AgResearch
Browse

File(s) not publicly available

Achieving unbiased predictions of national-scale groundwater redox conditions via data oversampling and statistical learning

journal contribution
posted on 2024-06-21, 03:50 authored by Scott Wilson, Murray Close, Phillip Abraham, Theo Sarris, Laura Banasiak, Roland Stenger, John Hadfield
An important policy consideration for integrated land and water management is to understand the spatial distribution of nitrate attenuation in the groundwater system, for which redox condition is the key indicator. This paper proposes a methodology to accommodate the computational demands of large datasets, and presents national-scale predictions of groundwater redox class for New Zealand. Our approach applies statistical learning methods to predict redox status in areas without sample data. A key achievement was to overcome the influence of sample selection bias on model training via oversampling. National maps are provided for redox class and probability at specified depths. Our model provides unbiased predictions at a scale relevant for environmental policy, and enables targeted interventions that can achieve the desired environmental outcome in a more cost-effective manner than non-targeted interventions.

Funding

Funded by the New Zealand Ministry for Business, Innovation and Employment's Our Land and Water National Science Challenge (Toitū te Whenua, Toiora te Wai) as part of project Sources and Flows

History

Publication date

2020-02-07

Language

  • English

Does this contain Māori information or data?

  • No

Journal title

Science of The Total Environment

Usage metrics

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC