2.4 Real-Time Air Quality Assessment and Management: Cascading Models in a Web Based Implementation

摘要

For the assessment of regional to local air quality in urban and industrial areas, a system of coupled models (AirWare) that cover several levels of nesting from regional (several hundred kilometers) to city level and street canyons has been developed in EUREKA E!3266 WEBAIR. The main objectives are to support regulatory tasks, compliance monitoring, and public information, and to provide a reliable basis for the multi-criteria optimization of emission control strategies to effectively and efficiently meet a range of environmental standards. Examples from Tehran, Iran, and Seoul/Gyeonggi-do, South Korea are used to demonstrate the system operation and performance. We compare two very large cities with different physiographical and socio-economic structures, management objectives, and policy options, yet a common problem. A web based implementation where all functions are accessed by a standard web browser provides ease of use for all major actors and stakeholders including public information functions. The system includes daily forecasts (meteorology and air quality) for 3-7 days, and hourly now-casts with data assimilation. In parallel to forecast and now-cast runs, the system supports interactive scenario analysis for EIA and emission control optimization tasks. The simulation tools include MM5 and WRF for meteorological forecasts based on downscaling global (GFS) weather forecasts, and dynamic emission modeling. The models are linked to several data bases (monitoring data and emission inventory) and an embedded GIS. CAMx is used for photochemistry and particulates, AERMOD for conservative substances. A high-resolution convolution model with a near-field mixing zone is based on AERMOD as a computational kernel for city-wide line sources and local concentration gradients with allows for near real-time simulation of thousands of road segments. Additional models include a 3D FD code for street canyons with explicit building obstacles for selected hot-spots, and a stochastic implementation of CAMx for probabilistic ozone forecasts.