The twentieth century contiguous US temperature changes indicated by daily data and higher statistical moments

Abstract

The daily surface air temperature data are used to assess the climate changes of the contiguous United States during the period of 1901 to 2000. The assessment is made through the first four statistical moments of the daily maximum, minimum, and mean temperature anomalies, the linear trends of the moments, and the changes of the anomalies’ probability density functions. The results on the first moment, i.e., the mean, are compared with the existing ones in terms of intra-annual means and their linear trends. Our first moment results agree with known ones and demonstrate a decrease from the 1930s to the 1960s and an increase from the 1970s to 2000. The temperature fluctuation is the smallest in the 1960s among the decades from 1931 to 2000. The trends of the higher (second-, third- and fourth-order) moments of the mean, maximum, and minimum surface air temperatures are calculated for the periods 1901–2000, 1910–1945, 1946–1975, and 1976–2000. The results show a decreasing trend of the second- and third-order moments of all the temperatures. The fourth-order moments of the mean and maximum surface air temperatures have increasing trends, but that of the minimum surface air temperature has a decreasing trend. The seasonal histograms of the mean, maximum, and minimum surface air temperatures are calculated for the three periods 1910–1945, 1946–1975, and 1976–2000 for the stations which have the largest trend of maximum daily surface air temperature. An obvious change has been identified in the probability density functions. Among the changes of statistical parameters, the ones for the minimum temperature are larger than those for the maximum and mean temperatures.