Trend and Change point Analysis of Extreme Temperature over India using Non-Parametric Methods and Empirical Mode Decomposition
Drisya S Dharan1, Anuja PK2, Govindan Unnithan3, Anand Vishnu B4, Adarsh S5
1Drisya S Dharan, TKM College of Engineering, Kollam India.
2Anuja PK, TKM College of Engineering, Kollam  India.
3Govindan Unnithan, TKM College of Engineering, Kollam India.
4Anand Vishnu B, TKM College of Engineering, Kollam India.
5Adarsh S, Assistant Professor, TKM College of Engineering, Kollam India.
Manuscript received on 05 December 2018 | Revised Manuscript received on 19 December 2018 | Manuscript published on 30 December 2018 | PP: 16-19 | Volume-8 Issue-2C, December 2018 | Retrieval Number: 100.1/ijeat.ICID-2018_EE_306/
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: This paper presents the trend and change point analyses of extreme temperature (Tmax and Tmin) datasets of annual and seasonal series of seven homogeneous regions of India for the period 1901-2007. The study used Mann-Kendall (MK) test for detection of trend, Cumulative Sum (CUSUM) test for detection of change point and Empirical Mode Decomposition (EMD) for extracting non-linear trend of all the 70 time series. The results of MK test showed that at all the seven regions except North-West, Tmax series of winter season showed a significant increase at 5 % significance level. The MK test detected a significantly increasing trend on annual and all seasonal Tmax series of West Coast (WC) and North East (NE) regions. The CUSUM test detected a change point within 1975-77 for minimum temperature series of East Coast (EC) region for all the seasons except monsoon, which is in agreement with the well debated climate shift of 1976-77 period. The test detected a change point in 1950 for Tmax series of winter season in all homogeneous regions except northwest (NW). The study also found that change point year estimated in the non linear trend fitting by EMD may differ from that based on statistical estimations.
Keywords: Trend, Change Point, Non-linear, Temperature.
Scope of the Article: Probabilistic Models and Methods