Premium
Annual precipitation in the northeast United States: Long memory, short memory, or no memory?
Author(s) -
Potter Kenneth W.
Publication year - 1979
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/wr015i002p00340
Subject(s) - autocorrelation , homogeneity (statistics) , series (stratigraphy) , hurst exponent , lag , precipitation , autoregressive model , partial autocorrelation function , long memory , time series , statistics , homogeneous , climatology , mathematics , environmental science , autoregressive integrated moving average , econometrics , geography , meteorology , geology , computer science , volatility (finance) , paleontology , computer network , combinatorics
Regional filtering and intervention analysis of nineteen 100‐year precipitation series from the northeast United States indicate that eight of these series are of doubtful homogeneity. Statistics of these series are generally distorted. Of particular significance is the inflation of the short‐lag autocorrelation function and Hurst coefficient. For the precipitation series which appear to be homogeneous the estimated autocorrelation functions and Hurst coefficients are remarkably consistent with a lag l autoregressive process, indicating that annual precipitation in the northeast United States is a short‐memory process. Based on these results, it is recommended that before long‐term memory is inferred from analyses of hydrologic series, the homogeneity of these series should be clearly established.