Best Practices for Multi-Sol Acoustic Surveys in Lava Tubes
The Problem
A BASALT-style analog crew at a Mauna Loa skylight budgeted 14 Earth-days for a lava tube soundscape study, expecting to operate on a simulated Mars timeline. By sol 3, they had collected 180 GB of audio that was 70% useless. Daytime winds at the surface had overwhelmed their deployed microphones during every recording window between 7 and 17 LMST, exactly when SEIS-InSight atmospheric noise data predicts convective turbulence dominates Mars signals. Their science team had never mapped Martian sols onto operational clock time, so they kept scheduling surveys during what felt like reasonable working hours on Earth.
The cost compounds across a multi-sol campaign. A Martian sol is 24 hours, 39 minutes, 35 seconds, so every sol drifts 39 minutes from Earth-day cadence. Power recharge cycles, DSN antenna passes, ambient wind regimes, and thermal-crack rates all follow sol-local time, not Earth-local time. A survey scheduled around Earth shift changes misses the 22-04 LMST quiet window when passive acoustic imaging actually resolves tube geometry. Multi-sol campaigns built without LMST-aware scheduling collect hours of data that cannot be stitched into a single coherent geometry, burning power budget and bandwidth on noise the operators then filter out in post.
The Apollo Passive Seismic Experiment ran for 8 years and logged roughly 13,000 lunar events, but only because each station recorded continuously and was analyzed against lunar time, not Earth schedules. Analog teams today often compress that insight into 7-sol windows without matching LMST discipline.
The cost surfaces in three places that are individually painful and collectively fatal. First, post-campaign filtering throws away roughly 60-70% of high-noise audio, which means the same ground-truth resolution would have been achievable with one-third the storage if the LMST gating had been applied at capture time. Second, DSN and Mars relay bandwidth gets spent on noisy patches that downselect later, displacing higher-signal patches that should have ridden the same pass. Third, the science narrative collapses because reviewers cannot see continuous geometry across the campaign — they see fragments separated by gaps that match the operator's schedule rather than the rover's environment. JPL's MSL operations team documented in their supratactical workflow analysis that misaligned LMST cadence is one of the top-three sources of ops-replanning in the first sols of a campaign, and the pattern holds for every subsequent rover that has tried to compress Earth-week thinking onto Mars-sol execution.
The Solution: LMST-Aware Quilt Scheduling
EchoQuilt's multi-sol scheduler operates on LMST directly and stitches recording windows into a single continuous quilt geometry across every sol of a campaign. Operators declare scientific objectives (tube width profile, skylight echo signature, structural microseismic baseline), campaign duration in sols, power envelope per sol, and DSN or relay contact windows. The scheduler then proposes recording windows positioned during quiet LMST bands, typically 22-04 local, when surface-derived wind noise drops below the instrument floor. This matches the approach validated by Keil 2026 in JGR Planets, which shows passive ambient noise cross-correlation resolves lunar lava tubes when recording windows are dense and low-noise.
Each sol in the quilt becomes a patch, and the scheduler maintains overlap between adjacent-sol patches so stitching boundaries are testable. If sol 3 captures a 22-03 LMST window and sol 4 captures 23-04 LMST, the overlap provides a repeatability check on features that should appear in both. When the overlap patches disagree beyond a tolerance, the tool flags the geometry for review rather than silently averaging it into the final map. This stitching discipline preserves the audit trail that planetary mission reviewers require.
The scheduler also handles sol drift against Earth calendar workflows. Mission ops teams run supratactical planning on Earth weeks, and MSL's published supratactical process formalized how to move from strategic campaign goals down to sol-level plans. EchoQuilt aligns its sol patches to supratactical checkpoints so science leads reviewing results every 3-7 sols see continuous quilt growth rather than scattered fragments. Each patch carries metadata about LMST window, instrument state, power draw, and parent campaign objective, which lets reviewers trace any feature in the final map back to the sol that produced it.
Terrestrial calibration follows the same cadence. Teams running campaigns at Kilauea, Lofthellir, or La Corona use EchoQuilt in LMST-equivalent mode to test their scheduling assumptions before committing to Mars-analog protocols. The approach borrows from terrestrial microseismic tube detection studies, which have demonstrated that multi-day passive campaigns outperform single-session active surveys for subsurface geometry.
These scheduling decisions connect directly to teleoperated workflows. Our guide to teleoperated workflows covers how the supervisor specifies sol targets under light-speed delay, and the supervisory command schema is what carries the LMST window assignment from Earth-side planning into rover-side execution.
Advanced Tactics
Budget your sols in tiers. Use the first 1-2 sols as instrument shake-out and wind-regime characterization; the scheduler will flag recording windows where noise is anomalously high so you can adjust microphone placement before locking into the core survey. Sols 3 through N-1 become the production quilt. Reserve the final sol for redundancy captures of any tiles that failed cross-check during stitching, which avoids ending a campaign without a complete map.
Tie DSN pass windows to compression targets, not raw download. If a pass window allocates 90 minutes and your quilt has produced 4 GB of new patches, the scheduler can downselect which patches ship first based on scientific priority metadata. Patches near known skylights or branching geometry go on the high-priority queue; baseline repeat tiles defer to later passes. This mirrors the cross-campaign coordination we describe for long entrapment sync, where the same patch-priority logic governs rescue-timeline decisions.
For seven-sol campaigns at Mauna Kea or Surtshellir analog sites, plan sol 0 as Earth-day calibration against known tube geometry before transitioning to LMST mode. The ground-truth anchoring lets you quantify stitching error per patch, so by sol 7 the EchoQuilt map carries confidence intervals rather than unqualified lines. NIAC-funded concept teams have used this pre-calibration pattern to move from TRL 3 to TRL 4 within a single campaign.
Track the wind-regime envelope per sol as a separate timeseries alongside your patch log. The scheduler exposes a wind-derived noise prediction for each upcoming LMST window, but the realized noise is what determines whether a patch was actually within tolerance. When the predicted-vs-realized noise diverges by more than 3 dB on more than 20% of windows in a campaign, that is a signal to reweight the meteorological model the scheduler uses for that site, not to abandon LMST gating. Mauna Loa, Mauna Kea, and Lofthellir all have characteristic wind regimes that the scheduler learns over the first 2-3 sols, and the learned model carries forward into subsequent campaigns at the same site, so the second deployment to a known site benefits from sol-0 calibration that has been pre-tuned by the first.
For campaigns extending past 30 sols, the multi-sol scheduler hands off to long-duration monitoring mode, where the cadence shifts from intensive scouting to baseline-with-drift-detection. The handoff between modes is explicit in the EchoQuilt configuration, so campaigns scoped for an initial scouting phase followed by long-duration baseline can plan the transition rather than rebuilding the operational rhythm mid-campaign.
Reserve at least one EchoQuilt sensor node as a quiet-window control. Place it in a known geometry-stable section of the tube and let it record continuously across all LMST bands, not just the scheduled quiet windows. The control node's data is your ground truth for what the noise floor actually was during scheduled-noisy bands, and it lets the scheduler validate its own gating decisions against direct measurement rather than just predicted wind models. This pattern adds roughly 5% to the campaign's storage budget but yields a self-consistency check that planetary mission reviewers consistently flag as raising the credibility of the resulting quilt.
Ready to Plan Your Next Multi-Sol Campaign?
Planetary analog researchers committing seven sols of field time deserve a scheduler that treats LMST as first-class and stitches sols into a continuous geometry rather than a collection of isolated recordings. EchoQuilt is built around that requirement, with integrations for BASALT-style supratactical review, DSN-aware downlink planning, and Electra UHF link-budget analysis sized to MRO and MAVEN proximity passes. Each pilot ships with a measured wind-regime envelope from prior Mauna Loa, Mauna Kea, and Lofthellir campaigns, an LMST gating configuration pre-tuned to BASALT-equivalent supratactical cadences, and a quiet-window control node configuration that validates the scheduler's gating decisions against direct measurement rather than predicted models. Pilot teams shape the sol-overlap tolerance defaults and the LMST calendar import format that the 2027 reference release will adopt for NIAC, MatISSE, and ESA PANGAEA campaign integration.
Priority goes to NIAC concept teams targeting TRL 4 advancement within a single seven-sol field season, JPL MSL operations alumni who have run supratactical workflow analyses, and ESA PANGAEA campaign coordinators planning multi-week analog deployments at Lanzarote or Hawaiian sites. Join the Waitlist for Planetary Analog Researchers to get early access to the multi-sol scheduler and shape how it integrates with your mission's planning cadence.