Geographic Scale Compatibility Study Based on Process Simulation: A Case Study of Meteorological and Air Quality Process

Although multiscale data and models are taken into account to study geography, due to their scale dependence, the scale mismatching may cause adverse results. Thus, scale compatibility is becoming crucial to decode geographic process, especially, for the interactive geographic processes. This thesis focused on scale compatibility in geographic process with a case study of multiscale meteorological and air quality simulation in Hong Kong Special Administrative Region (HKSAR) and the Pearl River Delta (PRD) region.;Firstly, based on the conceptual definition of scale, this dissertation identifies four groups of scale compatibility in geographic process research: multiple process level, dimension level, type level and component level. By illustrating the different operational levels from the abstract to operational, the author proposes a procedure to implement scale compatibility with potential criteria.;Secondly, scale compatibility in the reproduction of meteorological process is investigated between multiscale digital elevation model (DEM) data and the Weather Research and Forecasting Model (WRF). The experiments show that: DEM data with 3 and 30 arc sec resolutions are relatively more compatible with the WRF model of 1 km resolution to reproduce the meteorological field over Hong Kong; and uncertainty arising from scale mismatching between DEM data and the model may account for 38% of the variance in certain meteorological variables (e.g., temperature). This case study not only helps to improve meteorological simulations by taking the issue of scale compatibility into account, but also explains the significance and implementation of scale compatibility in geographic process research.;Thirdly, this thesis utilizes multiscale meteorological models to study the scale compatibility between dynamic models and interested geographic process. We conduct validation through three steps: daily statistics, spatial comparison and time series. The results support the idea that CALifornia METeorological model (CALMET) is more compatible to reproduce the meteorology process in Hong Kong; however, the discrepancy between the WRF and CALMET is spatially heterogeneous, with larger improvement over the area of complex topography and land use. The results also give evidence about the cross-scale interaction to interpret multi-scale geographic process.;Fourthly, applying the above findings, this dissertation presents a multiscale Virtual Geographic Environments (VGE) system to integrate geographical analysis and multiscale models in a cross-platform and parallel manner. With database system and Linux-Apache-MySQL-Perl (LAMP) architecture, users can manage and retrieve modeling concerning both data and model parameterization to help them reach a consensus on the simulation results and share modeling knowledge. Scale compatibility among data, models and analysis is also considered in the system design. Aided with high-resolved and regulable emission inventory, such multiscale system enables the practical application for various scenarios. As a case study, the VGE is applied to simulate and analyze the SO2 concentration process and local contribution in HKSAR.;Achievements of this dissertation should greatly contribute to a better cognition of multiscale issues and scale compatibility concerning geographic process, and the VGE is expected to contribute to better understanding and management of air quality for both HKSAR and PRD.