Author(s): Hiroyuki Tosaka; Faizah Che Ros; Hidayah Basri; Lariyah Mohd Sidek

Linked Author(s):

Keywords: Rainfall time-series; Homogeneity test; Mann–Kendall; El Niño; La Niñ

Abstract: The Kelantan River Basin, situated in the northeastern Malaysian Peninsula, suffers serious flood/inundation damage, related to the northeast monsoon season (November–January), every few years. In this river basin, rainfall observation systems have been progressively developed since 1948, and long-term time-series data at distributed rainfall stations have been accumulated. This study firstly investigated the homogeneity of the accumulated time-series data for the purpose of constructing a reliable database for various hydrologic analyses. The homogeneity of rainfall time-series data was established using four absolute homogeneity tests: the Pettitt test, standard normal homogeneity test, Buishand range test, and von Neumann ratio test. It was found that among 50 rainfall stations within the river basin, 9 were flagged by the tests. Of these, inhomogeneous time-series data from four stations were omitted from further analysis. Secondly, using the homogenous time-series rainfall data, a trend analysis of annual rainfall variability was conducted by means of the Mann–Kendall (MK) test. To investigate long- and short-term characteristics of rainfall variability, three sampling methods of the MK test were applied: 30-year, sequentially increased, and 10-year moving segmented sampling. The MK test using the 30-year sampling showed a decreasing trend between 1957 and 1987, and an increasing trend between 1981 and 2011. The MK test using the sequentially increased sampling detected neither significant increasing nor decreasing trends through the same 70-year period. The MK test using 10-year moving segmented sampling clearly showed significant rainfall variability, which corresponded to the El Nino and La Nina events of 1972, 1982, 1988, the mid-1990s, and 2000–2004. Extending the plot, by comparing the turning points with the occurrence of El Nino and La Nina events, the results showed that the influence of the El Nino southern oscillation (ENSO) is the most likely to significantly influence rainfall trends in the Kelantan River Basin.

DOI: https://doi.org/10.1080/15715124.2015.1105233

Year: 2016