Precipitation Isotopes New Zealand (PINZ): improvements in precipitation isoscapes with machine learning

A. F. Hill; A. McKenzie; B. D. Dudley; D. B. Nelson

doi:10.1071/EN24098

RESEARCH ARTICLE (Open Access)

Previous Next Contents Vol 22(3)

Precipitation Isotopes New Zealand (PINZ): improvements in precipitation isoscapes with machine learning

A. F. Hill

^A , A. McKenzie ^B , B. D. Dudley ^B ^* and D. B. Nelson ^C

+ Author Affiliations

- Author Affiliations

^A National Institute of Water and Atmospheric Research, 217 Akersten Street, Port Nelson, Nelson, 7010, New Zealand.

^B National Institute of Water and Atmospheric Research, 10 Kyle Street, Riccarton, Christchurch, 8011, New Zealand.

^C Department of Environmental Sciences-Botany, University of Basel, CH-4056 Basel, Switzerland.

^* Correspondence to: bruce.dudley@niwa.co.nz

Handling Editor: Zongbo Shi

Environmental Chemistry 22, EN24098 https://doi.org/10.1071/EN24098

Submitted: 24 October 2024 Accepted: 9 April 2025 Published: 14 May 2025

© 2025 The Author(s) (or their employer(s)). Published by CSIRO Publishing. This is an open access article distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND)

Abstract

Environmental context

Stable water isotopes δ¹⁸O and δ²H in atmospheric precipitation are valuable hydrologic tracers with varied applications, but measuring them over large scales is impractical. Precipitation isotope models (isoscapes) are instead needed to predict isotopes through space and time. Precipitation isotopes are affected by numerous variables ranging from orographic effects to phase changes to moisture mass origin, making them challenging to predict in complex geographies such as New Zealand. Improving data density together with recent advances in modelling approaches hold promise for improving accuracy of precipitation isoscapes in New Zealand and similar locations.

Rationale

Precipitation isotope models based on global networks, such as the Global Network of Isotopes in Precipitation (GNIP), lack sufficient spatial and temporal resolution for hydrological applications in New Zealand due to sparse data coverage. Models driven by regional datasets are needed to improve accuracy of isotope predictions to increase utility in environmental research applications.

Methodology

We develop Precipitation Isotopes New Zealand (PINZ) δ¹⁸O and δ²H isoscapes employing the Extreme Gradient Boosting (XGBoost) machine learning algorithm driven by a regional precipitation isotope database with high station density across New Zealand, and diverse climate and geographic predictors. We produce a national monthly precipitation isotope model at 5-km resolution. Model uncertainty associated with monthly predictions is evaluated using ‘leave one out’ cross validation (LOOCV, or jackknife) tests.

Results

PINZ shows significant spatial and temporal variability in isotope values, with notable patterns related to New Zealand’s complex topography and climatic influences. Model performance on test data is described by a root mean square error of 1.83‰.

Discussion

Although the model performs well overall, it exhibits limitations in predicting extreme precipitation δ¹⁸O values (most enriched and most depleted) and predicting δ¹⁸O of precipitation during periods of extreme weather conditions. Leveraging detailed local data and advanced algorithms, we demonstrate one way forward for improving isoscapes in complex environments.

Keywords: isoscape, isotope hydrology, machine learning, monthly isoscape, New Zealand, oxygen and hydrogen stable isotopes, precipitation isotope model, tree boosting algorithm.

References

Allen ST, Jasechko S, Berghuijs WR, Welker JM, Goldsmith GR, Kirchner JW (2019) Global sinusoidal seasonality in precipitation isotopes. Hydrology and Earth System Sciences 23, 3423-3436.
| Crossref | Google Scholar |

Baisden WT, Keller ED, van Hale R, frew RD, Wassenaar LI (2016) Precipitation isoscapes for New Zealand: enhanced temporal detail using precipitation-weighted daily climatology. Isotopes in Environmental and Health Studies 52, 343-352.
| Crossref | Google Scholar | PubMed |

Bartelink EJ, Chesson LA (2019) Recent applications of isotope analysis to forensic anthropology. Forensic Sciences Research 4, 29-44.
| Crossref | Google Scholar | PubMed |

Benettin P, Rodriguez NB, Sprenger M, Kim M, Klaus J, Harman CJ, van der Velde Y, Hrachowitz M, Botter G, Mcguire KJ, Kirchner JW, Rinaldo A, Mcdonnell JJ (2022) Transit Time Estimation in Catchments: Recent Developments and Future Directions. Water Resources Research 58, e2022WR033096.
| Crossref | Google Scholar |

Bowen GJ, Revenaugh J (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39, 1299.
| Crossref | Google Scholar |

Chen TCG (2016) XGBoost: a scalable tree boosting system. In ‘KDD ‘16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining’, 13–17 August 2016, San Francisco, CA, USA. (Eds B Krishnapuram, M Shah, ASmola, C Aggarwal, D Shen, R Rastogi) pp. 785–794. (Association for Computing Machinery: New York, NY, USA) 10.1145/2939672.2939785

Crawford J, Hughes CE, Parkes SD (2013) Is the isotopic composition of event based precipitation driven by moisture source or synoptic scale weather in the Sydney Basin, Australia? Journal of Hydrology 507, 213-226.
| Crossref | Google Scholar |

Dansgaard W (1964) Stable isotopes in precipitation. Tellus 16, 436-468.
| Crossref | Google Scholar |

Dudley BD, Marttila H, Graham SL, Evison R, Srinivasan M (2018) Water sources for woody shrubs on hillslopes: an investigation using isotopic and sapflow methods. Ecohydrology 11, e1926.
| Crossref | Google Scholar |

Dudley B, Hill A, Mckenzie A (2024) A seasonal precipitation isoscape for New Zealand. Journal of Hydrology: Regional Studies 52, 101711.
| Crossref | Google Scholar |

Friedman I, O’Neil JR (1977) Compilation of stable isotope fractionation factors of geochemical interest. Professional Paper 440-KK, US Geological Survey, US Department of the Interior. 10.3133/pp440KK

Gibson JJ, Holmes T, Stadnyk TA, Birks SJ, Eby P, Pietroniro A (2021) Isotopic constraints on water balance and evapotranspiration partitioning in gauged watersheds across Canada. Journal of Hydrology: Regional Studies 37, 100878.
| Crossref | Google Scholar |

Guerrieri R, Belmecheri S, Ollinger SV, Asbjornsen H, Jennings K, Xiao J, Stocker BD, Martin M, Hollinger DY, Bracho-Garrillo R (2019) Disentangling the role of photosynthesis and stomatal conductance on rising forest water-use efficiency. Proceedings of the National Academy of Sciences 116, 16909-16914.
| Crossref | Google Scholar | PubMed |

Hastie T, Tibshirani R, Friedman JH, Friedman JH (2009) ‘The Elements of Statistical Learning: Data Mining, Inference, and Prediction.’ (Springer)

Hill AF, Rittger K, Dendup T, Tshering D, Painter TH (2020) How important is meltwater to the Chamkhar Chhu headwaters of the Brahmaputra River? Frontiers in Earth Science 8, 81.
| Crossref | Google Scholar |

Hollister J, Shah T, Nowosad J, Robitaille A, Beck M, Johnson M (2023) jhollist/elevatr: CRAN Release v0.99.0. Zenodo 2023, v0.99.0.
| Crossref | Google Scholar |

International Atomic Energy Agency (2021) ‘Reference Sheet for Certified Reference Material GRESP, Greenland Summit Precipitation.’ (IAEA: Vienna, Austria)

Jasechko S (2019) Global isotope hydrogeology―review. Reviews of Geophysics 57, 835-965.
| Crossref | Google Scholar |

Kerr T (2013) The contribution of snowmelt to the rivers of the South Island, New Zealand. Journal of Hydrology 52, 61-82.
| Google Scholar |

Kerr T, Srinivasan M, Rutherford J (2015) Stable water isotopes across a transect of the Southern Alps, New Zealand. Journal of Hydrometeorology 16, 702-715.
| Crossref | Google Scholar |

Kirchner JW, Allen ST (2020) Seasonal partitioning of precipitation between streamflow and evapotranspiration, inferred from end-member splitting analysis. Hydrology and Earth System Sciences 24, 17-39.
| Crossref | Google Scholar |

Lachniet MS (2009) Climatic and environmental controls on speleothem oxygen-isotope values. Quaternary Science Reviews 28, 412-432.
| Crossref | Google Scholar |

Lin W, Barbour MM, Song X (2022) Do changes in tree-ring δ¹⁸O indicate changes in stomatal conductance? New Phytologist 236, 803-808.
| Crossref | Google Scholar | PubMed |

Liu F, Williams MW, Caine N (2004) Source waters and flow paths in an alpine catchment, Colorado Front Range, United States. Water Resources Research 40, W09401.
| Crossref | Google Scholar |

Lorrey AM, Boswijk G, Hogg A, Palmer JG, Turney CS, Fowler AM, Ogden J, Woolley J-M (2018) The scientific value and potential of New Zealand swamp kauri. Quaternary Science Reviews 183, 124-139.
| Crossref | Google Scholar |

Marttila H, Dudley B, Graham S, Srinivasan M (2018) Does transpiration from invasive stream side willows dominate low‐flow conditions? An investigation using hydrometric and isotopic methods in a headwater catchment. Ecohydrology 11, e1930.
| Crossref | Google Scholar |

Nelson DB, Basler D, Kahmen A (2021) Precipitation isotope time series predictions from machine learning applied in Europe. Proceedings of the National Academy of Sciences 118, e2024107118.
| Crossref | Google Scholar | PubMed |

Prince HD, Cullen NJ, Gibson PB, Conway J, Kingston DG (2021) A climatology of atmospheric rivers in New Zealand. Journal of Climate 34, 4383-4402.
| Crossref | Google Scholar |

Purdie H, Bertler N, Mackintosh A, Baker J, Rhodes R (2010) Isotopic and elemental changes in winter snow accumulation on glaciers in the Southern Alps of New Zealand. Journal of Climate 23, 4737-4749.
| Crossref | Google Scholar |

Salinger MJ (1980) New Zealand climate: I. Precipitation patterns. Monthly Weather Review 108, 1892-1904.
| Crossref | Google Scholar |

Stadnyk TA, Holmes TL (2023) Large scale hydrologic and tracer aided modelling: a review. Journal of Hydrology 618, 129177.
| Crossref | Google Scholar |

Stewart MK, McDonnell JJ (1991) Modeling base flow soil water residence times from deuterium concentrations. Water Resources Research 27, 2681-2693.
| Crossref | Google Scholar |

Stewart MK, Mehlhorn J, Elliott S (2007) Hydrometric and natural tracer (oxygen‐18, silica, tritium and sulphur hexafluoride) evidence for a dominant groundwater contribution to Pukemanga Stream, New Zealand. Hydrological Processes: An International Journal 21, 3340-3356.
| Crossref | Google Scholar |

Tait A, Macara G (2014) Evaluation of interpolated daily temperature data for high elevation areas in New Zealand. Weather and Climate 34, 36-49.
| Crossref | Google Scholar |

Tait A, Woods R (2007) Spatial interpolation of daily potential evapotranspiration for New Zealand using a spline model. Journal of Hydrometeorology 8, 430-438.
| Crossref | Google Scholar |

Tait A, Henderson R, Turner R, Zheng X (2006) Thin plate smoothing spline interpolation of daily rainfall for New Zealand using a climatological rainfall surface. International Journal of Climatology: A Journal of the Royal Meteorological Society 26, 2097-2115.
| Crossref | Google Scholar |

Tait A, Sturman J, Clark M (2012) An assessment of the accuracy of interpolated daily rainfall for New Zealand. Journal of Hydrology 51, 25-44.
| Google Scholar |

Terzer-Wassmuth S, Wassenaar LI, Welker JM, Araguas-Araguas LJ (2021) Improved high-resolution global and regionalized isoscapes of δ¹⁸O, δ²H and d-excess in precipitation. Hydrological Processes 35(6), e14254.
| Crossref | Google Scholar |

Vystavna Y, Harjung A, Monteiro LR, Matiatos I, Wassenaar LI (2021) Stable isotopes in global lakes integrate catchment and climatic controls on evaporation. Nature Communications 12, 7224.
| Crossref | Google Scholar | PubMed |

West JB, Bowen GJ, Cerling TE, Ehleringer JR (2006) Stable isotopes as one of nature’s ecological recorders. Trends in Ecology & Evolution 21, 408-414.
| Crossref | Google Scholar | PubMed |

West JB, Bowen GJ, Dawson TE, Tu KP (2009) ‘Isoscapes: Understanding Movement, Pattern, and Process on Earth through Isotope Mapping.’ (Springer)

Yang J, Dudley BD, Montgomery K, Hodgetts W (2020) Characterizing spatial and temporal variation in ¹⁸O and ²H content of New Zealand river water for better understanding of hydrologic processes. Hydrological Processes 34, 5474-5488.
| Crossref | Google Scholar |