Redefining southern Australia’s climatic regions and seasons

Climate scientists routinely rely on averaging over time or space to simplify complex information and concisely communicate findings. Currently, no consistent definitions of ‘warm’ ‘cool’ seasons for southern Australia exist, making comparisons across studies difficult. Similarly, numerous climate in use either arbitrarily defined areas the Natural Resource Management (NRM) clusters perform spatial averaging. While NRM regions were informed by temperature rainfall information, they remain somewhat arbitrary. Here we weather type influence clustering methods quantitatively define climatic Australia. Three are explored: k-means two agglomerative methods, Ward linkage average linkage. K-means was found be preferred temporal clustering, while method clustering. For as a whole, cool season April–September warm October–March, though note that three-season split may provide more nuanced analysis. We also show different experience demonstrate changing types with seasons, which aid regionally seasonally specific Division into 15 shows localised agreement where distinct differences amounts exist. However, here better represent importance topographical aspect inland extent particular types. suggest these would analysis if widely adopted. A key requirement is simplification data sets both averaged subsets. This simplification, grouping like done number reasons scientific practical, including help understand patterns variability, underlying drivers trends weather, large concisely, reduce amount required processing (which becomes increasingly important higher resolution model output), simply draw physical boundary between other purposes, such flora fauna habitat analysis, appropriate agricultural practices water management.