eds), Springer Publishers, ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Clark P., (1998): The Implementation And Performance Of A 1D Model Coupled To NWP Forcing for Low-Cost Site-Specific Forecasting. AMSD 2nd Urban Symposium, Albuquerque, NM, Nov 1998.

Clarke, J. A. (1985): Energy simulation in building design, Adam Hilger, Bristol, 1985.

COST-728, 2007: Overview of the urban sublayer/exchange parameterizations in meso-meteorological and air pollution models. Report of the WG1 COST728, available from: http://cost728.org

Dandou A., M. Tombrou (2009): Sensitivity Tests in the Dynamical and Thermal Part of the MRF-Urban PBL Scheme in the MM5 Model. In Urbanization of Meteorological and Air Quality Models; Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou (Eds), Springer Publishers, ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Dandou A., M. Tombrou., E. Akylas, N. Soulakellis and E. Bossioli E. (2005): Development and evaluation of an urban parameterization scheme in the Penn State/NCAR Mesoscale Model (MM5), Journal of Geophysical Research-Atmospheres, 110, D10102, doi: 10.1029/2004JD005192, 2005.


28

Davenport, A., S. Grimmond, T. Oke and J. Wieringa (2000): The revised Davenport roughness classification for cities and sheltered country, 3rd Symp. On the Urban Environment, 14-18 Aug. 2000, Davis, Ca, AMS Proceedings, 9-10, 2000.

Doms, G., Schättler, U. (1999): The nonhydrostatic limited-area model LM, part1: scientific docu-mentation. R&D Dep. of DWD, 169 p., 1999.

Dupont, S and Mestayer, P.G.: Evaluation of the urban soil model SM2-U on the city center of Marseille (France), Fifth Symposium on the Urban Environment, 23-27 Aoüt, Vancouver, BC, Canada, AMS Proceeding CD (9.14), 2004.

Dupont, S., P.G. Mestayer, (2006): Parameterisation of the Urban Energy Budget with the Sub-mesoscale Soil Model, J. Applied Meteorology and Climatology, 45, No 12, 1744-1765

Dupont, S., P.G. Mestayer, E. Guilloteau, E. Berthier and H. Andrieu, (2006): Parameterisation of the Urban Water Budget with the Submesoscale Soil model, J. Applied Meteorology and Cli-matology, 45, No. 4, 624–648

Dupont, S. (2001): Modélisation Dynamique et Thermodynamique de la Canopée Urbaine: Réalisa-tion du Modèle de Sols Urbains pour SUBMESO, Doctoral thesis, Université de Nantes, France, 2001.

Ellefsen, R. (1991): Mapping and measuring buildings in the canopy boundary layer in ten U.S. cities. Energ. Buildings. 15-16 (3-4): 1025-1049, 1991.

Esau, I., (2004): Simulation of Ekman boundary layers by large eddy model with dynamic mixed subfilter closure, Environ. Fluid Mech., 4, 273-303.

Essery, R. , Best, M. an Cox, P. (2001): MOSES 2.2 Technical Documentation, Hadley Centre Technical Report No. 30, Met Office Hadley Centre

Essery, R. L. H. , Best, M. J., Betts, R. A., Cox, P. M., (2003): Explicit Representation of Subgrid Heterogeneity in a GCM Land Surface Scheme, Journal of Hydrometeorology, 4, 530-543

Fay, B. and L. Neunhaeuserer (2005): Evaluation of high-resolution simulations with the Lo-kalmodell of the German Weather Service for urban air pollution episodes in Helsinki and Oslo. Atm. Chem&Phys.

Fay, B., L. Neunhäuserer, J. L. Palau, G. Pérez-Landa, J. J. Dieguez, V. Ødegaard,G. Bonafé, S. Jongen, A. Rasmussen, B. Amstrup, A. Baklanov, U. Damrath (2005): Evaluation and inter-comparison of operational mesoscale models for FUMAPEX target cities. FUMAPEX Report for D3.4, DWD Offenbach, Germany, June 2005, 110p., 2005.

Fehrenbach, U., Scherer D., and Parlow E. (2001): Automated classification of planning objectives for the consideration of climate and air quality in urban and regional planning for the exam-ple of the region of Basel/Switzerland. Atmos. Environ., 35(32), 5605-5615, 2001.

Fisher, B., J. Kukkonen, M. Piringer, M. Rotach, M. Schatzmann (2005): Meteorology applied to urban air pollution problems: Concepts from COST 715. Atm. Chem&Phys.

Fisher, B., S. Joffre, J. Kukkonen, M. Piringer, M. Rotach, and M. Schatzmann (Eds.) (2005a): COST-715 ‘Meteorology applied to urban air pollution problems’. Final Report. CEC Publi-cation EUR , 2005a.

Garratt, J.R., (1992): The Atmospheric Boundary Layer, Cambridge University Press, Cambridge, pp. 316.

Gastellu-Etchegorry, J.P., Martin E. and Gascon, F. (2004): DART: a 3-D model for simulating satellite images and surface radiation budget. Intern. J. Remote Sensing, 25(1), 73-96, 2004.

Gavrilova Yu., A. Mahura, S. Smyshlayev, A. Baklanov (2009): Thermal and Dynamical Urban


29

Effects of Saint-Petersburg Metropolitan Area – Winter Case Study. Abstracts of the Interna-tional Conference on Computational Information Technologies for Environmental Sciences: CITES-2009, 11-15 Jul 2009, Krasnoyarsk, Russia, pp. 63-64.

Gavrilova Yu., Mahura A., Smyshlaev S., Baklanov A. (2010): Urban Effects in Numerical weather prediction model at Saint-Petersburg Metropolitan Area for winter. Geophysical Research Abstracts, Vol. 12, EGU2010-314.

Grell, G., Dudhia, J., and Stauffer, D. (1994): A Description of the Fifth-Generation PENN STATE/NCAR Mesoscale Model (MM5), NCAR Technical Note, NCAR/TN-398+STR. Boulder, Colorado, National Center for Atmospheric Research. 138 pp., 1994.

Grimmond, C. S. B., J. A. Salmond, T. R. Oke, B. Offerle, and A. Lemonsu (2004): Flux and turbu-lence measurements at a densely built-up site in Marseille: Heat, mass (water and carbon di-oxide), and momentum J. Geophys Res., 109, D24101, 2004.

Grimmond, C.S.B. and Oke, T.R.: Aerodynamic properties of urban areas derived from analysis of surface form. J. Appl. Meteor., 38 (9), 1262-1292, 1999b.

Grimmond, C.S.B. and Oke, T.R.: Heat storage in urban areas: Local-scale observations and evaluation of a simple model. J. Appl. Meteor., 38, 922-940, 1999a.

Grimmond, C.S.B., Cleugh, H.A. and Oke, T.R.: An objective urban heat storage model and its comparison with other schemes. Atmos. Environ., 25B, 311-326, 1991.

Groleau, D., Fragnaud F., Rosant J.M.: Simulation of the radiative behaviour of an urban quarter of Marseille with the Solene model. Fifth International Conference on Urban Climate, Lodz, Pologne, 1-5 September 2003, 2003.

Gryning, S. E., Batchvarova, E. and De Bruin, H. A. R.: 2001, 'Energy Balance of Sparse Conifer-ous High-Latitude Forest under Winter Conditions', Boundary-Layer Meteorol., 99,465-488.

Gryning, S.-E. and Batchvarova, E.: 1999, 'Regional Heat Flux over the NOPEX Area Estimated from the Evolution of the Mixed Layer', Agric. For. Meteorol. 98-99, 159-167.

Hamdi, M. and G. Schayes (2005): Improving Martilli's urban boundary layer scheme: offline validation over different urban sites. Atm. Chem & Phys.

Harman, I. N., Barlow, J. F., Belcher, S. E., 2004: Scalar fluxes from urban street canyons, Part II: Model. Boundary-Layer Meteorol., 113, 387-410.

Harman, I. N., Best, M. J., Belcher, S. E., 2004, Radiative exchange in an urban street canyon. Boundary-Layer Meteorol., 110, 301-316.

Hasager, C. B. Nielsen, N. W., Boegh, E., Jensen, N. O., Christensen, J. H., Dellwik,E. and Soe-gaard, H.: Effective roughnesses calculated from satellite-derived land cover maps and hedge information and used in a weather forecasting model. Boundary-Layer Meteorology, 109: 227-254, 2003.

Högström, U.: Review of Some Basic Characteristics of the Atmosphere Surface Layer, Boundary-Layer Meteorol. 78, 215-246, 1996.

Joffre S.M.: Momentum and Heat Transfers in the Surface Layer over a Frozen Sea. Boundary Layer Meteorology, 24, 211-229, 1982.

King J. C., W. M. Connolley and S. H. Derbyshire: Sensitivity of modeled Antarctic climate to surface and boundary-layer flux parameterizations, Q. J. R. Meteorol. Soc., 127, 779-794, 2001

Kusaka, H., Kondo, H., Kikegawa, Y., and Kimura, F.: A Simple Single-Layer Urban Canopy Model for Atmospheric Models: Comparison with Multi-Layer and SLAB Models, Boundary-


30

Layer Meteorol. 101, 329-358, 2001.

Lagouarde, J-P., Moreau P., Solliec F., Groleau D., Irvine M., Bonnefond J-M. and Voogt J.: Etude expérimentale et modélisation des effets directionnels dans l'IRT en zone urbaine. Atelier de Modélisation de l'Atmosphère 2002, Météo France, Toulouse, 17-19 Dec. 2002. 111-114, 2002.

Letzel, M. O., Krane, M. and Raasch, S., 2008: High resolution urban large-eddy simulation studies from street canyon to neighbourhood scale, Atmos. Environ., 42, 8770-8784.

Louis, J. F.: A parametric model of vertical eddies fluxes in the atmosphere, Boundary-Layer Meteorol., 17, 187-202, 1979.

Mahura, A., K. Sattler, C. Petersen, B. Amstrup, A. Baklanov (2005a): DMI-HIRLAM Modelling with High Resolution Setup and Simulations for Areas of Denmark. DMI Technical Report 05-12.

Mahura A., Leroyer S., Mestayer P., Calmet I., Dupont S., Long N., Baklanov A., Petersen C., Sattler K., Nielsen N.W. (2006): Large Eddy Simulation of Urban Features for Copenhagen Metropolitan Area. Atm. Chem. & Phys.

Mahura A., Baklanov A., Hoe S., J.H. Sorensen, C. Petersen (2009): Changes in Meteorological and Atmospheric Transport and Deposition Patterns due to Influence of Metropolitan Areas. Ukrainian Hydrological Journal, Vol. 4, pp. 187-194.

Mahura A., Baklanov A., Petersen C., N. W. Nielsen, B. Amstrup (2008a): Verification and Case Studies for Urban effects in HIRLAM Numerical Weather Prediction, pp. 143-150. In “Ur-banization of Meteorological and Air Quality Models”, (Eds. Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou), ISBN: 978-3-642-00297-7.

Mahura A., Leroyer S., Baklanov A., Mestayer P., Korsholm U.S., Calmet I., (2008b): Temporal and Spatial Variability of Fluxes in Urbanized Areas, pp-219-232; In “Urban Climate and Bioclimate”, (Eds. Klysik K., Wibig J., Fortuniak K.), ISBN: 978-83-7525-243-9.

Mahura A., Petersen C., Baklanov A., Amstrup B., (2008c): Evaluation of Building Effect Parame-terization Module for Urbanized Numerical Weather Prediction Modelling, pp-371-380; In “Urban Climate and Bioclimate”, (Eds. Klysik K., Wibig J., Fortuniak K.), ISBN: 978-83-7525-243-9.

Martilli A., J.L. Santiago (2009): How to use Computational Fluid Dynamics Models for Urban Canopy Parameterizations. In Urbanization of Meteorological and Air Quality Models; Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou (Eds), Springer Publishers, ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Martilli, A. and J. L. Santiago, 2007: CFD simulation of airflow over a regular array of cubes. Part II: analysis of spatial average properties. Boundary-Layer Meteorology 122, 635-654.

Martilli, A., 2009: On the Derivation of Input Parameters for Urban Canopy Models from Urban Morphological Datasets. Bound-Layer Meteo., 130, 301-306.

Martilli, A., A. Clappier, M. W. Rotach, 2002: An urban surface exchange parameterisation for mesoscale models’, Bound-Layer Meteor., 104, 261-304.

Masson, V.: A Physically-Based Scheme for the Urban Energy Budget in Atmospheric Models, Boundary-Layer Meteorol. 98, 357-397, 2000.

Mestayer, P. and M. Bottema: Parameterisation for roughness parameters in urban areas. In: Rotach M., Fisher B., Piringer M. (Eds.), COST Action 715 Workshop on Urban Boundary Layer Parameterisations (Zurich, 24-25 May 2001). Office for Official Publications of the European Communities, EUR 20355, pp.51-61, 2002.


31

Mestayer, P. G., Durand, P., Augustin, P., Bastin, S., Bonnefond, J.-M., Bénech, B., Campistron, B., Coppalle, A., Delbarre, H., Dousset, B., Drobinski, P., Druilhet, A., Fréjafon, E., Grimmond, S., Groleau, D., Irvine, M., Kergomard, C., Kermadi, S., Lagouarde, J.-P., Lemonsu, A., Lo-hou, F., Long, N., Masson, V., Moppert, C., Noilhan, J., Offerle, B., Oke, T., Pigeon, G., Puy-grenier, V., Roberts, S., Rosant, J.-M., Saïd, F., Salmond, J., Talbaut, M., and Voogt, J.: The Urban Boundary Layer Field Experiment over Marseille. UBL/CLU-ESCOMPTE: Experi-mental Set-up and First Results, Boundary-Layer Meteorol. 114, 315-365, 2005.

Neunhäuserer, L., B. Fay, A. Baklanov, N. Bjergene, J. Kukkonen, V. Ødegaard J. L. Palau, G. Pérez Landa, M. Rantamäki, A. Rasmussen, I. Valkama: Evaluation and comparison of op-erational NWP and mesoscale meteorological models for forecasting urban air pollution epi-sodes – Helsinki case study. In: Suppan, P. (Ed.), Proceedings of the 9th International Con-ference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, 1–4 June 2004, Garmisch-Partenkirchen, Germany. Vol. 2. pp. 245–249, 2004.

Noilhan, J. and Planton, S.: A Simple Parametrization of Land Surface Processes for Meteorologi-cal Models, Mon. Wea. Rev. 117, 536-549, 1989.

Oke, T.R., Spronken-Smith, R., Jauregui, E. and Grimmond, C.S.B.: Recent energy balance obser-vations in Mexico City. Atmos. Environ., 33, 3919-3930, 1999.

Oke, T.R.: Boundary layer climates. London, Methuen & Co Ltd, J. Wiley & Sons, New York, 1978.

Pigeon, G., P. Durand, V. Masson: Quantification of anthropogenic heat releases over Toulouse (France) during the CAPITOUL field program. Short Papers of the 5th International Confer-ence on Urban Air Quality Valencia, Spain, 29-31 March 2005, 2005.

Piringer, M. and S. Joffre (Eds.): The urban surface energy budget and mixing height in European cities: Data, models and challenges for urban meteorology and air quality/ Baklanov, A., J. Burzynski, A. Christen, M. Deserti, K. De Ridder, S. Emeis, S. Joffre, A. Karppinen, P. Me-stayer, D. Middleton, M. Piringer and M. Tombrou. Final Report of WG2 COST Action 715. 194 pp. , 2005.

Pourquie, M., Breugem, W.P. and B.J. Boersma, 2009: Some issues related to the use of immersed boundary methods to represent square obstacles, Int. J. Multiscale Comp. Eng., 7, 509-522.

Rotach, M.W., Vogt R., Bernhofer C., Batchvarova E., Christen A., Clappier A., Feddersen B., Gryning S-E., Martucci G., Mayer H., Mitev V., Oke T.R, Parlow E., Richner H., Roth M., Roulet YA., Ruffieux D., Salmond J., Schatzmann M., and Voogt J.: BUBBLE – an Urban Boundary Layer Meteorology Project. Theoretical and Applied Climatology. In press, 2005.

Rotach, M.W.: Determination of the zero-plane displacement in an urban area, Boundary-Layer Metorol., 67, 187-193, 1994.

Rotach, M.W.: On the influence of the urban roughness sublayer on turbulence and dispersion, Atmospheric Environment, 33, 4001-4008, 1999.

Salamanca, F., Krayenhoff, E. S., Martilli, A., 2009, On the Derivation of Material Thermal Properties Representative of Heterogeneous Urban Neighborhoods, Journal of Applied Meteorology and Climatology, 48, 1725-1732.

Salamanca, F., Krpo, A., Martilli, A., Clappier, A., 2009, A new building energy model coupled with an Urban Canopy Parameterization for urban climate simulations – Part I. Formulation, verification and sensitivity analysis of the model. Theoretical and Applied Climatology. DOI 10.1007/s00704-009-0142-9

Salamanca, F., Martilli, A., 2009, A new building energy model coupled with an Urban Canopy Parameterization for urban climate simulations – Part II. Validation with one dimension off-line simulations. Theoretical and Applied Climatology. DOI 10.1007/s00704-009-0143-8.


32

Salamanca, F., Martilli, A., tewari, M., Chen, F. 2010, A study of the Urban Boundary Layer using different urban parameterizations and high-resolution land cover data sets with WRF (The case of Houston), manuscript in preparation.

Santiago, J. L., A. Martilli, and F. Martin, 2007: CFD simulation of airflow over a regular array of cubes. Part I: 3-D simulation of the flow and validation with wind tunnel measurements. Boundary-Layer Meteorology 122. 609-634.

Santiago, J. L., Coceal, O., Martilli, A., Belcher S., 2008, Variation of the sectional drag coefficient of a group of buildings with packing density, Boundary-Layer Meteorology. 128. 445-457

Sass B.H., N.W. Woetmann, J.U. Jorgensen, B. Amstrup, M. Kmit, K.S. Mogensen, 2002: The operational DMI-HIRLAM system 2002 - version. DMI Technical Report N 02-05, 60p.

Troen I. and L. Mahrt: A simple model for the atmospheric Boundary-Layer: Sensitivity to surface evaporation, Boundary-Layer Meteorol., 37, 129-148, 1986.

Unden, P., L. Rontu, H. Järvinen, P. Lynch, J. Calvo, G. Cats, J. Cuhart, K. Eerola, etc. 2002: HIRLAM–5 Scientific Documentation. December 2002, HIRLAM–5 Project Report, SMHI.

Wood N., Mason P.: The Influence of Static Stability on the Effective Roughness Lengths for Mo-mentum and Heat Transfer. Q.J.R. Meteorological Society, Vol 117, pp. 1025-1056, 1991.

Zilitinkevich S.S.: Dynamics of the Atmospheric Boundary Layer. Godrometizdat, Leningrad, USSR, (in Russian), 1970.

Zilitinkevich, S. S., Hunt, J. C. R., Grachev, A. A., Essau, I. N., Lalas, D. P., Akylas, E., Tombrou, M., Fairall, C. W., Fernando, H. J. S., Baklanov, A., Joffre, S. M.: The effect of large eddies on the convective heat/mass transfer over complex terrain: advanced theory and its validation against experimental and LES data. Croatian Meteorological Journal, 40, 20-26, 2005b.

Zilitinkevich, S., I. Mammarella, A. Baklanov, S. Joffre (2008): The effect of stratification on the surface resistance for very rough vegetated and urban surfaces. Boundary-Layer Meteorol., Vol. 129, pp. 179-190, 2008.

Zilitinkevich, S., Baklanov, A.: An analytical model of the mean-wind and the momentum flux profiles in the urban roughness layer. Ch. 3 in DMI Scientific Report #04-08, ISBN: 87-7478-510-9, pp. 42-46. (also to be submitted to Boundary-Layer Meteorol.), 2005.


33

PART II - Model Urbanization Strategy

Baklanov A., Ching J., Grimmond CSB., Martilli A., Mahura A.

Abstract The available urban canopy models, modules and parameterisations are very different in terms of

the sophistication of process descriptions, computing resources required and in the associated

difficulties in implementing in meteorological models. Many publications consider separate aspects

of urban features but none provide necessary algorithms and steps required. Summaries,

discussions and recommendations on the best practice and strategy for urbanisation of different

types of meteorological and air quality models are given.


34

1. Introduction Urban features essentially influence atmospheric flow and microclimate, strongly enhance atmospheric turbulence, and modify turbulent transport, dispersion, and deposition of atmospheric pollutants (e.g., Piringer et al., 2007). Increased resolution in numerical weather prediction (NWP) models allows for a more realistic reproduction of urban air flows and air pollution processes, however most of the operational models still do not consider, or consider very poorly, the urban effects. This has triggered new interest in model development and investigation of processes specific to urban areas. Recent developments performed as part of the European project FUMAPEX on integrated systems for forecasting urban meteorology and air pollution (Baklanov et al., 2002, 2005), the US EPA and NCAR communities for MM5 (Dupont et al., 2004; Bornstein et al., 2006; Taha et al., 2008), WRF models (Chen et al., 2006), and other relevant studies (see e.g. PBL, 2007) have shown many opportunities in the “urbanization” of weather forecasting and atmospheric pollution dispersion models. Atmospheric models for urban areas have different requirements (e.g. relative importance of the urban boundary layer (UBL) and urban surface sublayer (USL) structure) depending on: (i) the scale of the models (global, regional, city, local, micro, etc.); (ii) the functional type of the model, e.g.:

• Forecasting or assessment type of models, • Urban or regional climate models, • Research meso-meteorological models, • Numerical weather prediction models, • Atmospheric pollution models (city-scale), • Emergency preparedness models, • Meteo-preprocessors (or post-processors).

A wide range of approaches have been taken to incorporating urban characteristics. In addition there are a wide range of processes which includes: characteristics of the urban canopy sublayer, components of urban surface energy balance (net radiation, sensible and latent heat fluxes, storage heat flux, etc.), and water transport. This results in a wide range of models (e.g., Brown & Williams, 1998; Oke et al., 1999; Grimmond & Oke, 1999; Kusaka et al., 2001; Masson, 2000; Dupont, 2001; Martilli et al., 2002). Most urban NWP or meso-meteorological models modify the existing non-urban approaches (e.g., the Monin-Obukhov similarity theory, MOST) for urban areas by parameterisation or finding proper values for the effective roughness lengths, displacement height, and heat fluxes, including the anthropogenic heat flux, heat storage capacity, albedo and emissivity change, etc. The main limitation is when there is a need to resolve meteorological profiles within the urban canopy, where the MOST assumption of a constant flux surface layer is invalid. This is obviously important as it is a layer into which pollutants are emitted and in which people live. The sophistication of urbanization within research mesoscale models has increased during the last 10 years, starting with the work of Brown & Williams (1998), which included urban effects in their TKE scheme. Masson (2000) then included a detailed canyon energy balance scheme into his surface energy balance equation. Martilli et al. (2002) expanded on the work of these two studies to include effects from canyon walls, roofs, and streets in each prognostic PBL equation. A similar, but less complex urbanization scheme has been developed by Kusaka et al. (2001). A drawback to these advanced urbanization schemes is that they require detailed (i.e., on scale of a few 10s of meters) urban morphological data, including land use and land cover, surface roughness, building thermal characteristics, and anthropogenic heat fluxes. The given below recommendations are based on detailed evaluation of the (i) urban physiographic data classification and utilisation of surface satellite data; (ii) parameterisations and models of urban


35

soil/heat, roughness sublayer and internal boundary layers; (iii) urbanisation of existing meso-meteorological models including th enumerical weather predictionmodels, (iv) urban sublayer models, parameterisations and meteo-preprocessors for urban air quality and emergency preparedness models, and (v) incorporation of urban effects into regional climate models.

2. Model Urbanization Strategy: Summaries, Recommendations and Requirements The urban canopy (UC), the layer of the atmosphere between the ground and the top of the highest buildings, is the region where people live and human activities take place. Because of this importance (e.g., human health, preservation of buildings) significant efforts have been dedicated to its investigation. Such studies shed light on the high complexity of atmospheric circulations in the UC, primarily because of the presence of obstacles (buildings) large enough to strongly modify air flow and the thermal exchanges between these surfaces and the atmosphere. The high level of heterogeneity of the UC has been a challenge for atmospheric modeling in urban areas, even for mesoscale models with a typical resolution of the order of 1 km; the basic characteristics of the perturbations induced by the obstacles still remaining unresolved at this model resolution. Over the last decade, with the increase of computational processing unit (CPU) power, several mesoscale modeling systems, each with different urban canopy parameterization (UCP) schemes, have been developed and applied with the primary aim of representing the subgrid effects of urban surfaces on their mean variables.

2.1 “Fitness-for-purpose” guidance UCP schemes used in models cover a wide spectrum ranging, from simple ones with a limited number of parameters, such as basic roughness and scale length for thermal or density stability, to multi-parameter sets that include vertical profile descriptions of building and vegetation size and shape. As their level of detail increases, the computational demands for running such models also increase. We note that there are no existing rules governing the appropriate levels of detail and specificity of UCPs that a model must have. However, it is of practical importance to achieve a balance between the level of detail and precision desired to describe the urban boundary layer with the computational costs and availability of commensurate descriptive data to run such models. This leads to a practical guideline that the choice of level of descriptive complexity of these UCPs be based both on “fitness- for-purpose” and the appropriate grid resolution of the requisite application. Here we list and highlight the requirements of five common applications. (1) Air quality exposure studies to assess the impact of atmospheric pollutants on human health.

Model concentration outputs are needed that accurately characterize pollution “hot spots” or gradients at a sufficiently fine grid resolution commensurate to the extent in which significant exposure impacts occur.

(2) Urban climatology studies and development of strategies for mitigating the intensity of heat islands. Information is needed to estimate human comfort and stress based on air temperature, relative humidity, and solar radiation. Model parameterization schemes need information about physical attributes of the underlying surface (buildings and vegetation), such as albedo, soil moisture, building material’s thermal conductivity, and capacity as well as anthropogenic sources of heating.

(3) Emergency response and predicting for site locations where toxic gases have been released. Needs improved methods and modeling of urban-scale transport and building and street canyon resolved dispersion and inverse modeling approaches for determining release location.


36

(4) Advanced air quality and weather forecasting to improve on the predicted future state of the atmosphere (clouds, rain, air temperature, winds, etc.) and to inform and provide guidance to the public on adverse air quality conditions.

(5) Urban planning to evaluate local climate and air quality impacts caused by urban developments and three-dimensional (3D) urban morphological structures.

Air quality exposure, urban climatology, emergency response, and urban planning models need detailed resolution of UC features, whereas weather and air quality forecasts are more focused on estimating the gross vertical exchange of heat, momentum, and pollution between the top of the canopy and the atmosphere. Case studies supporting air quality assessments, urban climatology, and urban planning studies are not relatively constrained with large CPU demands to achieve their target accuracy and precision estimates; whereas weather forecasting and emergency response model applications must, for practical reasons, scale down the details of their UC descriptions to achieve the required rapid output response times. At some point, it will be necessary to perform evaluation of models based on their fitness for purpose. Depending on the type of application, the ranking of atmospheric variables by their roles or importance may be useful for operational model evaluation purposes (see Table 2.1). This exercise is somewhat subjective as the atmospheric variables are interconnected in some way. For example, wind speed and direction is considered more important for air quality and dispersion applications than for urban climatology studies as those variables control pollutant transport. However, the role of wind is of indirect importance because it affects the magnitude of heat exchange between surfaces (walls, roofs, and streets) and the atmosphere, thus impacting urban microclimates.

Table 2.1: Ranking of importance of variables by (example) application. /I – important; VI – very important; VVI – very-very important/

Application

vs. Importance

Air Quality

Urban Climatology

Emergency Response

Weather Forecasting

Urban Planning

Wind Speed VI I VI I (above the canopy) VI Wind Direction VI I VI I (above the canopy) VI

Temperature (and Humidity)

I VVI I VI (2-m temperature) VI

Pollutant Concentration

VVI VI VI

Turbulent Fluxes

VI VI VI (at top of the canopy) VI

Additional criteria are needed for a robust evaluation based on fitness of purpose concepts. For example, whereas predicted pollutant concentration is a crucial variable for air quality studies, it will be important, in some applications, to focus on different statistical measures. For example, when considering averaging time, one should be clear whether the focus is on the averaging period, on the peak or the number of hours above a certain threshold, or on some other discriminator. Similarly, it would be useful to set objectives based on the degree of precision needed (e.g., Is it sufficient to have a modeled wind speed within 1 m s-1 of measurements for air quality simulations?). Thus, practical targets to be reached in terms of level of precision of outputs for the UCP implemented into models would be established for the models’ intended use at the outset of the evaluation.


37

2.2 Strategy to urbanize different types of models Current types of UC schemes available for model implementation are reviewed in this section, in the context of their application requirements. Given different modeling objectives, there are several types of UC schemes and associated atmospheric models available. They can be separated into three primary categories:

(1) single-layer and slab/bulk-type UC schemes, (2) multilayer UC schemes, and (3) obstacle-resolved microscale models.

The first two categories are sufficiently simple (in their grid-averaged representation of urban morphological features as parameters) to be coupled into classical numerical atmospheric models. The third corresponds to computational fluid dynamic (CFD)-type explicit building-scale resolved models.

Com

puta

tiona

l Req

uire

men

ts

Number of Parameters

Com

puta

tiona

l Req

uire

men

ts

Number of Parameters

Parameters difficult to get?

Too expensive to run?

Globally moreapplicable?

Parameters difficult to get?Parameters difficult to get?

Too expensive to run?Too expensive to run?

Globally moreapplicable?Globally moreapplicable?

Figure 2.1: Schematic diagramme depicting computational requirement increases with the inclusion of increased levels of UCP sophistication in UC models

(from workshop presentation by Grimmond et al., 2008). The simplest approaches, which include the traditional Reynold’s averaging scheme (using roughness [Zo] and displacement length), are single-layer schemes that link the UC effects to the atmospheric boundary layer through the model’s lowest layer. In these methods, the urban scheme is implemented through parameterization of each grid’s radiative and turbulent flux values. Moreover, details regarding drag aspects typically are addressed through various ad hoc approaches. For example, simple analytical wind profile formulations for applications inside the canopy typically are introduced. Removing this limitation requires implementation of urban schemes with multi-layers in which the flux quantities interact with the atmospheric variable (Martilli et al., 2002; Dupont et al., 2004). This approach requires additional terms in the prognostic equations of the atmospheric models (e.g., drag term in momentum equations, heating term in temperature equations, production term in turbulent kinetic energy equations). Such models require the addition of layers from the surface to the top of the highest urban feature, thus representing the morphological features as functions of height for each grid. This allows the schemes to model the interactions between air and the urban environment at several heights. Thus, it is possible to simulate the in-canopy flows


38

with greater precision than in single-layer models. However, additional prognostic equations and vertical model levels are required for this type of implementation. Consequently, whereas the effects of surface features are better represented, the computation burden is increased because of the increased integration time step and treatment of additional modeling details. This additional burden presents a limit in the use of multilayer canopy models in NWP forecasting. Clearly, as seen in Figure 2.1, care must be taken at the outset to understand and balance the need for greater precision obtainable with full canopy details and model turnaround time.

2.3 Overview of major applications

2.3.1 Numerical weather prediction and meso-meteorological models The simplest approach for meteorological models is to modify the existing non-urban approaches (e.g., the Monin-Obukhov similarity theory, MOST) for urban areas by introducing different values to represent each grid’s effective roughness lengths, displacement height and components and parameters of heating, including the anthropogenic heat flux, heat storage capacity, albedo, and emissivity for each urban land use class. Operational forecasts for urban areas using models with increasingly more sophisticated urban schemes will require significant advancements in computer power. Beginning with Brown & Williams (1998), who included urban effects in their turbulence closure scheme, methods with increasing levels of sophistication have been introduced into today’s mesoscale models. Masson (2000) included a detailed canyon energy balance scheme into the surface energy balance, whereas Martilli et al. (2002) and Dupont et al. (2004) included the effects from canyon walls, roofs, and streets in each prognostic planetary boundary layer (PBL) equation. A similar, but less complex urbanization scheme that shows promise toward capturing fine-scale urban weather phenomena, was a single-layer scheme developed by Kusaka & Kimura et al. (2004ab). With these advances came the requirement for detailed urban morphological data (i.e., on the scale of a few meters), including land use and land cover, surface roughness, building geometric and thermal characteristics, and anthropogenic heat fluxes (Ching et al., 2009). Thus, depending on fit-for-purpose analyses for specific urban applications, the next level of sophistication in NWP models may be through implementation of advanced single-layer UCP schemes. This approach is a relatively inexpensive and practical means of improving on the modified MOST approach.

2.3.2 Urban air pollution and emergency response models Urban and regional-scale atmospheric pollution models, can operate in either a prognostic mode, or in post event analysis i.e., retrospective mode. Each mode has prioroties and requirements that do not necessarily overlap. On prognostic mode, air quality forecasts are produced using meteorological information from forecasts of the operationally run meteorological models. The retrospective mode is used in air quality simulations necessary to conducting regulatory impact and cost-benefit analyses, developing source control strategies, and performing human exposure assessments. Such simulations require the highest precision and accuracy possible based on the most complete and highly detailed meteorological simulations for specific meteorological scenarios of interest, typically those for which air quality is poorest. For retrospective assessments, the precision and accuracy of the meteorological simulation is more important than timeliness. For air quality forecasts, the sophistication of UCP schemes in numerical weather prediction models is governed by operational requirements. To obtain products, that will help with guidance in


39

reducing poor air quality in street canyons, from forecast mode, may require special ad hoc urban meteorological post processing. However, for applications such as emission control strategies, where vertical profiles of the meteorological and turbulent characteristics are needed in great detail, similar post-processing might not be sufficient. Studies to assess air pollution health effects are an important objective that may be satisfied best with retrospective approaches. Population exposure modeling will require highly detailed multi-pollutant and multi-scale air quality models, as well as high-resolution urban morphology, population distribution, and human activities databases (Baklanov et al., 2007). For these and similar applications, the emphasis will be on implementing urban schemes at a grid resolution that can provide the appropriate transport and turbulence details within the UC. The fitness-for-purpose analysis also governs the choice of local-scale emergency preparedness modeling for accidental biological, chemical, or nuclear releases, and moreover, is clearly one of scale. For direct response and for operational purposes where timeliness is important urban meteorological observations and/or forecast products coupled with dispersion models are appropriate. For planning and assessment purposes and for the near-sources region, obstacle-resolved modeling approaches (e.g., CFD modeling) may be required. Such approaches will require careful linkage to outputs of urban-scale models, and both will require basic building and vegetation descriptions. Ideally, specific urban feature effects that should be incorporated into this type of application will include the following: • Impact of urban surfaces on pollutant deposition (e.g., vertical walls, building materials and

structure, vegetation) • Information regarding chemical transformation such as lifetime of chemical species (e.g., inside

street canyons), heterogeneity of solar radiation (street shadows, albedo, and emissivity), and specific aerosol dynamics in street canyons (e.g., resuspension processes)

• Very detailed, high-resolution data on the mobile emission of pollutants • Indoor-outdoor pollutant exchange information

2.3.3 Multiscale atmospheric environment modeling Air quality in urban areas is influenced by local pollutant emission sources as well as transport of species on regional and global scales. In turn, transport of air pollutants from urban areas will impact on regional and global scale air quality. Current atmospheric-chemical-transport (ACT) models apply model nesting approaches as a means for treating the up- and down-scaling to account for this multi-scale dimension (Moussiopoulos, 1995; Fernando et al., 2001; Baklanov, 2007). For down-scaling, a chain of urban models of different scales with sub-domain nests using finer grid sizes is applied. A common approach is to use outputs of large-scale models as boundary condition inputs to domains employing smaller grids successively from global to urban and street scales. It is well recognized that transport and transformation are nonlinear in scale (especially for reactive and rapidly deposited species), and parameters controlling atmospheric processes are typically grid-size dependent. Usually, the microscale (street canyon) models are obstacle-resolving and consider the detailed geometry of the buildings and UC, whereas the up-scaled city-scale (sub-meso) or mesoscale models consider parameterizations of urban effects or statistical descriptions of the urban building geometry. Downscaling from regional (or global) meteorological models to the urban-scale meteorological models, with statistically parameterized building effects, and further downscaling to microscale obstacle-resolved, CFD-type models was included in the methodology. Likewise, methods are needed for regional- and global-scale models to properly account for


40

downwind transport of pollutant species from urban sources in regional- and global-scale contexts. This is because the modeled composition by species is grid-size dependent. Thus, for global and climate change models, the mesoscale model can provide a proper pollutant species accounting from biogenic and urban sources ranging from small urban areas to megacities to regional and global scales. It serves investigations of the evolution of pollutants from large urban plumes (e.g., Sarrat et al., 2006) or from major industrial and power-generation point sources. Such plumes are subgrid phenomena for the regional-global models that have the highest resolution (between 10- and 100-km grid sizes) in the zoomed areas. Therefore, urban-scale models can provide appropriate composition mix for the regional-global model. Currently, to understand the impact of aerosols and gas-phase compounds emitted from local/urban sources on regional and global scales, at least three scales of the integrated atmosphere-chemistry-aerosol and general circulation models are being considered: (1) local, (2) regional, and (3) global. Note that two-way nesting approaches are ideal for situations in which the scale effects in both directions (from the mesoscale on the microscale and from the microscale on the mesoscale) are important. However, such approaches are difficult to implement.

2.3.4 Urban pollution and climate integrated modeling Integrated air quality modeling systems are tools that help in understanding impacts from aerosols and gas-phase compounds emitted from urban sources on the urban, regional, and global climate. The integration of urbanized numerical weather prediction and atmospheric chemical transport models is a strategic approach to providing the science-based tools for assessments of urban air quality and population exposure in the context of global to regional to urban transport and climate change. This is reasonable because meteorology governs the transport and transformations of anthropogenic and biogenic pollutants, drives urban air quality and emergency preparedness models; meteorological and pollution components have complex and combined effects on human health (e.g., hot spots, heat stresses); and pollutants, especially urban aerosols, influence climate forcing and meteorological events (precipitation, thunderstorms, etc.). The online integration of mesoscale meteorological models and atmospheric aerosol and chemical transport models enables the utilization of all meteorological 3D fields in ACT models at each time step and the consideration of feedback among air pollution (e.g., urban aerosols), meteorological processes, and climate forcing (e.g., Enviro-HIRLAM: Baklanov et al., 2008, WRF-Chem: Grell et al., 2005). Chemical species in the atmosphere, such as CO2 and ozone act as greenhouse gases to influence weather and atmospheric processes. Aerosols such as sea salt, dust, primary and secondary particles of anthropogenic and natural origin are also airborne and contribute to atmospheric processes in a complex manner. Some aerosol components (black carbon, iron, aluminum, and polycyclic and nitrated aromatic compounds) warm the air by absorbing solar and thermal-infrared radiation, whereas others (water, sulphate, nitrate, and most organic compounds) cool the air by backscattering incidental short-wave radiation into space. The effects of urban aerosols and other chemical species on meteorological parameters have many different pathways (direct, indirect, semi-direct effects, etc.) that these online, coupled modeling systems are capable of addressing.

2.4 Database and evaluation aspects of urbanized models It is evident that there are a large range of applications that involve an urban focus. Moreover, given the wide range of model complexities, operational and data input requirements, and diverse applications, we find that there is no “one-size-fits-all” modeling approach that addresses the wide range of modeling objectives. Thus, for urban applications, the fitness-for-purpose concept is a


41

relevant and important consideration. In this survey, we have identified a number of considerations; some of the major ones are outlined below.

2.4.1 Database requirements Models of urban areas will be required to provide reliable predictions at fine 3D resolution of turbulent exchanges, air flow and thermodynamic characteristics. To meet these requirements, parameterizations are being developed and implemented with varying degrees of detail in terms of features and sophistication relative to the actual physiographic features of individual cities. One limitation to the degree of complexity in the model parameterizations is the availability of appropriate morphology information. For operational needs, the requirements are fulfilled using specifications associated with limited numbers of urban land use categories, each with specified surface properties such as roughness, displacement lengths, albedo, moisture availability, and thermal properties. For research and development applications, models that can capture more detailed effects of urban morphological features and underlying surfaces and building materials, at increasingly higher spatial resolutions, employ more explicit and highly detailed sets of 3D canopy parameters and within-grid land use classes. A common requirement for environmental models is the description of the underlying surface layer. Technological advancements allow increasingly sophisticated definitions of land cover characteristics (e.g., shape files with high resolution [~1 m] definition of buildings and vegetation). Data of this type are now becoming routinely available for many urban areas of the world with the information technology available to facilitate dissemination. In the United States, a pilot project is underway to serve as a community-based technology enabler of such data; this or comparable systems can be developed to handle the needs on an international basis (Ching et al., 2009). A community-based system should decrease administrative barriers and increase international collaborative efforts to advance modeling tools.

2.4.2 Evaluation Once the target variables and degree of precision needed for the application purpose are identified (Table 2.1), it is necessary to determine whether the parameterizations are capable of reaching these targets. Several techniques are available. • Real scale measurements. As measurements are taken in a real city, a model should be able to

reproduce them; however, very often it is difficult to have enough measurements, and, where measurements are taken, its representativeness of the gridded fields must be ascertained. The model computes the equivalent of a spatial average over the grid cell (usually a few kilometers or, at best, several hundreds of meters). Outputs from models that introduce vertical resolution within the UC and capture the effects of urban building and vegetation features are virtual fields, and the task of evaluating such outputs is challenging. Model-predicted vertical profiles of variables in the canopy reflect the aggregated influence of all the canopy features as virtual elements within the grid. In reality, such features take up finite volumes, and building-induced flows are subgrid features. Thus, any single or set of measurements will not provide a representation of the gridded fields but will, more or less, be under the influence of the nearest buildings or obstacles. This is a design feature that has yet to be resolved in developing field measurement strategies to evaluate predictions of within-canopy fields. Future guidance may come from insights gained using coupled UC models and building-resolved flows, both of which are driven by the same set of building datasets. Currently, evaluations performed above the canopy layer (blending layer) should not be subject to this conceptual difficulty, but, in and of itself, it does not provide the requisite within-canopy evaluation.


42

• Remote sensing data. A variety of satellite platforms do now provide data on surface variables and for urban areas. In particular, skin temperature is considered a very important variable because it exerts a strong control on boundary layer processes and the intensity of the heat island. Such data would be very useful toward diagnostic evaluation of urban model predictions. Of course, care is required to address scale issues of observation and models. There are some critical assumptions in the derivation of the remotely sensed variables (e.g., emissivities, mixed pixels). The comparison also is biased to conditions that the remotely sensed data are operational (e.g., clear sky conditions for surface temperature, time of overpass). Also, because models have varying treatments for handling subgrid land use and coverage, some of the resulting differences between observed and modeled skin temperature may be result, in part, from these treatments.

• Scale-model measurements. Wind tunnels have the advantage that external conditions can be controlled and are repeatable, but are limited by certain conditions (e.g., Reynolds number may be a factor of 100 less than in the real world; no concurrent radiative moisture forcing; typically treats only neutral stratification cases). Numerical models allow for a wide range of conditions with real meteorological forcing to be compared (Kanda, 2009). To date these models remain simple in morphology and arrangement.

• CFD (large eddy simulation [LES] or Reynolds-averaged Navier-Stokes [RANS]) models. Such building resolving models can be run over a limited part of a city to investigate flow properties to be used in UCP. Using CFD models, it is possible to derive the spatial averages required for UCP (Galmarini et al., 2009; Martilli & Santiago, 2009). CFD-RANS lacks the accuracy for some complex configurations. CFD-LES is more accurate but much more expensive in CPU time, which, thereby, limits its use.

• Operational testing. Real-scale routine data from weather networks are used for evaluation, most typically for weather forecasts (e.g., Bohnenstengel & Schlünzen, 2009).

2.5 Potential community activities There is a wide range of activities that are needed to support the recent improvements to the state of urbanization of models. These fall into a variety of categories. To date, a systematic evaluation of urban land surface schemes has not taken place as it has for vegetated environments. The model comparison outlined in Grimmond et al. (2009) takes some initial steps to address this. As they note, it is anticipated that there will be need for further observations. There is a clear need for both intensive and extensive observational data sets to allow the wide range of variables to be evaluated over a wide range of synoptic conditions. The development of urban testbeds and urban atmospheric observatories (e.g., Helsinki, Shanghai, London, Paris, Hanover, Phoenix, Oklahoma City, Houston, New York City, and Washington, D.C.) and long-term urban campaigns (e.g., CAPITOUL, BUBBLE) enable these issues to be addressed. For example, studies evaluating the Martilli scheme (Martilli et al., 2002) show that it is able to reproduce the generation of the urban heat island effect and to represent correctly most of the behavior of the fluxes over Basel and Marseilles city centers (Hamdi & Schayes, 2005). There is a continuing need for modelers and observers to communicate. As models are used for a variety of purposes, there is a need for increasing the range of variables observed to ensure as complete a range of evaluation as possible. This may mean having testbeds and observatories with different objectives and dataset richness. There is a wide range of processes and variables that need to be evaluated over a broad spectrum of conditions (meteorological, morphological, geographical setting, etc.). For example, a deeper understanding of urban PBL dynamics requires development of long-term urban testbeds in a variety of geographic regions (e.g., inland, coastal, complex terrain) and in many climate regimes, with a variety of urban core types (e.g., deep versus shallow, homogeneous versus heterogeneous).


43

The conceptual issue of evaluation of model prediction of the flow within the canopy is not satisfactorily resolved at this time, and a framework to address this is needed. Ideal urban testbeds would include quasi-permanent mesoscale networks, with surface, canyon, rooftop, and PBL meteorological and air quality observations. These real-time, quality-assured data would be used for real-time urban-scale weather and air quality forecasts, as well as for emergency response actions after releases of air toxins (with an indoor-outdoor linkage) and for climate change impact studies. In addition, the testbeds should be able to accommodate intensive short-term field observational studies that could involve turbulent flux and pollutant tracer measurements. Problems also exist in the evaluation of microscale CFD meteorological model results by use of field study or canyon wind tunnel observations (e.g., wind tunnel wall effects, the isolated nature of wind tunnel urban domains, the periodic LES and CFD lateral boundary conditions). When comparisons are done with these limitations in mind (e.g., only compare model results with wind tunnel results over urban centers), however, they show good agreement among the methods. Obviously, with increasing evaluation, there will be enhanced development of the models. It is also clear that, within the chain of needs between meteorological forcing and applications, there is a range of new developments needed. Finally, user friendly and multifaceted urban databases and enabling technology are critical and core capabilities for advancing urban modeling and boundary layer research. We see the National Urban Database and Access Portal Tool as a research and development resource toward future improved UCP descriptions and scientific bases for advanced urban modeling applications. With careful thought to its implementation, the concept of this prototype system is extensible on an international basis. For such an enterprise, we suggest several guiding principles be adopted. First, that this type of database be open and community-wide and available both universally and in as an unrestricted form as possible. Second, that both protocols and mechanisms should be established for its maintenance, upgrading, updating, and archiving. Further, issues of availability and sources of high-resolution data sets will need to be addressed.


44

References Baklanov A., U. Korsholm, A. Mahura, C. Petersen, A. Gross, 2008: ENVIRO-HIRLAM: on-line

coupled modelling of urban meteorology and air pollution. Advances in Science and Research, 2, 41-46.

Baklanov, A., 2007: Urban air flow researches for air pollution, emergency preparedness and urban weather prediction. Chapter 9 in: Flow and transport processes with complex obstructions: Applications to cities vegetative canopies and industry. Eds. Ye. A. Gayev and J.C.R. Hunt, Springer, 311-357.

Baklanov, A., A. Rasmussen, B. Fay, E. Berge and S. Finardi, 2002: Potential and Shortcomings of Numerical Weather Prediction Models in Providing Meteorological Data for Urban Air Pollution Forecasting. Water, Air and Soil Poll.: Focus, 2(5-6): 43-60.

Baklanov, A., O. Hänninen, L. H. Slørdal, J. Kukkonen, N. Bjergene, B. Fay, S. Finardi, S. C. Hoe, M. Jantunen, A. Karppinen, A. Rasmussen, A. Skouloudis, R. S. Sokhi, J. H. Sørensen, and V. Ødegaard, 2007: Integrated systems for forecasting urban meteorology, air pollution and population exposure. Atmos. Chem. Phys., 7:855-874.

Baklanov, A., S. Joffre, and S. Galmarini (Editors), 2005: "Urban Meteorology and Atmospheric Pollution (EMS-FUMAPEX)". Atmospheric Chemistry and Physics, Special Issue 24. http://www.atmos-chem-phys.net/special_issue24.html.

Baklanov A., J. Ching, CSB Grimmond, A. Martilli (2009): Model Urbanization Strategy: Recom-mendations and Requirements. In Urbanization of Meteorological and Air Quality Models; Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou (Eds), Springer Publishers, ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Baklanov A., S. Grimmond, A. Mahura, M. Athanassiadou (Eds) (2009): Urbanization of Meteoro-logical and Air Quality Models. Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Bohnenstengel, S., and H. Schlünzen, 2009: Performance of different sub-grid-scale surface flux parameterizations for urban and rural areas. In Urbanization of Meteorological and Air Quality Models. Baklanov, Grimmond, Mahura, Athanassiadou (Eds). Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Bornstein, R., R. Balmori, H. Taha, D. Byun, B. Cheng, J. Nielsen-Gammon, S. Burian, S. Stetson, M. Estes, D. Nowak, and P. Smith, 2006: Modeling the effects of land-use land cover modifications on the urban heat island phenomena in Houston, Texas. SJSU Final Report to Houston Advanced Research Center for Project No. R-04-0055, 127 pp.

Brown M. and Wiliams M., 1998: An Urban canopy parameterization for Mesoscale Meteorological Models. AMS 2nd Urban Environment Symposium, Albuquerque, NM USA.

Chen, F., Kusaka, H., Tewari, M., Bao, J-W, and Hirakuchi, 2004: “Utilizing the coupled WRF/LSM/urban modeling system with detailed urban classification to simulate the urban heat island phenomenon over the greater Houston area”, Paper 9.11, American Meteorological Society Fifth Symposium on the Urban Environment, 23-27 August 2004, Vancouver, British Columbia.

Chen, F., M. Tewari, H. Kusaka and T. L. Warner, 2006: Current status of urban modeling in the community Weather Research and Forecast (WRF) model. Sixth AMS Symposium on the Ur-ban Environment, Atlanta GA, January 2006.

Ching, J., A. Hanna, F. Chen, S. Burian, and T. Hultgren, 2008: Facilitating advanced urban


45

meteorology and air quality modeling capabilities with high resolution urban database and access portal tools. In Urbanization of Meteorological and Air Quality Models. Baklanov, Grimmond, Mahura, Athanassiadou (Eds). Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Dupont, S., 2001 : Modelisation Dynamique et Thermodynamique de la Canopee Urbaine: Realisa-tion du Modele de Sols Urbains pour SUBMESO, Doctoral thesis, Universite de Nantes, France.

Dupont, S., T.L. Otte, and J.K.S. Ching, 2004: Simulation of meteorological fields within and above urban and rural canopies with a mesoscale model (MM5) Boundary-Layer Meteor., 2004, 113:111-158.

Fernando H.J.S., S.M. Lee, J. Anderson, M. Princevac, E. Pardyjak, and S. Grossman-Clarke, 2001: Urban fluid mechanics: air circulation and contaminant dispersion in cities. J Environ. Fluid. Mech. 1(1):107-164.

Galmarini, S., J.-F. Vinuesa, and A. Martilli, 2009: Relating small-scale emission and concentration variability in air quality models. In Urbanization of Meteorological and Air Quality Models. Baklanov, Grimmond, Mahura, Athanassiadou (Eds). Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Grell, G. A., S. E. Peckham, R. Schmitz, S. A. McKeen, G. Frost, W. C. Skamarock, and B. Eder, 2005: Fully coupled “online“ chemistry within the WRF model, Atmos. Environ., 39(37), 6957–6975.

Grimmond, C.S.B. and Oke, T.R., 1999: Heat storage in urban areas: observations and evaluation of a simple model, J. Appl. Meteorol., 38, 922-940.

Grimmond, C.S.B., M. Best, and J. Barlow, 2008: Urban surface energy balance models: model characteristics and methodology for a comparison study.

Hamdi, R., and G. Schayes, 2005: Validation of the Martilli urban boundary layer scheme with measurements from two mid-latitude European cities. Atmos Chem. Phys. Discuss. 5, 4257-4289.

Kanda, M., 2009: Review of Japanese urban models and a scale model experiment. In Urbanization of Meteorological and Air Quality Models. Baklanov, Grimmond, Mahura, Athanassiadou (Eds). Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.

Kusaka, H., and F. Kimura, 2004a: Coupling a single-layer urban canopy model with a simple atmospheric model: Impact on urban heat island simulation for an idealized case. J. Meteor. Soc. Japan, 82:67-80.

Kusaka, H., and F. Kimura, 2004b: Thermal effects of urban canyon structure on the nocturnal heat island: numerical experiment using a mesoscale model coupled with an urban canopy model. J. Appl. Meteor., 43:1899-1910.

Kusaka, H., Kondo, H., Kikegawa, Y., and Kimura, F., 2001: A Simple Single-Layer Urban Canopy Model for Atmospheric Models: Comparison with Multi-Layer and SLAB Models, Boundary-Layer Meteorol. 101, 329-358.

Martilli A., Clappier A., and Rotach MW. 2002. An urban surface exchange parameterization for mesoscale models. Boundary-Layer Meteorology 104: 261-304.

Martilli, A. and J. L. Santiago, 2009: How to use computational fluid dynamics models for urban canopy parametrizations. In Urbanization of Meteorological and Air Quality Models. Bakla-nov, Grimmond, Mahura, Athanassiadou (Eds). Springer Publishers, 169p., ISBN 978-3-642-00297-7; DOI 10.1007/978-3-642-00298-4.


46

Masson, V., 2000: A physically-based scheme for the urban energy budget in atmospheric models. Boundary-Layer Meteorol. 98:357-397.

Moussiopoulos, N., 1995: The EUMAC Zooming Model, a tool for local-to-regional air quality studies. Meteorol. Atmos. Phys., 57, 115-133.

Oke, T.R., Spronken-Smith, R., Jauregui, E. and Grimmond, C.S.B., 1999: Recent energy balance observations in Mexico City. Atmos. Environ., 33, 3919-3930.

Piringer, M., S. Joffre, A. Baklanov, A. Christen, M. Deserti, K. De Ridder, S. Emeis, P. Mestayer, M. Tombrou, D. Middleton, K. Baumannstanzer, A. Dandou, A. Karppinen, J. Burzynski, 2007: The surface energy balance and the mixing height in urban areas – activities and recommendations of COST Action 715. Bound.-Layer. Meteor. 124: 3-24, DOI : 10.1007/s10546-006-9124.

Sarrat, C., A. Lemonsu, V. Masson, G. Guedalia, 2006: Impact of urban heat island on regional atmospheric pollution. Atmos. Environ., 40:1743-1758.

Taha, H., 2008: Sensitivity of the urbanized MM5 (uMM5) to perturbations in surface properties in Houston Texas. Boundary-Layer Meteorology, 127: 193-218.


47

Acknowledgements The research leading to these results has received funding from the European Union’s Seventh Framework Programme FP/2007-2011 under grant agreement n°212520. The contributions from the COST Action 728, FUMAPEX, and other projects have been analysed and summarized in this deliverable report. The materials published in the quarterly MEGAPOLI project newsletters have been also used in this report. Thanks to Roman Nuterman (DMI) for assistance with checking of references used.


48

Previous MEGAPOLI reports Previous reports from the FP7 EC MEGAPOLI Project can be found at: http://www.megapoli.info/

Collins W.J. (2009): Global radiative forcing from megacity emissions of long-lived greenhouse gases. Deliverable 6.1, MEGAPOLI Scientific Report 09-01, 17p, MEGAPOLI-01-REP-2009-10, ISBN: 978-87-992924-1-7

http://megapoli.dmi.dk/publ/MEGAPOLI_sr09-01.pdf

Denier van der Gon, HAC, AJH Visschedijk, H. van der Brugh, R. Dröge, J. Kuenen (2009): A base year (2005) MEGAPOLI European gridded emission inventory (1st version). Deliverable 1.2, MEGAPOLI Scientific Report 09-02, 17p, MEGAPOLI-02-REP-2009-10, ISBN: 978-87-992924-2-4


Baklanov A., Mahura A. (Eds) (2009): First Year MEGAPOLI Dissemination Report. Deliverable 9.4.1, MEGAPOLI Scientific Report 09-03, 57p, MEGAPOLI-03-REP-2009-12, ISBN: 978-87-992924-3-1


Allen L., S Beevers, F Lindberg, Mario Iamarino, N Kitiwiroon, CSB Grimmond (2010): Global to City Scale Urban Anthropogenic Heat Flux: Model and Variability. Deliverable 1.4, MEGAPOLI Scientific Report 10-01, MEGAPOLI-04-REP-2010-03, 87p, ISBN: 978-87-992924-4-8 http://megapoli.dmi.dk/publ/MEGAPOLI_sr10-01.pdf

Pauli Sievinen, Antti Hellsten, Jaan Praks, Jarkko Koskinen, Jaakko Kukkonen (2010): Urban Morphological Database for Paris, France. Deliverable D2.1, MEGAPOLI Scientific Report 10-02, MEGAPOLI-05-REP-2010-03, 13p, ISBN: 978-87-992924-5-5 http://megapoli.dmi.dk/publ/MEGAPOLI_sr10-02.pdf

Moussiopoulos N., Douros J., Tsegas G. (Eds.) (2010): Evaluation of Zooming Approaches De-scribing Multiscale Physical Processes. Deliverable D4.1, MEGAPOLI Scientific Report 10-03, MEGAPOLI-06-REP-2010-01, 41p, ISBN: 978-87-992924-6-2 http://megapoli.dmi.dk/publ/MEGAPOLI_sr10-03.pdf

Mahura A., Baklanov A. (Eds.) (2010): Hierarchy of Urban Canopy Parameterisations for Different Scale Models. Deliverable D2.2, MEGAPOLI Scientific Report 10-04, MEGAPOLI-07-REP-2010-03, 49p, ISBN: 978-87-992924-7-9 http://megapoli.dmi.dk/publ/MEGAPOLI_sr10-04.pdf


49

MEGAPOLI

Megacities: Emissions, urban, regional and Global Atmos-

pheric POLlution and climate effects, and Integrated tools for

assessment and mitigation

EC FP7 Collaborative Project

2008-2011

Theme 6: Environment (including climate change) Sub-Area: ENV-2007.1.1.2.1:

Megacities and regional hot-spots air quality and climate

MEGAPOLI Project web-site http://www.megapoli.info

MEGAPOLI Project Office Danish Meteorological Institute (DMI) Lyngbyvej 100, DK-2100 Copenhagen, Denmark

E-mail: [email protected] Phone: +45-3915-7441 Fax: +45-3915-7400

MEGAPOLI Project Partners

• DMI - Danish Meteorological Institute (Denmark) - Contact Persons: Prof. Alexander Baklanov (co-ordinator), Dr. Alexander Mahura (manager)

• FORTH - Foundation for Research and Technology, Hellas and University of Patras (Greece) - Prof. Spyros Pandis (vice-coordinator)

• MPIC - Max Planck Institute for Chemistry (Germany) - Dr. Mark Lawrence (vice-coordinator)

• ARIANET Consulting (Italy) – Dr. Sandro Finardi • AUTH - Aristotle University Thessaloniki (Greece)

- Prof. Nicolas Moussiopoulos • CNRS - Centre National de Recherche Scientifique

(incl. LISA, LaMP, LSCE, GAME, LGGE) (France) – Dr. Matthias Beekmann

• FMI - Finnish Meteorological Institute (Finland) – Prof. Jaakko Kukkonen

• JRC - Joint Research Center (Italy) – Dr. Stefano Galmarini

• ICTP - International Centre for Theoretical Physics (Italy) - Prof. Filippo Giorgi

• KCL - King's College London (UK) – Prof. Sue Grimmond

• NERSC - Nansen Environmental and Remote Sensing Center (Norway) – Dr. Igor Esau

• NILU - Norwegian Institute for Air Research (Norway) – Dr. Andreas Stohl

• PSI - Paul Scherrer Institute (Switzerland) – Prof. Urs Baltensperger

• TNO-Built Environment and Geosciences (The Netherlands) – Prof. Peter Builtjes

• MetO - UK MetOffice (UK) – Dr. Bill Collins • UHam - University of Hamburg (Germany) – Prof.

Heinke Schluenzen • UHel - University of Helsinki (Finland) – Prof.

Markku Kulmala • UH-CAIR - University of Hertfordshire, Centre for

Atmospheric and Instrumentation Research (UK) – Prof. Ranjeet Sokhi

• USTUTT - University of Stuttgart (Germany) – Prof. Rainer Friedrich

• WMO - World Meteorological Organization (Switzerland) – Dr. Liisa Jalkanen

• CUNI - Charles University Prague (Czech Repub-lic) – Dr. Tomas Halenka

• IfT - Institute of Tropospheric Research (Ger-many) – Prof. Alfred Wiedensohler

• UCam - Centre for Atmospheric Science, Univer-sity of Cambridge (UK) – Prof. John Pyle

Work Packages

WP1: Emissions (H. Denier van der Gon, P. Builtjes)

WP2: Megacity features (S. Grimmond, I. Esau)

WP3: Megacity plume case study (M. Beekmann, U. Baltensperger)

WP4: Megacity air quality (N. Moussiopoulos)

WP5: Regional and global atmospheric composition (J. Kukkonen, A. Stohl)

WP6: Regional and global climate impacts (W. Collins, F. Giorgii)

WP7: Integrated tools and implementation (R. Sokhi, H. Schlünzen)

WP8: Mitigation, policy options and impact assessment (R. Friedrich, D. van den Hout)

WP9: Dissemination and Coordination (A. Baklanov, M. Lawrence, S. Pandis)

Documents

MEGAPOLI Scientific Report 10-04 Hierarchy of Urban Canopy