Anchor spectral transmission entry to selected point readout

[Jesssullivan]

Linear: https://linear.app/tinyland/issue/TIN-1739/anchor-spectral-transmission-entry-to-selected-point-readout

Problem

The historical transmission tracker still describes the old LayerRail row (i) / global CTA model. That entry point is now misleading for V3: spectral transmission is a selected-point/directable-boresight analysis, not a global layer-rail command.

The useful inputs for T(λ) are bound to the chosen location and time: clicked-point atmospheric readout, PWV/RH/visibility, PM2.5-derived or manual AOD, band guidance, and selected boresight assumptions. A LayerRail CTA suggests the analysis belongs to a layer toggle and worsens the mobile HUD competition.

Scope

• Remove the global LayerRail spectral-transmission CTA and per-atmospheric-row (i) entry points.
• Add the transmission entry to PointReadout.svelte, gated on a selected point with atmospheric data.
• Open the existing TransmissionSheet from that selected point and keep recomputation seeded from current readout/time state.
• Update guided-tour copy so users discover the path by tapping the map, then opening T(λ) from the point readout.
• Update Playwright specs to drive the point-anchored flow through a reusable helper that stubs /api/atmospheric/point; sheet-render specs should not depend on live Open-Meteo availability.
• Keep this compatible with Mobile field HUD: non-competing point readout, transmission sheet, toolbar, and twilight rail #312 / TIN-1734 so the readout CTA and sheet participate in the non-competing mobile HUD layout.

Acceptance

• No transmission CTA or (i) transmission button remains in LayerRail.svelte.
• A selected map point with atmospheric data shows a clear “Spectral transmission T(λ)” action in the point readout.
• The action opens the existing atmospheric transmission sheet and recomputes from selected-point state.
• Existing transmission e2e specs no longer depend on the old LayerRail entry point.
• Remote CI/GitHub Actions provide validation; no local tests, builds, switches, or checks are required.

Related

• PR-H: Transmission widget — (i) chevron on atmospheric rows → bottom sheet with T(λ) chart #158 / TIN-1662: original LayerRail-bottom-sheet tracker, now outdated in entry-point shape.
• Mobile field HUD: non-competing point readout, transmission sheet, toolbar, and twilight rail #312 / TIN-1734: mobile field HUD layout arbitration.
• Add frequency-specific atmospheric analysis requests for laser/RTX planning #250 / TIN-1703: frequency-specific field planning.
• Transmission AOD source modes: manual versus OpenAQ-derived PM2.5 estimate #306 / TIN-1728: AOD source modes.
• Effect PM2.5 field service: cached OpenAQ viewport sensors plus point kernel sample #304 / TIN-1726 and V2: explicit PM2.5 kernel-diffusion field feeding the transmission estimator #275 / TIN-1731: PM2.5 field and kernel-diffusion handoff.
• Define atmospheric observation provenance and no-data semantics before transmission auto-fill #271 / TIN-1711: provenance/no-data semantics before auto-fill.

[Jesssullivan]

Implementation PR opened: #318

This is the feat/tx-readout-entry branch currently in the root worktree. It moves spectral transmission from the old LayerRail/global CTA model to the selected PointReadout, adds a reusable Playwright helper for the point-anchored flow, and updates the stale e2e expectations that still targeted the removed LayerRail buttons.

Validation is intentionally remote-only. PR CI should provide the proof.