Modified singular spectrum analysis in identifying rainfall trend over Peninsular Malaysia

Shazlyn Milleana Shaharudin

Abstract

Identifying the local time scale of the torrential rainfall pattern through Singular Spectrum Analysis (SSA) is useful to separate the trend and noise components. However, SSA poses two main issues which are torrential rainfall time series data have coinciding singular values and the leading components from eigenvector obtained from the decomposing time series matrix are usually assesed by graphical inference lacking in a specific statistical measure. In consequences to both issues, the extracted trend from SSA tended to flatten out and did not show any distinct pattern. This problem was approached in two ways. First, an Iterative Oblique SSA (Iterative OSSA) was presented to make adjustment to the singular values data. Second, a measure was introduced to group the decomposed eigenvector based on Robust Sparse K-means (RSK-Means). As the results, the extracted trend using modification of SSA appeared to fit the original time series and looked more flexible compared to SSA.

References

[1] F.W. Githui, A. Opere, and W. Bauwens, “Statistical and trend analysis of rainfall and river discharge: Yala River basin, Kenya,” in Proc. International Conference of UNESCO, Netherlands, 2003.

[2] M. Khaleghi, H. Zeinivand, and S. Moradipour, “Rainfall and river discharge trend analysis: a case study of Jajrood watershed, Iran,” International Bulletin of Water Resources and Development, vol.2, no.3, 2014.

[3] RA. Mondal, S. Kundu, and A. Mukhopadhyay, “Rainfall trend analysis by Mann-Kendall test: a case study of North-Eastern part of Cuttack district, Orissa. International Journal of Geology, Earth and Environmental Sciences, vol. 2, no. 1, pp.70-78, Jan. 2012.

[4] T. Alexandrov, N. Golyandina, and A. Spirov, “Singular spectrum analysis of gene expression profiles of early drosophila embryo: exponential-in-distance patterns,” Research Letters in Signal Processing, vol. 2008, no. 2008, pp. 1-5, June 2008.

[5] MAR. Khan, D.S. Poskitt, “Moment tests for window length selection in singular spectrum analysis of short- and long-memory processes,” Journal of Time Series Analysis, vol.34, no.2, pp. 141-155, 2013.

[6] RO. Awichi, W. Muller, “Improving SSA predictions by inverse distance weighting,” REVSTAT, vol.13, no.1, pp. 105-119, 2013.

[7] N. Itoh, N. Marwan, “An extended singular spectrum transformation (sst) for the investigation of kenyan precipitation data,” Nonlinear Processes in Geo-physics, vol. 20, no. 4, pp. 467-481,2013.

[8] N. Golyandina, and A. Shlemov, “Variations of singular spectrum analysis for separability improvement: non-orthogonal decompositions of time series,” Statistics and Its Interface, vol. 8, no.3, pp. 277-294, Feb. 2014.

[9] H. Hassani, “Development of the theoretical and methodological aspects of the singular spectrum analysis and its application for analysis and forecasting of economics data”. Cardiff, Wales: Cardiff University, 2009.

[10] FJ. Alonso, DR. Salgado, J. Cuadrado,P. Pintado, “Automatic smoothing of raw kinematics signals using ssa and cluster analysis,” Euromech Solid Mechanics Conference Lisbon, Portugal, pp. 1-9, September 2009.

[11] YS. Maslennikova, VV. Bochkarev, “20 training algorithm for neuro-fuzzy based on singular spectrum analysis,” World Conference on Information Technology, Barcelona, Spain, pp. 606-610, November 2012.

[12] SM. Shaharudin, N. Ahmad, NH. Zainuddin, NS. Mohamed, “Identification of rainfall patterns on hydrological simulation using robust principal component analysis,” Indonesian Journal of Electrical Engineering and Computer Science (IJEECS),” vol. 5, no. 3, pp. 401-408, January 2017.

[13] SM. Shaharudin, N. Ahmad, Fadhilah Yusof, “Effect of window length with singular spectrum analysis in extracting the trend signal on rainfall data,”AIP Conf. Proc., vol. 1643, no. 321, pp. 321-326, 2015.

[14] AP. Putra, A. Buono, BP. Silalahi, “ Modeling singular value decomposition and k-means of core image in classification of potential nickel,” TELKOMNIKA Indonesian Journal of Telecommunication, Computing,Electronics and Control,” vol. 13, no.3, pp. 561-567, March 2016.

[15] Y. Kondo, Robustification of the Sparse K-Means Clustering Algorithm. West Mall, Vancouver: The University of British Columbia, 2009

[16] Y. Kondo, M. Salibian-Barrera, and R. Zamar, “A robust and sparse k-means clustering algorithm,”arXiv:1201.6082v1, pp. 1-20, Jan. 2012.

[17] N. Golyandina, and A. Korobeynikov, “Basic singular spectrum analysis and forecasting with R,” Computational Statistics and Data Analysis, vol. 71, pp. 934- 954, Apr. 2013.