4.2.2 The synoptic precipitation with typhoon impact
Compared with the existence of multiple types of isotopic variations in normal precipitation, the time-series of typhoon-related precipitation shows similar trend and can be divided into three main stages (Fig. 4). The low humidity in the first stage is probably due to influence of the frontal subtropical high pressure of the typhoon, which have a strong re-evaporation effect (Fig. 8), meanwhile the δ18O was the highest due to it being in the initial stage of precipitation (Xu et al., 2019). As the typhoon center approached, the typhoon rain brands gradually controlled the precipitation in the study area. The isotopic composition of water vapor was depleted due to the high precipitation amount (Fudeyasu et al., 2008), resulting in increase of atmospheric relative humidity and decrease of δ18O value. When the typhoon rain brands pass, the atmospheric relative humidity decreases again, and the residual water vapor with high δ18O value becomes the main source of moisture for precipitation, while evaporation also gradually increases, resulting in the δ18O values of precipitation increasing again. The results suggest that the moisture source which affected by the precipitation and re-evaporation in different stages of the typhoon are the main factor controlling the variation of δ18O values during precipitation in the study area.
[Insert Figure 8]
Although the time-series patterns of precipitation are the same, there are large differences between the isotopic values of precipitation from different typhoons, which may be influenced by the season of typhoon occurrence, the source of moisture, and other factors (Table 1) (Mullaugh et al., 2013; Xu et al., 2019; Zhang et al., 2015b). Typhoons in the western Pacific Ocean occur mainly during the summer months when the East Asian Summer Monsoon are prevalent (e.g., Typhoon Maria, TD, and Typhoon Bailu in this study), but occasionally typhoons occur during the withdraw of the summer monsoon period or even the beginning of the winter monsoon period (e.g., Typhoon Yutu in this study), which may lead to differences in the isotopic composition of precipitation between them. The moisture of typhoon-related precipitation mainly originated from oceanic air masses (Fudeyasu et al., 2008), but there are significant differences in moisture sources due to the differences in the origin, route, and intensity of different typhoons (Fig. 9). The backward trajectory results show that the moisture source of different typhoon-related precipitation varies greatly, resulting in different values of δD, δ18O, and D-excess (Table 1, Fig. 9). The moisture trajectory during Typhoon Maria, TD, and Typhoon Bailu showed obvious characteristics of anticlockwise spiral (Fig. 9 a, b, and d), especially the moisture at 2000 m heights, which was controlled by the cyclone structure of typhoon. Compared with TD and Typhoon Bailu, the moisture trajectories at 1000 m and 2000 m heights during Typhoon Maria suggested moisture sources from local region, so the D-excess value of Typhoon Maria related precipitation was significantly higher than the others (Table 1). The moisture source of Typhoon Yutu related precipitation was significantly different from the other three typhoon-related precipitations (Fig. 9). The moisture at 2000 m height mainly originated from the oceanic air mass (west Pacific), which basically agreed with the path of typhoon, whereas the moisture at 500 and 1000 m heights originated from the continental air mass (Mongolian region) and underwent long-distance transport, which may dominant by the winter monsoon (Fig. 9 c). It is likely that moisture carried by typhoon (2000 m height) should be much larger than that by general air transport (500 and 1000 m heights) in the typhoon-related precipitation (Xie et al., 2011). As a results, the D-excess value of the Typhoon Maria related precipitation mainly shows a clear ocean-source character. At the same time, the moisture of this Typhoon Yutu related precipitation was influenced by the mixing of land-source moisture (500 and 1000m heights), so the δD and δ18O value was relatively higher than that of other typhoon-related precipitations.
[Insert Figure 9]
Previous studies have shown that the factors affecting the δD and δ18O of atmospheric precipitation include two main aspects. The first one is climatic background of the regional environment, including the source and nature of moisture and all changes in hydrogen and oxygen isotopes during water vapor production and transport (Zwart et al., 2016; Quezadas et al., 2021). The other one is local geographic factors, including various meteorological parameters during precipitation, as well as local latitude and altitude (Langebroek et al., 2011; Xie et al., 2011). The effects of regional climate on δD and δ18O are synchronous and cause equilibrium fractionation of isotopes in moisture, thus having a significant control on the D-excess of atmospheric precipitation (Guo et al., 2021). The correlation analysis results show that the D-excess of typhoon-related precipitation is significant correlated with wind direction (Table 2), which may be due to the influence of the cyclonic structure of the typhoon on the transport of moisture for typhoon-related precipitation. Local geographic factors mainly affect the dynamic fractionation of isotopes during precipitation, such as re-evaporation, which leads to higher δD and δ18O of precipitation and is influenced by meteorological factors such as temperature, pressure and relative humidity during precipitation. Thus, the δD and δ18O of precipitation show significantly correlation with temperature, relative humidity, and pressure in this study (Table 2, Fig. S1). There is a significant negative correlation between δD and δ18O of precipitation and temperature, showing an inverse temperature effect, mainly related to the seasonal variation of isotopic values, which is consistent with the results of monthly average result (Fig. 6). It is noteworthy that such correlations are strengthened in typhoon-related precipitation in the Pearson’s correlation analysis (Table 2 c). However, the linear regression results indicate that the correlation between δ18O of typhoon-related precipitation and temperature is a spurious signal. When the precipitation samples collected during Typhoon Yutu (which occurred in November and had significantly lower temperatures than the rest of the typhoon occurrences) are removed, there is no significant correlation between δ18O value and temperature for the remaining samples (Fig. S1 c). Therefore, the Pearson’s correlation coefficient alone cannot be simply considered when analyzing the relationship between δ18O value and meteorological parameters. The lack of correlation between δ18O value of typhoon-related precipitation and temperature may be due to the short duration of typhoon-related precipitation and the small temperature variation during the same precipitation event, as well as the fact that typhoons mostly occur in summer and the temperature difference between different typhoon-related precipitation events is small, which eventually leads to no significant correlation between δ18O value and temperature. Both Pearson’s correlation coefficients and linear regression results indicate a significant correlation between δ18O value of typhoon-related precipitation and relative humidity and pressure (Table 2, Fig. S1), which may be related to the significant re-evaporation effect in the first and third stages of typhoon-related precipitation. Absolutely, due to the limited sample size for typhoon-related precipitation, the results of correlation analysis may have some deviations, but it can still be a good reference for related studies. In the future work, we will collect more samples and data for a more detailed and in-depth study.
[Insert Table 2]