Skip to content
LunarTraceDocs

Documentation

LunarTrace

Inspect lunar terrain without losing the evidence. Python computes; the workbench makes every threshold, source and limit visible.

Get startedThe one-call StudyPython API

LunarTrace is an experimental Python package and web workbench for event-conditioned lunar observation Studies. Version 0.2 puts the temporal method behind one call: name a site, an event interval and a product type, and get back a Study that can be checked, saved and replayed.

Replay the archived Vikram run
import lunartrace as lt


study = lt.temporal.study(site=(-69.3741, 32.32),
                          event="2023-08-23T12:33Z/12:35Z",
                          products="LRO/LROC/CDRNAC4",
                          archive="vikram-2026-10-14")
study.frontier("angles")              # the 13 pairs nothing beats
study.why("pair:002ea7a82e1060f1")    # names the pair that beat it
study.save("vikram.study.json")       # replays to the same sha256

archive="vikram-2026-10-14" replays the archived pages of the 14 October 2026 run. The snapshot ships inside the package, so the call needs no network and returns the same result every time:

Products
108
Pairs
2,340
Digest
sha256:3d75a662ce75f00cd34dd9ba288012ad471c73b1e56ecf0abe36896d0b199dd0

Install it with pip install lunartrace (Install). The same three steps are available on the command line as lunartrace study, lunartrace why and lunartrace verify; see the one-call Study.

The second vertical is the frozen PGDA LOLA Site01 slope quality result. The method lunartrace.slope_quality/v1 evaluates it into a deterministic result, an execution receipt and a typed browser projection.

Start here

Section titled “Start here”
QuickstartInstall with pip, then replay the archived Vikram run in three lines.
Installpip, a local wheel or the locked checkout, and the optional extras.
The one-call StudyFrontiers, why, save and replay for the temporal Study.
Reproduce Site01Verify frozen source bytes, run the study, and keep the result and receipt.
Run the workbenchExplore, Analyze, Compare, Publish and Present over Python-authoritative results.
Method and limitsThe closed cell rule, the independent oracle and the limits of the evidence.

Who it is for

Section titled “Who it is for”
  • Researchers who need to know which pre-event and post-event lunar observations are comparable, and need that answer to replay.
  • Anyone inspecting a source-backed terrain-quality calculation.
  • Python users reproducing the frozen Site01 example.
  • Reviewers who need the method, the source witness and the proof ceiling beside every displayed number.

Module contracts are in the generated API reference; source custody is in Provenance.

What LunarTrace is

Section titled “What LunarTrace is”
  • A one-call temporal Study: the observations before and after an event, every PRE × POST pair, the Pareto frontiers of the pairs nothing beats, a why answer that names the beating pair, and a digest that shows nothing moved. Method: lunartrace.temporal_prepost/v1.
  • A deterministic implementation of the Site01 slope-quality method. Identical inputs give identical canonical result bytes.
  • A custody-aware STUDY path that resolves a frozen source-witness manifest from a local content-addressed store and re-hashes every input before use.
  • Generated, typed projections (JSON Schema, TypeScript types and AJV validators) that the browser reads without recomputing science.

What LunarTrace is not

Section titled “What LunarTrace is not”
  • It is not a landing-safety assessment, a surface-change detector or a physical-area calculation. Temporal adjacency is not a surface-change claim.
  • It is not a drop-in upgrade of the older PyPI releases. 0.2.0 is on PyPI; 0.1.0 and 0.1.1 under the same name are a different codebase, and 0.2.0 removes their API and command line. See the changelog.
  • It does not redistribute source data. Source and derived-data rights are UNVERIFIED; keep all inputs and outputs local and private.