Temporal Change Analysis Based on Data Characteristics and Nonparametric Test

Based on data characteristics and nonparametric test, a new statistical temporal change analysis approach is proposed. The new approach consists of data characteristics analysis, temporal change analysis (including both change point and trend analysis), and result interpretation. Data characteristics are firstly investigated, especially with respect to the assumptions of independence and normality. Then proper nonparametric methods are chosen based on the detected characteristics of the observed data to analyze change points and monotonous linear trend for each of the segments divided by the change points. To avoid shortcoming of the traditional approach of carrying out the trend analysis before change point analysis, it is proposed in this paper that change point detection be performed before trend analysis. At last, statistical analysis results are interpreted according to the physical mechanism of observations. As a study case, the proposed approach has been carried out on three annual discharge series of the Yangtze River at the Yichang hydrological station. The investigations of data characteristics show that the observed data do not meet the assumptions of being independent and identically Gaussian-distributed. So the nonparametric Pettitt’s test was adopted to detect abrupt changes in the mean levels, followed by trend analysis using the nonparametric Mann-Kendall (MK) test. Results indicate the proposed approach is both reliable and reasonable for the temporal change analysis.

[1]  L. Perreaulta,et al.  Bayesian change-point analysis in hydrometeorological time series . Part 1 . The normal model revisited , 2000 .

[2]  Chong-Yu Xu,et al.  Observed trends of annual maximum water level and streamflow during past 130 years in the Yangtze River basin, China , 2006 .

[3]  H. Theil.  A Rank-Invariant Method of Linear and Polynomial Regression Analysis , 1992 .

[4]  Francesco Serinaldi,et al.  Analyses of seasonal and annual maximum daily discharge records for central Europe , 2011 .

[5]  Sheng Yue,et al.  The influence of autocorrelation on the ability to detect trend in hydrological series , 2002 .

[6]  Demetris Koutsoyiannis,et al.  Nonstationarity versus scaling in hydrology , 2006 .

[7]  M. Kendall,et al.  Rank Correlation Methods , 1949 .

[8]  Bernard Bobée,et al.  Bayesian change-point analysis in hydrometeorological time series. Part 1. The normal model revisited , 2000 .

[9]  D. Cox,et al.  SOME QUICK SIGN TESTS FOR TREND IN LOCATION AND DISPERSION , 1955 .

[10]  David R. Brillinger,et al.  Time Series: Data Analysis and Theory. , 1982 .

[11]  Ximing Cai,et al.  Detecting gradual and abrupt changes in hydrological records , 2013 .

[12]  Bernard Bobée,et al.  Bayesian change-point analysis in hydrometeorological time series. Part 2. Comparison of change-point models and forecasting , 2000 .

[13]  G. Villarini,et al.  On the stationarity of annual flood peaks in the continental United States during the 20th century , 2009 .

[14]  E. Kahya,et al.  Trend analysis of streamflow in Turkey , 2004 .

[15]  Lihua Xiong,et al.  Effects of the Three Gorges Reservoir on the hydrological droughts at the downstream Yichang station during 2003–2011 , 2013 .

[16]  Chong-Yu Xu,et al.  Variability of Water Resource in the Yellow River Basin of Past 50 Years, China , 2009 .

[17]  S. Yue,et al.  The Mann-Kendall Test Modified by Effective Sample Size to Detect Trend in Serially Correlated Hydrological Series , 2004 .

[18]  R. Hirsch,et al.  Techniques of trend analysis for monthly water quality data , 1982 .

[19]  M. Kendall.  A NEW MEASURE OF RANK CORRELATION , 1938 .

[20]  Dingfang Li,et al.  Exploring the ability of the Pettitt method for detecting change point by Monte Carlo simulation , 2014, Stochastic Environmental Research and Risk Assessment.

[21]  Harry F. Lins,et al.  Streamflow trends in the United States , 1999 .

[22]  Hans von Storch,et al.  Monte Carlo experiments on the effect of serial correlation on the Mann-Kendall test of trend , 1992 .

[23]  H. B. Mann.  Nonparametric Tests Against Trend , 1945 .

[24]  Gregory J. McCabe,et al.  A step increase in streamflow in the conterminous United States , 2002 .

[25]  S. Yue,et al.  Canadian streamflow trend detection: impacts of serial and cross-correlation , 2003 .

[26]  R. Anderson.  Distribution of the Serial Correlation Coefficient , 1942 .

[27]  M. Slivitzky,et al.  Detection of changes in precipitation and runoff over eastern Canada and U.S. using a Bayesian approach , 1999 .

[28]  Richard M. Vogel,et al.  Trends in floods and low flows in the United States: impact of spatial correlation , 2000 .

[29]  Thomas C. Piechota,et al.  Changes in U.S. Streamflow and Western U.S. Snowpack , 2008 .

[30]  S. Becker,et al.  Precipitation, temperature and runoff analysis from 1950 to 2002 in the Yangtze basin, China / Analyse des précipitations, températures et débits de 1950 à 2002 dans le bassin du Yangtze, en Chine , 2005 .

[31]  M. Kendall.  Rank Correlation Methods , 1949 .

[32]  David R. Maidment,et al.  Handbook of Hydrology , 1993 .

[33]  R. Bartels.  The Rank Version of von Neumann's Ratio Test for Randomness , 1982 .

[34]  J. Neumann.  Distribution of the Ratio of the Mean Square Successive Difference to the Variance , 1941 .

[35]  W. L. Lane,et al.  Applied Modeling of Hydrologic Time Series , 1997 .

[36]  Witold F. Krajewski,et al.  Examining Flood Frequency Distributions in the Midwest U.S. 1 , 2011 .

[37]  P. Sen.  Estimates of the Regression Coefficient Based on Kendall's Tau , 1968 .

[38]  D. Burn,et al.  Detection of hydrologic trends and variability , 2002 .

[39]  N. Matalas.  Stochastic Hydrology in the Context of Climate Change , 1997 .

[40]  G. Villarini,et al.  Flood peak distributions for the eastern United States , 2009 .

[41]  A. Pettitt.  A Non‐Parametric Approach to the Change‐Point Problem , 1979 .

[42]  T. A. Buishand,et al.  Tests for detecting a shift in the mean of hydrological time series , 1984 .