1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
multi-agent simulation, urban climate, pedestrian behaviour, micro-modelling, biometeorology
Start Date
1-7-2002 12:00 AM
Abstract
The urban canopy layer is characterised by a specific climate which significantly differs from theaverage regional climate conditions. Especially inside urban structures such as street canyons, shopping mallsor public places, the local microclimate depends directly on the physical properties of the surrounding surfacesand objects, producing well-known effects such as wind speed decrease, local jets, increased turbulence orincreased thermal loads. These phenomena can have a strong influence on the comfort of pedestrians usingthese areas for different tasks of urban life or leisure activities. Assessing urban climate means to answer thequestion, how the citizens feel under the given circumstances and to decide, whether the urban design, theusage of urban structures and the needs of the pedestrians fit together or not. This complex assessment processrequires a queue of different methods starting with numerical climate models that provide the meteorologicalparameters ending up with a set of models that simulate the comfort level or the behaviour of citizens actinginside the structure.In this paper, the Multi-Agent model climBOT is presented which aims to simulate the interactions betweenlocal climate and pedestrian behaviour. The different modules needed for that task are outlined and the generalperception-decision approach for the agents is shown.
Multi-Agent Simulations as a tool for the assessment of urban microclimate and its effect on pedestrian behaviour
The urban canopy layer is characterised by a specific climate which significantly differs from theaverage regional climate conditions. Especially inside urban structures such as street canyons, shopping mallsor public places, the local microclimate depends directly on the physical properties of the surrounding surfacesand objects, producing well-known effects such as wind speed decrease, local jets, increased turbulence orincreased thermal loads. These phenomena can have a strong influence on the comfort of pedestrians usingthese areas for different tasks of urban life or leisure activities. Assessing urban climate means to answer thequestion, how the citizens feel under the given circumstances and to decide, whether the urban design, theusage of urban structures and the needs of the pedestrians fit together or not. This complex assessment processrequires a queue of different methods starting with numerical climate models that provide the meteorologicalparameters ending up with a set of models that simulate the comfort level or the behaviour of citizens actinginside the structure.In this paper, the Multi-Agent model climBOT is presented which aims to simulate the interactions betweenlocal climate and pedestrian behaviour. The different modules needed for that task are outlined and the generalperception-decision approach for the agents is shown.
