Isotopic Disentanglement of Moisture Sources: Insights into the Precipitation Histories of Mountain River Systems

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Application of stable isotopes in the fingerprinting of moisture sources has its perks and falls. While we can establish with some degree of accuracy the process and provenance of a moisture parcel, our assumptions largely rely on the Rayleigh distillation model and the isotopic end-member parametrization of different phases of water. To this end, the potential for discriminating water based on 18O, D and d-excess values become limited. The isotopic systematics involved in the water cycle i.e. from ocean to atmosphere to land and back to the ocean is fairly simple if each stage of fractionation can be identified and characterized. The water cycle progresses in a binary fashion where local environmental variables determine the evolution of the daughter isotopes. We demonstrate a physics-based ML-powered architecture founded on satellite-derived and hydrological input parameters which mimic the water cycle at any spacetime coordinate to yield optimum isotopic signature of water from snow, rain, glacier melt, groundwater, and other sources. In addition, the architecture back propagates the isotopic composition of a parcel of water (e.g. in a river) to its corresponding sources (e.g. snowmelt + groundwater influx) and their relative contribution percentages, given the location and time of the sample. We derived this model for the Himalayan river basins, arguably one of the most complex climatological systems in the globe due to its intricate topography, multiple sources and types of precipitation, and anomalous cryospheric input. Model output identified the amount and source of moisture received in different parts of the mountain range from the westerlies, summer monsoon, and others with a striking accuracy of 95%.
Original languageEnglish
Title of host publicationAGU Fall Meeting Abstracts
Publication statusPublished - 2021

Fingerprint

Dive into the research topics of 'Isotopic Disentanglement of Moisture Sources: Insights into the Precipitation Histories of Mountain River Systems'. Together they form a unique fingerprint.

Cite this