Study on the influencing factors of spring water in Jinan City based on wavelet transform method

In order to find out the influence of two main factors—precipitation and groundwater exploitation—on the gushing of spring in Jinan urban area, we selected the exploitation data of the western suburb in Jinan (the water source area of Emei Mountain, Dayangzhuang and Lashan), Jinan urban area (the water source area of Quancheng road, Jiefangqiao, Yinhuchi, and Linan) and the eastern suburb in Jinan (the water source area of Huaneng road). We also selected dynamic observation data of urban spring water level and precipitation during the same period. Based on cross wavelet change, wavelet coherence spectrum and multiple wavelet coherence spectrum, we analyzed the time-lag relationship, comprehensive impact and impact degree between atmospheric precipitation, groundwater exploitation and spring water in Jinan City, in order to provide a basis for the optimal allocation of groundwater development and utilization and the promotion of spring protection in this city.

The following conclusions are drawn in this study. (1) By cross wavelet transform, we analyzed the resonance period, significant period and phase relationship between the spring level and the influencing factors in the time-frequency domain. It is found that during the research period, there was a frequency domain period of about 1 year between spring water level, precipitation and exploitation quantity of groundwater. The urban spring level and precipitation passed the 95% red noise test and there was a frequency domain period of 0.69-1.16 a outside COI. In a hydrological year, the urban spring level lagged behind the precipitation for 133.22 days. When the urban spring level and groundwater exploitation passed the 95% red noise test and there was a frequency domain period of 0.82-1.23 a outside COI, the response time lag between the urban spring level and the western suburb and the eastern suburb of Jinan was 125.43 days and 83.85 days, respectively. There existed a hydraulic relationship between spring water in Jinan urban area and groundwater in the western suburb and the eastern suburb of this city. The cross wavelet transforms between urban spring level and the exploitation quantity in the western suburb of Jinan, and between the urban spring level and the exploitation quantity in the urban area also showed a dispersed low energy region of 0.39-0.58 a, and the 95% red noise test reflected the seasonal characteristics of the exploitation of Jinan water source. (2) We carried out the wavelet coherence analysis of urban spring water level, precipitation and exploitation quantity of groundwater to explore the change characteristics of urban spring water level and influencing factors in the whole period. The wavelet coherence average value (AWC) between the urban spring level and precipitation was 0.58. The AWC value between the urban spring level and the respective exploitation quantity in the western suburb, in the city area and in the eastern suburb is 0.47, 0.40 and 0.32. PASC was 13.90%, 16.81% and 10.09%, respectively. During the research period, both precipitation and karst groundwater exploitation exerted obvious effects on urban spring levels. In terms of groundwater exploitation in different locations, the urban exploitation exerted the greatest influence on the urban spring level, followed by the western suburb. The eastern suburb of Jinan was influenced least. (3) According to exploitation conditions during the research period, the combination of precipitation and exploitation quantity in the western suburb of Jinan, and its combination in the urban area were regarded as the biggest influencing factors for the change of spring water level. Under the influence of the three factors, MWC produced by the combination of precipitation and exploitation quantities in the western suburb and the urban area reached 0.91, and PASC increased by more than 5%. Under the combined action of the four influencing factors, MWC reached the maximum value of 0.94, while PASC decreased. One reason was that the PASC threshold with statistical significance increased after additional factors were added, and the other reason was that the overlapping effect caused by collinearity among influencing factors reduced the variance contribution of some factors. Therefore, the multivariate wavelet coherence spectrum showed that the combination of precipitation and the exploitation quantities in the western suburb and the urban area is the most influential factor for the change of urban spring water level under the exploitation conditions during the research period.