Computing Solar Eclipses — Research

Reports

updated 2026-09-16

Five pages, written to be read in order and to stand without the raw notes.

  • Executive summary: the findings that carry the most weight, in one page.
  • Pipeline design: the stages, their inputs, formulas, constants, outputs and validation cases.
  • Error budget: every input ranked by its effect in seconds and metres, and the two product modes that follow from it.
  • Reading list and reference matrix: the papers, code and datasets in the order a developer should take them.
  • Open questions: the artefacts that would most change the conclusions.

Every claim here is carried from a raw note, and the citation travels with it. The raw research holds the full derivations.

In this section

Executive summaryOne algorithm, four constants that decide the edges, one published state-of-the-art method, and no open-source implementation of it.workingPipeline designNine stages from ephemeris kernels to limb-corrected umbra polygons, with the formulas, constants, outputs and validation case for each.workingError budgetEvery input to an eclipse prediction ranked by its effect in seconds of contact time and metres on the ground, and the two product modes that follow.workingReading list and reference matrixThe papers, code and datasets a developer should take in order, and what each one uniquely provides.workingOpen questionsThe artefacts that would most change the conclusions, ranked, each naming the document, dataset or line of code to obtain.working