Extended Abstract (40K)11.8
An Intrigued Analysis to Quantify the Causes for Urban Heat Island by the Revised Architecturerban-Soi-Uuban Simultaneous Simulation Model, RIVISED–AUSSSM Part.2 Quantitative Analysis based on huge numerical experiments
Jun Tanimoto, Kyushu University, Fukuoka, Japan; and A. Hagishima and T. Katayama
By the revised AUSSSM, precisely explained on its modeling background in the Part 1 that is also presented at this conference, huge numerical experiments were done. These really challenging and great endeavors do target to identify a bunch of causes leading to Urban Heat Island (UHI) in quantitative format. The causes for UHI are seemed to be so complicated and so widely ranging that the design of experiments should be considered in a certain strategic scheme. We are very proud of the revised AUSSSM because its light-computing load but still keeping relatively high accuracy let us possible to observe lots of potential Factors existing behind UHI. In these numerical experiments we surely aimed at to present quantitative, useful, practical and precise output to build actual provisions against UHI. For an initial step, all possible significant causes on UHI that can be treated in the revised AUSSSM were selected through a primary screening. A set of those varies from Factors relating building design, HVAC system performance and other various anthropogenic heat generations in urban canopy space, up to, so-called urban planing Factors such as building allocation and vegetation in urban area. Actual procedure for the numerical experiments consists of two stages. As the primary experiment, so-called Variation Study was tried to pick up relatively significant Factors affecting on several assumed Characteristic Values. Let alone, the Variation Study is a technical term in the field of simulation engineering. For instance, a Variation Study having M Factors with N Levels for each Factor requires you to carry on 1+M*(N-1) times calculations, which is much less than the Perfect Experiment requiring NM times. The Characteristic Values were chosen as an air temperature increase in whole canopy space, another air temperature 1.2 m above the ground level, HVAC thermal requirements to extract from rooms, sensitive or latent heat flux at the top of Surface Boundary Layer (SBL) and more. And then, if you focus on the air temperature affecting on you in case of walking down the street, you can never neglect influences both the Factors of surface vegetation and building HVAC system. Whereas, if you think much about the temperature of whole canopy air, another significant Factor of the building density, roughness inspired by the buildings in SBL in other words, is coming up. The next stage in the procedure is a really huge experiment by the Design of Experiment Theory, where L81(340) was adopted as an Orthogonal Array. Relatively significant Factorial Effects for each Characteristic Value confirmed by the Variation Study were precisely investigated at this moment. Needless to say, a process based on the Design of Experiment Theory diminishes number of required simulation running as compared with the Perfect Experiment, even though almost equivalent outcome is at your hands. And this brought a significant result with the ANOVA Table and Table for Estimation of Factorial Effects. which could be regarded as a quantitative diagnosis for discussing UHI.
Session 11, Urban canopy layer: models
Thursday, 23 May 2002, 8:00 AM-1:30 PM
Previous paper Next paper