Scientific model and method

LunarTrace’s contribution is not a single map, image viewer, or screening algorithm. It is an explicit chain from a research reference to observations, comparisons, evidence, and reviewable claims.

Three questions should remain answerable throughout that chain:

What population was examined? A source query has a scope, a retrieval time, a closure state, and limits. The table on screen may show only a view of that population.

What operation produced the result? Temporal partitioning, pair construction, missing-metadata exclusions, and multi-objective screening are distinct operations with explicit predicates.

What does the result warrant? A nondominated comparison is not proof of image comparability. Matching bytes are not proof of scientific truth. A historical aggregate is not independently replayable when its original rows are absent.

Reading sequence

Start with the workflow overview. Use the coordinate and discovery chapters when defining a source query, the methodology chapter when interpreting a frontier, and the case study when examining the retained historical example.

• From a reference to a scientific record
  • The research chain
  • Three deliberately different outputs
  • Why the separation matters
  • Interoperability rather than enclosure
• Concepts and vocabulary
  • Research objects
  • Source and population objects
  • Evidence and authority
  • The distinctions to remember
• Coordinates and lunar context
  • The Lab’s user-defined reference
  • The historical anchor remains separate
  • Query geometry
  • Basemap provenance and limits
  • A useful interpretation checklist
• Source discovery, closure, and currentness
  • Fixed source profile
  • Retrieval sequence
  • What closure means
  • Currentness is not capture time
  • Normalization without invention
  • Captured bytes and rendered values
  • Empty results and failures
  • Public-service limits
• Event-conditioned comparison methodology
  • 1. Define the reference interval
  • 2. Normalize and identify observations
  • 3. Construct the candidate population
  • 4. Calculate explicit metrics
  • 5. Apply a declared objective set
  • 6. Interpret Pareto nondominance
  • 7. Keep computation bounds visible
  • 8. Carry interpretation beyond screening
• Chang’e 6: a retained historical case
  • Event anchor and its limits
  • What was retained
  • Reference identities
  • What was not retained
  • Historical currentness
  • What a new Lab search means
  • The scientific lesson
• Benchmarks, tests, and what they establish
  • Historical benchmark recovery
  • Deterministic software tests
  • Runtime and browser qualification
  • No invented performance results
  • Reproduction commands
• Limitations and non-claims
  • Capability boundaries
  • Historical incompleteness
  • Source completeness and freshness
  • Geometry and image processing
  • Computational limits
  • Operational limits
  • Publication and availability
  • Responsible interpretation