Current studentsExisting staffContact us
 

High resolution monthly precipitation isotope estimates across Australia from machine learning

Simple item page
dc.contributor.authorFalster, G.
dc.contributor.authorAbramowitz, G.
dc.contributor.authorHobeichi, S.
dc.contributor.authorHughes, C.
dc.contributor.authorTreble, P.
dc.contributor.authorAbram, N.J.
dc.contributor.authorBird, M.I.
dc.contributor.authorCauquoin, A.
dc.contributor.authorDixon, B.
dc.contributor.authorDrysdale, R.
dc.contributor.authorJin, C.
dc.contributor.authorMunksgaard, N.
dc.contributor.authorProemse, B.
dc.contributor.authorTyler, J.J.
dc.contributor.authorWerner, M.
dc.contributor.authorTadros, C.V.
dc.date.issued2026
dc.description.abstract<jats:p>Abstract. The stable isotopic composition of precipitation (δ2HP, δ18OP; “water isotopes”) is a powerful tool for tracking water through the atmosphere, as well as fingerprinting land-surface water masses and identifying water cycle biases in isotope-enabled climate models. Water isotopes also underpin our understanding of multi-decadal to multi-centennial water cycle variability via their retrieval from palaeoclimate archives. Water isotopes thereby increase our understanding of past and present – and hence future – water cycle variability. Understanding the drivers of spatial and temporal water isotope variability is a critical first step in applying these tracers for a better understanding of the water cycle. However, water isotope observations are sparse in both space and time. Here we develop and apply a machine learning (random forest) approach to predict spatially continuous monthly δ2HP and δ18OP across the Australian continent at 0.25° resolution from 1962–2023. We train the random forest models on monthly δ2HP (n=5199) and δ18OP (n=5217) observations from 60 sites across Australia. We also predict the deuterium excess of precipitation (dxsP, defined as δ2HP-8×δ18OP). Out-of-sample δ2HP and δ18OP prediction skill is high both geographically and temporally. Skill is slightly lower for the secondary parameter dxsP, likely reflecting the larger reliance of spatio-temporal dxsP variability on moisture source conditions. The random forest models accurately capture both the seasonal cycle of precipitation isotopic variability and long-term annual-mean precipitation isotopic variability across the continent, and outperform estimates from an isotope-enabled atmosphere general circulation model over an equivalent time period. We show that spatio-temporal variability in precipitation amount, precipitation intensity, and surface temperature are particularly important for monthly δ2HP and δ18OP variations across the continent, with local surface pressure also important for dxsP. Drivers of site-level δ2HP, δ18OP, and dxsP are more varied. Overall, the new random forest modelled dataset reveals clear spatial and temporal variability in δ2HP, δ18OP, and dxsP across the Australian continent over the past decades – providing a robust foundation for hydrology, ecology, and palaeoclimate research, as well as an accessible framework for predicting water isotope values in other locations.</jats:p>
dc.description.statementofresponsibilityGeorgina Falster, Gab Abramowitz, Sanaa Hobeichi, Catherine Hughes, Pauline Treble, Nerilie J. Abram, Michael I. Bird, Alexandre Cauquoin, Bronwyn Dixon, Russell Drysdale, Chenhui Jin, Niels Munksgaard, Bernadette Proemse, Jonathan J. Tyler, Martin Werner, and Carol V. Tadros
dc.identifier.citationHydrology and Earth System Sciences (HESS), 2026; 30(2):289-315
dc.identifier.doi10.5194/hess-30-289-2026
dc.identifier.issn1027-5606
dc.identifier.issn1607-7938
dc.identifier.orcidFalster, G. [0000-0001-8567-7413]
dc.identifier.orcidTyler, J.J. [0000-0001-8046-0215]
dc.identifier.urihttps://hdl.handle.net/2440/149625
dc.language.isoen
dc.publisherCopernicus Publications
dc.relation.granthttp://purl.org/au-research/grants/arc/DE250100071
dc.relation.granthttp://purl.org/au-research/grants/arc/CE170100023
dc.relation.granthttp://purl.org/au-research/grants/arc/CE230100012
dc.relation.granthttp://purl.org/au-research/grants/arc/CE170100015
dc.relation.granthttp://purl.org/au-research/grants/arc/FL140100044
dc.relation.granthttp://purl.org/au-research/grants/arc/FT230100648
dc.rights© Author(s) 2026. This work is distributed under the Creative Commons Attribution 4.0 License.
dc.source.urihttps://doi.org/10.5194/hess-30-289-2026
dc.subjectsotope estimates
dc.subjectAustralia
dc.subjectmachine learning
dc.titleHigh resolution monthly precipitation isotope estimates across Australia from machine learning
dc.typeJournal article
pubs.publication-statusPublished

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
hdl_149625.pdf
Size:
12.12 MB
Format:
Adobe Portable Document Format
Description:
Published version
Download

Collections

Research Outputs