Answers are grounded in this paper and cite their sources.
Groundwater potential maps depend on analytical choices that can be hidden by a single deterministic zonation. We developed a principal component analysis (PCA)-GIS assessment for the Lower Namoi Alluvium, Australia, using sixteen environmental factors on a common 30-m grid. The five-component Kaiser baseline explained 74.47% of predictor variance; the four-component parallel-analysis solution was a retention sensitivity alternative. We compared six weighting methods, factor omission, suitability functions, classification, and resolution and evaluated conditional PCA-weight uncertainty using spatial block bootstrapping. Clay omission changed 40.39% of fixed-threshold class assignments. The high plus very-high area was 33.09% under Fisher-Jenks and 6.18% under equal intervals (κ = 0.334). Gaussian class probabilities were checked against empirical bootstrap frequencies, with small average but material local differences. An exploratory comparison with 54 existing positive-yield bore records gave Spearman ρ = 0.205 (permutation p = 0.133), without establishing predictive validity. The zones represent relative environmental favorability, rather than validated yield classes. Conditional weight variability and sensitivity to model structure should therefore be reported alongside continuous indices and the limitations of external hydrogeological evidence.
No takes yet. Share an insight, caveat, or question.
Yazdanpanah et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: