Urban Land-Surface Characterisation for Modelling of Atmospheric Processes

Urban areas need to be represented in a wide variety of atmospheric models (e.g. NWP, GCM, AQ, LAM, CTM). These range in scale of applicability, with users concerned with global processes and long term climate to users who have responsibilities related to processes within small areas (e.g. streets, blocks, neighbourhoods) of an individual city (e.g. dispersion, urban planning, flooding). The urban part of these models all need to address the same general principles of the impact of urban/human activities on surface-atmosphere exchanges. The approach used depends on the primary objective of the modelling, computational resources, and modelling design. However, there are some common features that need to be considered, these include: the nature, rates, and spatial distributions of emissions (e.g. heat, aerosols) to the atmosphere; and the radiative, thermal, and biophysical characteristics of the surface.

To date a number of approaches have been taken to address the problem of providing the requisite information to drive these models, so there are a range of ways to characterize the urban surface. Some of the approaches have been developed specifically for modelling urban land-atmosphere interactions (e.g. Jackson et al. 2010, WUDAPT) while others make use of data that were intended for a variety of different purposes but have become regularly used for atmospheric modelling (e.g. USGS, NCLD data).

This talk will consider the following questions:

• What are the current requirements (e.g. parameters, level of uncertainty, etc) for characterising the urban land surface for modelling and observations?

• How do these vary with application/objective and scale of interest?

• How are these data currently obtained and what are the current limitations?

• Which of these are currently provided to an acceptable degree of accuracy by existing data sources?

• Where is future effort needed in order to make improvements?