INTERNATIONAL JOURNAL OF CLIMATOLOGY Int. J. Climatol. 27: 1909–1918 (2007) Published online 15 October 2007 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/joc.1620 Current research and future challenges in urban mesoscale modelling Alberto Martilli* CIEMAT, Research Centre for Energy, Environment and Technology, Avenida Complutense 22, 28040 Madrid, Spain Abstract: Increases in environmental problems linked to urbanization and increases in computational power are proposed as the main mechanisms behind positive feedbacks between experimental investigation and numerical modelling in mesoscale urban studies. The focus here is on the modelling, with a description of state-of-the-art techniques used to parameterize urban-induced dynamical and thermal effects in mesoscale models. An analysis of possible future developments in urban mesoscale modelling is presented. Emphasis is laid on the link between mesoscale and microscale models, increases in numerical resolution, and improvement in anthropogenic heat flux estimation. Copyright  2007 Royal Meteorological Society KEY WORDS air quality; mesoscale modelling; urban climatology; urban heat island; urban parameterizations; weather forecast Received 12 September 2006; Revised 20 July 2007; Accepted 29 July 2007 INTRODUCTION The mesoscale ranges from a few kilometres to several hundred kilometres in the horizontal and tens of metres to the depth of the troposphere in the vertical, with a time scale of about 1 to 12 h (Pielke, 1984). Circula- tions such as land/sea breezes, mountain/valley flows, and urban breezes fall within this scale. For these types of circulations, the interactions with planetary boundary layer (PBL) processes are crucial. Owing to the com- plexity of the problem, it is only with the advent of computers that studies of mesoscale circulations received a strong boost. In terms of urban mesoscale modelling, Delage and Taylor (1970) and Bornstein (1975) were one of the first to investigate the impacts of the urban heat island on wind fields and boundary layer structure. Such studies were motivated by the need for a better understanding of the dynamics and thermodynamics of the urban PBL, with particular interest in pollutant dis- persion, but were still very idealized and simple (two dimensional). Urban heat island was also investigated, with even simpler models, by Atwater (1972) and McEl- roy (1973). The continuous increase of computer power during the 1980s allowed relaxation of assumptions made to resolve the equations of motions (e.g. from hydro- static to non-hydrostatic), with a consequent enlargement of the domain of applicability. Models were still used as research tools to investigate physical mechanisms, but the configurations analysed shifted from simple and idealized * Correspondence to: Alberto Martilli, CIEMAT, Edificio 70, P1.11, Avenida Complutense 22, 28040 Madrid, Spain. E-mail: alberto.martilli@ciemat.es to realistic and complex (e.g. see the study of McNider and Pielke, 1984, on mountain/valley flows). However, it is only in the mid-1990s that interest in mesoscale modelling in general, and in urban mesoscale modelling in particular, began to increase very rapidly, as can be seen from the number of publications on this subject in scientific journals (Figure 1). Such an increase in interest results from interactions between several factors, as sketched in Figure 2. • The continuous increase in the urban population. In 1800, only 3% of the world’s population lived in urban areas. By 1900, almost 14% were urbanites, and, in 1950, 30% of the world’s population resided in urban centres. Today about half of the world population lives in cities, and it is forecast that, by 2030, 60% of the world population will live in cities (UN report, 2003). Moreover, the size of cities is also increasing. Just eight cities had populations of 5 million or more in 1950. This number increased to 46 in 2003. By 2015, it is projected to be 61, 45 in less-developed countries (UN report, 2003). Such fast and very often chaotic growth leads to a number of environmental problems, some of them linked with the atmosphere. The social pressure to find tools to understand and manage such problems increased parallel to the increase of urban areas. • As already mentioned, another very important factor that explains the popularity of urban mesoscale mod- elling studies in the last decade is the improvement of computer power. The performance of the fastest computers has grown exponentially from 1945 to the present, averaging a factor of 10 every 5 years (Fos- ter, 1994). In recent years, the use of massively parallel Copyright  2007 Royal Meteorological Society