Insights into the implementation of synoptic weather-type classification using self-organizing maps: an Australian case study

Abstract

ABSTRACT:The two-fold utility (data projection and cluster analysis) of a two-phase batch self-organizing map (SOM) procedure (CP2) has been previously explored using the NCEP/NCAR geopotential height data for east Australia. That study focused on examining the performance of CP2 in comparison with a traditional cluster analysis procedure, CP1, for the purpose of synoptic typing. The present paper provides additional documentation on the implementation of CP2 for the same region, with broader considerations on the effect of SOM map size, seasonality, data standardization and the choice of neighbourhood functions.A total of 215 SOMs (classifications) were trained through CP2 with various data processing and parameter settings. The examination of these SOMs shows that the two-fold utility of CP2 leads to supplementary visualization of the dominant synoptic patterns over the study region. For SOMs of the same map size (i.e. number of synoptic types), cluster analysis via CP2 provides data groupings with relatively high accuracy and large separation but reduced level of pattern self-organization, while data projection via CP2 tends to create data groupings with a high level of pattern self-organization but reduced accuracy and separation. The choice of map size affects the accuracy, separation and self-organization of data groupings. As a compromise, a map size of 10-20 for cluster analysis and 20-30 for data projection is recommended for the study region. To account for the seasonality and latitudinal heterogeneity in the activity of synoptic systems, a relatively larger SOM size is needed to capture typical synoptic features prevailing in different seasons. Data standardization helps to provide a relatively balanced representation between larger-scale synoptic systems (e.g. anticyclones) and smaller-scale synoptic features (e.g. thermal lows), and also improves the level of pattern self-organization on the SOM across seasons. The additional documentation in this paper encourages a wider application of CP2 in environmental research.

Particulate Pollution in the Sydney Region: Source Diagnostics and Synoptic Controls

“…3) illustrate interactions between prevailing synoptic flows and local circulation features (e.g., sea breezes). These results agree well with Jiang et al (2013a) and Jiang et al (2014). Fig.…”

“…In this study, a synoptic type classification was derived through a two-phase batch SOM classification procedure (CP2) detailed in Jiang et al (2012Jiang et al ( , 2014. The SOM (Kohonen, 2001) is an unsupervised method for data classification and visualisation.…”

“…3, and the frequency of occurrence (percentage) by season is shown in Table 1. The resulting classification consisted of 20 types/nodes arranged in 5 rows and 4 columns, similar to Jiang et al (2014). Similar systems appear close together in the output space.…”

Visualising the relationships between synoptic circulation type and air quality in Sydney, a subtropical coastal‐basin environment

“…The study domain covers latitudes 15-50 ∘ S and longitudes 130-170 ∘ E at a 2.5 ∘ × 2.5 ∘ resolution, with Sydney located near the centre. Relatively larger variability in geopotential heights occurs at higher latitudes in the study domain (Jiang et al, 2015a). Hence, standardisation (i.e., taken as the difference from the mean divided by the standard deviation) was applied to the raw geopotential height time series for each grid point in order to obtain relatively more homogeneous data for the map classification.…”

On the use of self‐organizing maps for studying climate extremes

“…With only a few exceptions [e.g., Jiang et al , ], what appears less documented in the synoptic climatology literature is the influence of the end point of the radius parameter. At the point where the radius parameter exactly equals 1, the SOMs approach is shown to be equivalent to K‐means clustering as only the “winning node” is updated at each iteration (in contrast the winning node and surrounding nodes) [ Bação et al , ].…”