Time series trend analysis of rainfall in last five decades and its quantification in Aseer Region of Saudi Arabia

Abstract

Investigating the spatiotemporal variation of meteorological parameters in the context of changing climate, particularly in countries where rainfall is sparse and agriculture is predominant contributor to economy, is vital to assess climate-induced changes and plan necessary adaptation strategies. To gain insight, trend analysis has been employed to inspect and quantify the change of precipitation in the highland of Aseer region of Saudi Arabia. Forty-eight-year long historical rainfall data series (1970–2017) from 30 rain gauges scattered across the study region was used for the analysis. Considering high spatial and temporal variability of rainfall data, the temporal scale chosen for the study was annual, rainy season (Mar–Jun), and bi-monthly. The trend was investigated using Mann–Kendall (MK) test. For serially correlated rainfall data series, modified Mann–Kendall (MMK) was applied for trend detection. Two variants of modified Mann-Kendall test were used in this study, namely, modified Mann–Kendall test using variance correction approach (mk1ylag) and bootstrapped Mann–Kendall Trend (pbmk) test. For quantification of trend magnitude, Theil–Sen approach (TSA) was used for calculation of Sen’s slope. Further, the abrupt change year detection in the trend was carried out using sequential Mann–Kendall (SQMK). All tests were performed using R-script. The application of abovementioned statistical test has shown statistically significant decreasing trend across majority of the stations at all temporal scale. It is observed that 10 stations located along the north-west to south-west boundary in mountainous region do not exhibit any significant trend. However, seven stations numbered as 14, 15, 38, 46, 514, 31 and 538 showed decreasing trends during most of the temporal scale. On analysing the trend magnitude of these seven stations, top three highest decreasing trend was found at station 31, 538, and 514 with trend magnitude as −9.325, −9.059, −7.516 mm/year, respectively. These three stations are located on north-west upper part of the study region. To understand the decade during which most of the stations experienced abrupt change, the period of 1970–2017 was broken down into four decades, 1980–1990; 1990–2000; 2000–2010; 2010–2017. Using SQMK, it was inferred that the decade of 2000–2010 witnessed drastic shift. Particularly for rainy season time series data, it is observed that the shift had occurred during the initial part of the decade 2000–2010 which can be attributed to increasing population and accelerated urbanisation. This result of declining and erratic trend of rainfall should be considered to design future strategies for agricultural sector, food security policy of the country and future water resource planning and management.