将：从图5可知除夜间部分时段排水片区末端污水中34S在全监测周期中均高于自来水本底水平。基于34S研究结果的异常性进行了补充实验监测及分析工作。基于不同用水行为对硫稳定同位素的影响实验结果可知洗漱行为及餐厨行为整体致使34S的上升考虑到异常现象整体处于白天时段因此重点研究餐厨行为与排水片区末端污水的34S指标关联性并增加了各用水行为下水质监测点的密度以确保验证结果的高度代表性。于730-2

From Figure 5, it is known that except for some nighttime periods, the δ34S in the sewage at the end of the drainage area is higher than the background level of tap water throughout the monitoring period. Based on the abnormality of the δ34S research results, supplementary experimental monitoring and analysis work were carried out. The experimental results based on the influence of different water use behaviors on sulfur stable isotopes showed that washing and kitchen behaviors generally caused an increase in δ34S, and considering that the abnormal phenomenon was generally during the daytime, the correlation between kitchen behavior and δ34S index at the end of the drainage area was studied, and the density of water quality monitoring points under various water use behaviors was increased to ensure the high representativeness of the verification results. Instantaneous sampling was continuously performed every 30 minutes on the 1st restaurant, 2nd restaurant, and the M point at the end of the drainage area during 7:30-22:00, and samples were collected synchronously for sewage under washing, showering, toilet flushing, and dishwashing behaviors during the corresponding peak periods from 7:30 to 22:00. In addition, synchronous monitoring of tap water background quality was carried out.

The experimental and design experiments were separated by 27 days, and the comparison of δ34S monitoring data under different water use behaviors and at the end of the drainage area in the supplementary experiment is shown in Figure 6. According to the statistical results, the sewage δ34S under different water use behaviors in the supplementary experiment showed a similar change trend to the previous experiment: washing and kitchen behaviors caused an increase in δ34S, with an average increase ratio of 0.89% and 1.80%, respectively, compared to tap water, which was lower than the increase trend in the design experiment; toilet flushing behavior caused the largest decrease in δ34S, with an average decrease ratio of -17.93% compared to tap water; followed by showering behavior and dishwashing behavior, with average decrease ratios of -11.59% and -9.24%, respectively, compared to tap water. At the end of the drainage area, the average increase ratio of δ34S in sewage compared to tap water was 0.49%, which was also lower than the increase trend in the design experiment, preliminarily confirming the strong correlation between kitchen sewage and δ34S at the end of the drainage area. Based on the analysis of the dispersion degree of δ34S under different water use behaviors, the sewage δ34S under washing, showering, and toilet flushing behaviors showed excellent stability, with variation coefficients of 0.033, 0.046, and 0.035, respectively, during the monitoring period; the sewage δ34S under kitchen behavior showed excellent stability, with a variation coefficient of 0.056 during the monitoring period; the sewage δ34S under dishwashing behavior showed moderate to good stability, with a variation coefficient of 0.109 during the monitoring period. At the end of the drainage area, the sewage δ34S showed excellent stability, with a variation coefficient of 0.023 during the monitoring period.

To further evaluate the correlation between kitchen behavior and δ34S at the end of the drainage area, this study conducted monitoring and analysis of kitchen sewage from the 1st and 2nd restaurants and the sewage at the M point at the end of the drainage area. The monitoring results are shown in Figure 7. According to Figure 8, the δ34S change trend of kitchen sewage from the 1st and 2nd restaurants was not exactly the same. When the two trends were similar, the δ34S change trend at the end of the drainage area was synchronized with them; when the two trends were opposite, the δ34S change trend at the end of the drainage area was synchronized with the 2nd restaurant, but with a lag in synchronization. As shown in Figure 8, according to the correlation statistical results analysis, the δ34S index of the kitchen sewage from the 1st restaurant was negatively correlated with that of the sewage at the end of the drainage area in the afternoon, and positively correlated in the morning, noon, evening, and night. It maintained a negative correlation level higher than -0.400 from 7:30 to 11:00, and had a high positive correlation level higher than 0.900 from 11:00 to 15:30. After a short period of positive correlation, its level quickly returned to negative values, and maintained a negative correlation level close to -0.700 from 15:30 to 19:30. The dominant positive correlation period accounted for 31.03% of the total statistical period. The δ34S index of the kitchen sewage from the 2nd restaurant was positively correlated with that of the sewage at the end of the drainage area in the morning, evening, and night, and negatively correlated in the noon and afternoon. It maintained a positive correlation level close to 0.700 from 7:30 to 11:00, and had a high negative correlation level of -0.935 from 11:00 to 15:00. After a short period of negative correlation, its level quickly returned to positive values, and maintained a positive correlation level higher than 0.800 (part of which was higher than 0.9) from 16:30 to 22:00. The two dominant positive correlation periods accounted for 62.07% of the total statistical period. It was judged that the kitchen behavior of the 2nd restaurant had a greater influence weight on the δ34S in the daytime sewage at the end of the drainage area. Therefore, a comparison and analysis of the flow curves of the 1st and 2nd restaurants and the sewage at the end of the drainage area from 7:30 to 22:00 on the experimental day were carried out, and the results are shown in Figure 9. It can be seen from Figure 9 that the sewage flow of the 1st restaurant did not form a significant proportion during 7:30-20:30, and its cumulative proportion was only 8.68%; while the sewage flow of the 2nd restaurant had a large proportion weight compared to the sewage flow at the end of the drainage area, with a cumulative proportion of up to 69.73%. After 20:30, both the 1st and 2nd restaurants gradually lost their flow proportion advantages.

The above monitoring results verified the flow advantage of the 2nd restaurant during the daytime and nighttime periods and its highly correlated δ34S in kitchen sewage and sewage at the end of the drainage area, which explained the change trend of δ34S in sewage at the end of the drainage area: it was generally higher than the δ34S level of tap water throughout the monitoring period, and only experienced a decrease in the index compared to the background level during some nighttime period