Temporal and spatial patterns of rainfall variability using nonparametric methods and wavelet transform: a case study of Sinai Peninsula

Abstract

Recognizing spatiotemporal patterns of rainfall trends has been of major concern for better planning and management of water resources. In this study, characterization of the variability in rainfall patterns was performed at 11 stations in Sinai Peninsula during the period 1990–2014 using the wavelet transform and descriptive statistics. Discrete wavelet transform technique was utilized in accordance with Mann–Kendall test to investigate trends and dominant periodicities associated with the seasonal and annual rainfall. Furthermore, the evaluation of significant periods of variations corresponding to the rainfall time series was done using the continuous wavelet transform. The results indicated that the rainfall was in general decreasing and declined trends were recognized in both winter and annual rainfall series; however, observed trends were statistically significant only in the northeastern part of the study area. Moreover, the results wavelet transformation analysis revealed an overall periodic variation in annual records accompanied by 4-–8-month signal that is more significant in influencing the identified trends; however, most of the periodic components were statistically insignificant. Based on the results of the spatial analysis, significant negative trends were mostly identified at the investigated stations. Overall, the winter and the annual rainfall records over Sinai Peninsula exhibited a considerable temporal variability accompanied by significant negative trends in northeastern part throughout the investigated time frame.