Selection of homoclimates based on comparisons with single stations and using monthly rainfall and temperature data

Abstract

A method is proposed for the detection of homoclimates using the Canberra metric as a similarity coefficient to assess the large amounts of global monthly rainfall and temperature data, such as is available in Wernstedt (World Climatic Data). The method uses the dimensionless and ratio characteristics of the Canberra metric to compare stations on the basis of numerical climatic attributes. Ranking of the stations from the most similar to least similar to a reference station is carried out, enabling homoclimates to be identified. Analysis of the similarity of temperature, cumulative temperature, rainfall, cumulative rainfall and rainfall/temperature ratio attributes of homoclimates is also carried out. The method is used to compare 349 Mexican and Central American stations with five north Australian reference stations for both the wet season and the whole year. The necessity for defining homoclimates in Mexico and Central America arose from a recent plant collecting expedition to this region which had the objective of broadening the germplasm base of tropical forage legumes suitable for certain areas of northern Australia. The method maximizes the information content of the climatic data available and enables comparative climatic information for different regions of the world to be obtained quickly and cheaply.