219 These findings have practical implications for applications such as geological modeling, resource classification, and mine planning. In such contexts, spatial models are often used to support decisions that depend on the relative importance of competing categories, such as ore versus waste or alternative geological domains. Treating hierarchical ordering as a secondary or default choice may obscure its role in shaping spatial outcomes and, ultimately, decision risk. Explicitly acknowledging and testing alternative hierarchies can therefore improve transparency and robustness in spatial decision making workflows. It is important to emphasize that this study intentionally adopts a simplified experimental setup to isolate the effect of hierarchical ordering. The use of a single realization, isotropic variograms, and fixed parameters limits the scope of inference but strengthens causal interpretation. Future work may extend this framework by incorporating multiple realizations, geological rules, local proportions, or data driven tree optimization strategies, further exploring how decision structures interact with spatial uncertainty in more complex settings. From an operational perspective, the observed sensitivity to hierarchical ordering has direct implications for decision-making workflows in mining. Spatial models derived from HSIS are often used as inputs for downstream tasks such as resource classification, ore/waste discrimination, and mine planning. In these contexts, alternative spatial realizations are not merely descriptive, but prescriptive, as they influence economic and operational decisions. The results of this study suggest that different hierarchical orderings may lead to alternative spatial classifications even when based on identical probabilistic inputs. This implies that two models, both statistically consistent and conditioned to the same data, may support different decisions depending solely on how the decision tree is structured. Such divergence is particularly relevant in threshold-based decisions, where small spatial differences may result in changes in classification boundaries or resource categorization. Under this perspective, hierarchical ordering can be interpreted as an implicit decision policy embedded within the simulation algorithm. Early-stage decisions in the hierarchy prioritize certain categories over others, effectively allocating spatial opportunity in a way that reflects modeling assumptions about importance or precedence. As a consequence, the choice of hierarchy influences not only spatial patterns, but also the distribution of decision-relevant outcomes. Recognizing hierarchical ordering as a decision variable rather than a fixed modeling parameter opens new possibilities for uncertainty analysis. Instead of evaluating uncertainty solely through multiple realizations under a fixed hierarchy, alternative hierarchies may be explored as competing decision scenarios. This approach enables a more comprehensive assessment of uncertainty, explicitly incorporating the impact of structural modeling choices on downstream decisions. 6. CONCLUSION This study demonstrates that hierarchical ordering in Hierarchical Sequential Indicator Simulation is a core modeling decision rather than a secondary operational choice. By isolating the decision tree hierarchy as the only variable component in a controlled experiment, the results show that different orderings yield distinct spatial allocations, even under identical probabilistic and geostatistical conditions. Interpreting HSIS as a stochastic decision tree evaluated in space provides a useful conceptual framework to understand this behavior. Within this framework, hierarchical ordering defines the sequence in which uncertainty is resolved and therefore governs how competing propositions are translated into spatial outcomes. Small changes in hierarchy definition, such as those induced by declustering, can propagate into large scale differences in spatial patterns.
RkJQdWJsaXNoZXIy MTM0Mzk2