Classification of synoptic weather types using the self-organising map and its application to climate and air quality data visualisation

The classification and visualisation utility of the self-organising map (SOM) method was explored in the New Zealand context using the NCEP/NCAR geopotential height reanalysis and local (Auckland) meteorological and air quality data. A new synoptic classification was derived from the geopotential height data for the New Zealand region, consisting of 25 types that are self-organised (topologically-ordered) on the SOM plane. The classification has not only reproduced the typical synoptic types previously identified in the literature, but also provided an opportunity to visualise the evolution of eastward-migrating synoptic systems over New Zealand. The topologically-ordered display of synoptic types on the SOM plane facilitated visualisation and identification of the synoptic types/local meteorology/air quality relationships in urban Auckland under a holistic framework. In particular, the projection of ozone (O3) and NOx (NO and NO2) data on the SOM plane provides important insights into how synoptic systems affect local NOx and O3 concentrations in urban Auckland. For example, the downward transport (from the upper troposphere/stratosphere) and mixing-in of O3 under cyclonic conditions and the advection of clean air from open ocean waters under blocking states are two of the main synoptic-scale mechanisms that modulate daily highs and lows of O3 and NOx in regional Auckland, bearing in mind that the coupling of local emissions, chemical processes and local meteorology is the major determinant of local air pollution. The results from this study are useful for air quality management in Auckland or similar regions, and also provide a basis for a more comprehensive assessment of the impact of weather and climatic conditions on the quality of the regional airshed.