Back to Blog
Mining & Resource Modelling
Why Your Resource Model Should Refuse to Answer
The most valuable thing an estimation system can do is decline to produce a number nobody can defend.
SpatialTechSolutions Team2026-08-116 min read
A default is a decision nobody made
The variogram model, the tail models, the search neighbourhood, and the stationarity mean decide the reported high-grade tonnage. When a tool pre-fills them, the resulting figure carries a choice no geologist ever consciously made — and it will not survive the first serious question about where it came from. Requiring those fields, empty, with an explanation of why no default exists, is slower on the first run and far cheaper on review.
An uninformed block is not an estimate
If the nearest sample to a block sits well beyond the variogram range, kriging it means interpolating from data the model itself says is uncorrelated. Widening the search until every block in the box gets a number produces a tidier model and a worse one. Refusing those blocks makes the informed envelope of the deposit visible, which is usually the more honest picture.
Refusals need structure, not prose
A refusal that arrives as an error string forces the consuming application to parse English to decide what to do. A refusal that arrives as a code, a category, and a stage can be routed: fix the form, explain the science, retry the execution, or escalate the infrastructure. The four cases demand genuinely different responses, and collapsing them into one message loses that.
The reward is a shorter review
Every refusal is a defect caught before it reaches a report. Teams that adopt this find the friction moves to where it is cheap — job submission — and disappears from where it is expensive: the technical review meeting where somebody asks which composite length was used and nobody is sure.
Related posts
Mining & Resource Modelling
Reproducibility as a Feature of Geoscience Software
Content-addressed results, recorded provenance, and what it takes for a reviewer to verify a number instead of trusting it.
AI & Geospatial AI
The Future of AI-Powered GIS
How AI agents, governed tools, and enterprise GIS platforms are changing spatial operations.
Enterprise GIS
Building Enterprise GIS Architectures
A practical view of architecture, governance, data flows, and deployment choices for enterprise GIS.
Request demo
Turn geospatial insight into an operational workflow
Explore SpatialTechSolutions products, services, and AI-enabled GIS architecture for your organization.