Understanding tropical convection through triple oxygen isotopes of precipitation from the Maritime Continent

• Main Authors: He, Shaoneng, Jackisc, Dominik, Samanta, Dhrubajyoti, Kho, Phyllis Yu Yi, Liu, Guangxin, Wang, Xianfeng, Goodkin, Nathalie Fairbank
• Other Authors: Asian School of the Environment
• Format: Article
• Language: English
• Published: 2021
• Subjects: Social sciences::Geography::Environmental sciences; Monthly Precipitation; Moisture Source Humidity; Stable Isotopes
• Online Access: https://hdl.handle.net/10356/149075
• Institution: Nanyang Technological University

Monthly precipitation samples from Singapore were collected between 2013 and 2019 for stable isotope analysis to further our understanding of the drivers of tropical precipitation isotopes, in particular, 17O-excess. d18O ranges from -11.34‰ to -2.34‰, with a low correlation to rainfall (r=-0.31, p=0.014), suggesting a weak amount effect. d-excess is relatively consistent and has an average value of 10.89±3.45‰. Compared to high-latitude regions, 17O-excess in our samples generally falls in a narrower range from 2 to 47 per meg with an average of 21±11 per meg. Moreover, 17O-excess shows strong periodic variability; spectral analysis reveals 3-month, 6-month and 2.7-year periodicities, likely corresponding to intra-seasonal oscillations, monsoons and the El Niño–Southern Oscillation (ENSO), respectively. In contrast, d-excess shows no clear periodicities. Although spectral analysis only identifies 6-month periodicity in the d18O time series, d18O tracks the Nino3.4 sea surface temperature variability; the average d18O value (-5.2‰) is higher during El Niño years than ENSO neutral years (-7.6‰). Therefore, regional convection associated with monsoons and ENSO has different impacts on d18O, d-excess and 17O-excess. 17O-excess and d-excess are anti-correlated, and do not relate to the relative humidity in moisture source regions. Extremely low humidity and drought conditions in moisture source regions would be required to account for high 17O-excess. Processes during transport and precipitation likely modify these two parameters, especially 17O-excess, which no longer record humidity conditions of moisture source regions. Our findings will be useful for further modeling studies to identify physical processes during convection that alter d-excess and 17O-excess.