New Urban Parameterizations in the Integrated ACASA Model

Approaches on modeling urban environment have often focused on specific components of cities complexity. In some simple models, cities have been simply described as a single-layer area showing aerodynamic and thermal variations from the surroundings. Multi-layer models have been used to describe vertical interactions within a multi-layered paradigm. These allow key features of turbulent exchange of physical quantities between levels, as well as vertically differentiated source and sink terms within the urban canopy. Another improvement is to describe anthropogenic fluxes of heat, later differentiating it between sensible and latent heat. However, usually heat fluxes from buildings have not also included traffic fluxes, and human-related fluxes have generally been ignored. Many of the building energy models used were usually run on a climatological basis, neglecting the actual daily and hourly local meteorological conditions. Furthermore, many models have ignored carbon dioxide fluxes , due both to the difficult modeling of anthropogenic emissions and to the lack of an adequate soil-canopy-atmosphere exchange model.

To overcome these limitations, the present work proposes an urbanized version of the UC Davis developed numerical SVAT model ACASA (Advanced Canopy-Atmosphere-Soil Algorithm). This model, initially created for vegetated canopies, has widely-tested and simulates CO2, moisture, radiation and sensible heat fluxes in natural environments. Its layered structure of soil and atmosphere permits a realistic third-order turbulent exchange of physical quantities. It can be run stand-alone (in-situ simulation on a footprint centered on a Eddy Covariance tower) or coupled interactively with the atmospheric model WRF (Weather Research and Forecasting) for regional scale simulation. The model ACASA has now been modified to include an improved, detailed urban scheme that accurately reproduces radiative and aerodynamic properties. A few basic anthropogenic fluxes (namely carbon dioxide from traffic and human respiration, and sensible heat from building heating) have been included, showing a discrete capability to reproduce observations. An additional, more complete set of urban parameterizations has been introduced, making the model able to simulate building and natural canopy in an integrated mode. No anthropogenic contributions are neglected, with the incorporation of fluxes of CO2, moisture and sensible heat from buildings, traffic and human population. Within the vertical ACASA structure, each of these source/sink contributions is assigned the proper vertical level, then the associated scalars and vector quantities are turbulently diffused between layers.

The present work simulates fluxes in a complex urban environment, starting from meteorological data and footprint morphology evaluation. These serve as inputs for the ACASA model that predicts fluxes of carbon dioxide, water vapour, and sensible heat. Predictions are then validated against experimental eddy-covariance data, focusing on impact of the newly introduced anthropogenic fluxes parameterization. Model performance improvements are objectively assessed through evaluation of widely accepted statistical parameters. Results are further compared with existing scientific literature.