Spatio-temporal long-term (1950–2009) temperature trend analysis in North Carolina, United States

Abstract

This study analyzed long-term (1950–2009) annual and seasonal time series data of maximum and minimum temperature from 249 uniformly distributed stations across the State of North Carolina, United States. The Mann-Kendall and Theil-Sen approach were applied to quantify the significance and magnitude of trend, respectively. A pre-whitening technique was applied to eliminate the effect of lag-1 serial correlation. For most stations over the period of the past 60 years, the difference between minimum and maximum temperatures was found decreasing with an overall increasing trend in the mean temperature. However, significant trends (confidence level ≥ 95 %) in the mean temperature analysis were detected only in 20, 3, 23, and 20 % of the stations in summer, winter, autumn, and spring, respectively. The magnitude of the highest warming trend in minimum temperature and the highest cooling trend in maximum temperature was +0.073 °C/year in the autumn season and −0.12 °C/year in the summer season, respectively. Additional analysis in mean temperature trend was conducted on three regions of North Carolina (mountain, piedmont, and coastal). The results revealed a warming trend for the coastal zone, a cooling trend for the mountain zone, and no distinct trend for the piedmont zone. The Sequential Mann-Kendall test results indicated that the significant increasing trends in minimum temperature and decreasing trend in maximum temperature had begun around 1970 and 1960 (change point), respectively, in most of the stations. Finally, the comparison between mean surface air temperature (SAT) and the North Atlantic Oscillation (NAO) concluded that the variability and trend in SAT can be explained partially by the NAO index for North Carolina.