Air temperature change due to human activities in Taiwan for the past century

Abstract

The purpose of this study was to statistically examine air temperature change and to analyse the anthropogenic factors potentially influencing air temperature changes in Taiwan for the period 1897–2005 using homogenized temperature measurement time series. In this study, the standard normal homogeneity test was used to attain homogeneity of all air temperature series. To analyse air temperature data from 1897 to 2005, a time-series regression model and a product–moment correlation were used. The annual mean maximum temperatures (T_max), annual mean minimum temperatures (T_min) and the number of days in which the air temperature ≥ 30 °C increased significantly in most regions in Taiwan, and the rate of increase for T_min was higher than that of T_max. In other words, in the past century the night temperatures have become higher and the period of hot days in Taiwan have become longer. However, over the past century and before 1962, the trend of T_max in central Taiwan (i.e. Jihyuehtan and Taichung) decreased significantly, which is inconsistent with greenhouse gas warming and general warming observed in other sites in Taiwan. Furthermore, during the economic growth experienced between 1962 and 2005, the rates of increase for T_max and T_min in major cities with similar climate conditions (i.e. Tainan and Kaohsiung) were significantly different, as were the correlation coefficients between temperature changes and the number of various economic sectors in the cities. The results of these statistical analyses suggest that human activities strongly affected air temperature changes in these urban areas. The obvious lack of spatial homogeneity indicates that various non-greenhouse gas processes have played roles in climate change in Taiwan, and these factors must be considered when analysing long-term air temperature variations. Caution should be exercised when using grid data to assess temperature trends at the grid-box level. Copyright © 2009 Royal Meteorological Society