Investigation of the trend and cycles using the long-term records of annual precipitation in Shandong, China

Abstract

Abstract The length of historical records is important for analyzing trends and multidecadal variabilities in precipitation. Conventional studies usually used historical precipitation data less than 60 years, which may cause inaccurate precipitation predictions. To better understand the effects of the study period on precipitation trends and cycles, the Mann–Kendall test and Sen's slope test were applied to analyze the trend and wavelet, while the multi-temporal analysis was used to study the cycles of long historical precipitation data at four rain gauge stations in Shandong Province, China. The analysis results using long-term records show an insignificant upward trend and a longer cycle in the annual precipitation, whereas a downward trend is found when using short-term records. Furthermore, it is found that selecting the length of the representative cycle derived from the precipitation data series as the time scale for the study proved to be more reasonable. A thorough consideration of the impact of cycles on trends should, therefore, be taken to facilitate a more accurate precipitation trend analysis.