Publication Date:
2006-02-14
Description:
Data of daily mean temperatures recorded at the Kennedy Center during the period of 1957-1977 were analyzed to forecast daily mean temperatures and their thirty-day moving averages for a period of ten to fifteen days in a given month. Since it is found that the standard deviation is linear in the mean, a logarithmic transformation of the data is used for finding an integrated moving average process IMA by the Box-Jenkins aproach. The first differences of the transformed data seem to fit a moving average model with parameter value 2, MA(2). The consideration of seasonality factor makes the fit worse.
Keywords:
METEOROLOGY AND CLIMATOLOGY
Type:
The 1981 NASA(ASEE Summer Fac. Fellowship Program; 13 p
Format:
text