Project topics and materials
Get Complete Project Material File(s) Now! »
Classification of synoptic types and identification of COLs
The focus of this paper is to relate interannual variability in seasonal rainfall to synoptic types. Synoptic types that represent archetypical atmospheric states were classified by application of the SOM technique (Kohonen 2001). A 35-node SOM was developed from daily sea-level pressure (SLP) anomaly fields derived from the daily average SLP fields from the National Centers for Environmental Prediction (NCEP) reanalysis data (Kalnay et al. 1996) for the region bounded by 45°S-32.5°S and 10°E-40°E. The SLP anomaly fields are effectively gradient fields that drive circulation (Scheunemann et al. 2009). Over the Cape south coast, low-level circulation is important to facilitate moisture advection from the surrounding ocean (Rouault et al. 2002; Singleton and Reason 2006, 2007a) – hence the choice of the SOM domain. A full description on the development of the SLP anomaly SOM can be found in Engelbrecht et al. (2015). In this paper, the composite maps of the 850 hPa geopotential height fields based on the SLP anomaly SOM are employed to represent the archetypical atmospheric states. The decision to employ the 850 hPa geopotential height fields to present the synoptic types in this study is to allow for presenting the synoptic types over a domain extending further northwards. The 850 hPa pressure level takes the typical height of 1,500 m of the South African plateau into consideration. The presentation of the synoptic types over a domain that includes a larger part of the subcontinent can complement the description of the synoptic types as the circulation over South Africa and surrounding oceans normally consist of segments of several circulation types. It can be noted that the synoptic types identified by the SLP anomaly SOM and those represented by the 850 hPa geopotential height composite maps over the larger domain, are consistent as the 850 hPa composite maps are derived from the daily entries in each node of the SLP anomaly SOM. NCEP reanalysis data (Kalnay et al. 1996) are utilized to identify and track COLs over the period 1979-2011. COLs have a typical length scale of 1,000 km (Singleton and Reason 2007b) and are therefore well resolved by the 2.5° resolution of the NCEP data. A COL is defined as a cold cored closed-low at 500 hPa that is displaced from the westerly wind regime (Favre et al. 2012). The daily-average geopotential height and temperature fields at 500 hPa are utilized for identifying and tracking COLs in the domain bounded by 40°S-20°S and 10°E-40°E. All the COLs that occurred for at least 24-h over this domain are considered in this study. Firstly, closed-lows are identified by locating geopotential minima in a procedure where the geopotential at each grid point is compared to the geopotential values of the square of eight surrounding grid points on the latitude longitude grid. After closed-lows have been identified in the time series of 500 hPa geopotential fields, tracks are constructed by identifying the geopotential minima at time step t+1 nearest to the geopotential minima at time step t. The distance between the closed-lows at time step t and time step t+1 needs to be <1,000 km in order to secure a sound mean daily speed of the potential COL. The mean daily speed of a COL in the South African region does not exceed 42 km/h (Favre et al. 2012). Any geopotential minima can only be used in a single track. From this closed-low track dataset, tests described in Favre et al. (2012) to ensure that tracks are of extra-tropical origin, are detached from the westerlies and possess a cold-core are employed. Rainfall associated with cold-cored systems occurs mainly some hundreds of kilometres to the northeast, east and southeast of the centres of these systems. From the constructed COL dataset for the period 1979-2011, all the COLs that occurred west of 32.5°E, following Favre et al. (2013), were considered to be potentially responsible for rainfall over the region. Such COLs associated with rainfall over the region, at least at a single station, were defined as rainfall producing COLs.
Grouping of similar synoptic types into main circulation types
The synoptic types identified by the SOM consist of various configurations of the main circulation types. Each of the SOM nodes is classified into one of the main circulation types – following the grouping performed in Engelbrecht et al. (2015). These groups represent circulation types representative of troughs southwest of the subcontinent, troughs southeast of the subcontinent, ridging high pressure systems, ridges east/southeast of the subcontinent, tropical-temperate troughs and weak synoptic flow. The groups of synoptic types used here fall within the subdivisions of circulation types (anticyclones, cyclones, ridges, troughs and zonal flow) identified and described by Taljaard (1995). Troughs southwest or southeast of the subcontinent fall within the west wind trough circulation type (e.g. Taljaard 1995), with the distinguishing factor being the position of the westerly trough relative to the subcontinent. Cold fronts, occurring during winter and summer as well as leader fronts, a winter circulation type, typically fall within these two groups while ridging high pressure systems and ridges east/southeast of the subcontinent are variations of ridges as described by Taljaard (1995). Tropical-temperate troughs, although part of the trough family, are considered separately as these systems are captured explicitly in the SOM (e.g. Tozuka et al. 2014) and due to its importance with regard to rainfall contribution over South Africa (Harrison 1984; Washington and Todd 1999; Hart et al. 2013). COLs are part of the cyclone/low pressure family (Taljaard 1995) and co-occur with ridging high pressure systems (Katzfey and McInnes 1996; Favre et al. 2012; Engelbrecht et al. 2015) and tropical-temperate troughs (Hart et al. 2013; Engelbrecht et al. 2015). However, due to the infrequent occurrence of COLs (Engelbrecht et al. 2015), these systems are difficult to explicitly be captured in the SOM. COLs are therefore considered with ridging high pressure systems and tropicaltemperate troughs in the analysis presented in this paper, unless where it is indicated that COLs are considered explicitly as identified by the objective COL identification and tracking algorithm (see section 2.2). The group of similar synoptic types termed weak synoptic flow is similar to the zonal flow circulation type identified by Taljaard (1995). However, a difference may be that the circulation type zonal flow described by Taljaard (1995) can be associated with weak or tight pressure gradients, whereas the circulation type weak synoptic flow in this study is defined to be characterized by weak pressure gradients over the Cape south coast region.
Classification of the synoptic types into groups is feasible by visual inspection of the 850 hPa geopotential height node composites (typically as would be performed by an experienced weather forecaster), but introduces the challenge of objectivity due to the circulation over the study domain not normally characterized by a single circulation type. Classification of the synoptic types into groups representative of troughs southwest of the subcontinent, troughs southeast of the subcontinent, ridging high pressure systems, ridges east/southeast of the subcontinent, tropical-temperate troughs and weak synoptic flow was therefore achieved by careful consideration of the SOM node anomaly fields (850 hPa geopotential height composites) relative to climatology (not shown) and a hierarchical clustering method, Ward’s minimum variance method (Wilks 2011). Ward’s method was applied to the SOM nodes based on the SLP anomaly fields to aid in grouping of the synoptic types into the main circulation types.
Chapter 1: Introduction
1.1 Background
1.2 Research problem
1.3 Aim and objectives of the study
1.4 Thesis outline
1.5 References
Chapter 2: A synoptic decomposition of rainfall over the Cape south coast of South Africa Preface
2.1 Introduction
2.2 Data and methodology
2.3 Results
2.4 Conclusions
2.5 Acknowledgements
2.6 References
Chapter 3: Interannual variability of seasonal rainfall over the Cape south coast of South Africa and synoptic type association
Preface
3.1 Introduction
3.2 Data and methodology
3.3 Results
3.4 Discussion and conclusions
3.5 Acknowledgements
3.6 References
Chapter 4: Seasonal predictability of intraseasonal synoptic type variability over the Cape south coast of South Africa by making use of the Met Office Global Seasonal Forecast System 5
Preface
4.1 Introduction
4.2 Data and methodology
4.3 Results
4.4 Conclusions
4.5 Acknowledgments
4.6 References
Chapter 5: Summary and Conclusions
