Identification of rainfall homogenous regions in Saudi Arabia for experimenting and improving trend detection techniques

Abstract

In Saudi Arabia, identifying homogenous zones based on rainfall patterns is critical for ensuring a predictable and stable water resource and agriculture management strategy. As a result, the present research aims to identify Saudi Arabia's homogeneous rainfall zones and examine rainfall patterns in these areas. By proposing a novel trend analysis technique with a particular graphical representation, this study utilises and compares the traditional Mann-Kendall (MK) test, modified MK test, and basic Sen-innovative trend analysis (ITA) method. Another approach is to use the Pettit change point test to objectively identify subcategories as "low" or "high." The applications are based on 40-year rainfall records from 22 Saudi Arabian meteorological sites. K-means clustering and hierarchical clustering on principle component analysis (HCPC) were used to find homogeneous areas. The results of the homogeneous region identification revealed that the research area is divided into three clusters, each with three distinct climatic characteristics. Cluster 1 contains eight stations, whereas clusters 2 and 3 each have seven. The results of trend identification utilising the MK, MMK, and ITA tests revealed that cluster 1 had a falling rainfall trend, whereas cluster 2 had a very minor decreasing and increasing rainfall trend. Cluster 2 can be thought of as a transition zone. Cluster 3 observed an upward trend in rainfall. While the proposed new form of ITA produced similar results with more detailed analysis such as change point-based high and low value identification, and magnitude of decreasing and increasing trend, the proposed new form of ITA produced similar results with more detailed analysis such as change point-based high and low value identification. This study will serve as a foundation for future work by scientists and planners on the identification of climatic zones, the development of trend detection techniques, and the formulation of water resource management strategies.

Keywords: Clustering; Hierarchical clustering on principle component analysis (HCPC); Homogenous regions; Mann–Kendall trend test; Sen’s innovative trend analysis (ITA); Trend analysis.