Field science operations and the Mars laboratory

Field science on Mars is an operational chain from question to evidence, not a collection of interesting rocks

A scientific campaign begins before anyone walks outside. The team must define a question that observations can actually discriminate, decide what measurements or samples would change the interpretation, and identify the controls needed to distinguish a real signal from contamination or instrument drift. Without that structure, a crew can return with beautiful specimens and terabytes of data that do not answer the original question.

Field work is expensive because every station consumes EVA time, mobility range, suit consumables, communications attention, sample capacity and later laboratory effort. A good plan therefore separates must-have observations from desirable additions. The crew records context before disturbing the site, because stratigraphic position, texture, orientation and association can be impossible to reconstruct once a sample has been removed. Science quality depends on preserving relationships, not just objects.

The local laboratory is part of the campaign rather than a separate destination. Some analyses are non-destructive and can guide which samples deserve scarce destructive tests. Other procedures consume material, alter mineral phases or risk contamination. The team should therefore sequence measurements from least destructive to most destructive where scientifically appropriate, while preserving archive material for later instruments or for possible return to Earth.

Calibration, blanks, standards, duplicates and metadata are what make measurements interpretable months or years later. A number without instrument state, method, location, time, operator and sample history may be impossible to compare with a later result. Mars-Earth delay reinforces this need because the local crew must often decide whether to repeat a measurement or modify a traverse before specialists on Earth can respond.

Planetary protection and crew contamination control add another boundary. Samples, tools, suits and laboratory spaces can move terrestrial organisms, dust and chemicals between environments. Procedures must distinguish scientific contamination, biological protection and ordinary occupational cleanliness. These goals overlap but are not identical; a method that protects crew health may still ruin a trace-organic measurement if it introduces the wrong cleaning chemical.

Twelve practices that turn a traverse into defensible science

1. Start with a testable question

A useful field question identifies competing interpretations and the observation that would make one interpretation more plausible than another. “Study this outcrop” is a destination, not a scientific question. “Does this layer contain textures and mineral associations consistent with repeated water-rock interaction?” gives the crew a reason to photograph specific contacts, compare units and collect paired samples.

Testable questions also protect mission time. When a new feature appears, the crew can ask whether it bears on the campaign hypothesis or deserves a documented opportunistic observation rather than derailing the traverse. Curiosity remains essential, but it is managed against the finite operational budget.

2. Record context before touching the target

Images, scale, orientation, surrounding units, bedding relationships, surface condition and the exact sampling position should be captured before tools disturb the scene. A specimen can retain chemistry while losing the field relationship that explained how it formed. Context is therefore part of the sample and deserves the same chain of custody.

A standard observation sequence reduces omission under time pressure. Wide scene, intermediate context and close detail can be linked to a common station identifier. Notes should distinguish direct observation from interpretation so later reviewers can reconsider the hypothesis without losing what the crew actually saw.

3. Design sampling for variation, not for the most attractive specimen

Scientific sampling should represent the range needed to answer the question: different layers, distances from a contact, altered and unaltered material, or multiple points across a gradient. Collecting only visually striking material can create selection bias. The plan therefore specifies why each sample exists and what comparison it enables.

Replicates can reveal natural heterogeneity or measurement scatter. They are not wasted mass when variability is central to the question. Conversely, collecting many near-identical samples without a comparison logic consumes storage and laboratory capacity without increasing information proportionally.

4. Keep calibration, blanks and controls conceptually separate

Calibration connects instrument response to a known reference. A blank checks whether the method or handling path introduces a signal when the target substance should be absent. A control provides a comparison condition that helps interpret change. These tools answer different quality questions and should not be treated as interchangeable administrative rituals.

Field plans identify where each control enters the chain. A clean tool blank can test sampling contamination; a procedural blank can follow containers and preparation steps; a reference standard can check instrument drift. When a result is surprising, these controls help determine whether the surprise belongs to Mars or to the measurement process.

5. Metadata is the memory of the experiment

Every sample and data product should link to location, time, operator, instrument, software or method version, environmental conditions and relevant configuration. The amount of metadata must be manageable enough that crews actually capture it. Automated acquisition can reduce burden, but operators still need a way to add observations and anomalies that sensors do not know.

Identifiers should remain stable through field, transport, laboratory subsampling and archive. Renaming files casually or copying labels by hand creates opportunities for silent mix-ups. A robust system makes provenance visible from the final analytical result back to the original station.

6. Chain of custody protects both identity and condition

Custody is not only a legal concept. In science it means knowing who handled a sample, which container it occupied, what temperatures or atmospheres it experienced, and whether the seal was broken. This record lets researchers decide whether a later anomaly could have been introduced during handling.

For fragile or volatile samples, condition can matter as much as identity. Storage rules may include temperature limits, shielding from light, vibration protection or controlled atmosphere. The scientific value of a sample is the combination of material and preserved history.

7. Sequence laboratory work from informative to irreversible

Visual inspection, imaging, mass measurement or non-destructive spectroscopy may guide whether a sample deserves cutting, heating, dissolving or other destructive preparation. The best sequence depends on the scientific objective, but the decision should be explicit because consumed material cannot be recovered.

Archive fractions provide insurance against future instruments and revised hypotheses. A crew under pressure to produce immediate results may be tempted to use every gram. Preserving a documented fraction can be scientifically more valuable than one additional current measurement, especially when resupply or return is impossible.

8. Quality control needs repetition at the right places

Repeating every measurement is inefficient, but never repeating measurements makes it hard to recognize drift or random error. The plan selects critical checkpoints for duplicates, standards or independent methods. Repetition is most valuable where a result drives a major interpretation or an irreversible decision.

Unexpected disagreement should trigger diagnosis, not averaging by habit. The team examines calibration, sample heterogeneity, preparation, environmental conditions and software processing. Averaging two incompatible numbers can hide the very clue that identifies the failure mode.

9. Field autonomy requires pre-agreed decision rules

Earth specialists may not be able to answer before the crew must move on. The campaign can therefore define local triggers: if a measurement exceeds a threshold, take a second sample; if an instrument fails calibration twice, switch to the backup method; if a site consumes more than the allocated time, preserve a minimum observation set and continue.

These rules do not remove scientific judgment. They protect it by giving the crew authority within a shared framework. Later Earth review can still redirect future sorties, but the current EVA is not paralyzed by communication delay.

10. Science competes with maintenance and safety for the same people

A settlement does not have a separate unlimited science workforce. The geologist may also be an EVA specialist, mechanic or medical responder. Campaign planning therefore includes crew-hours, recovery time and the probability that urgent maintenance cancels a sortie. A scientifically elegant schedule that assumes no operational interruptions is fragile.

Priorities should preserve rare opportunities without endangering settlement reliability. A transient atmospheric event or seasonally accessible site may deserve protection in the schedule, while routine sampling can move. Explicit priority makes trade-offs defensible when operations become crowded.

11. Data volume is not the same as information value

High-resolution instruments can produce more data than can be reviewed locally or transmitted promptly. The team should preserve raw data where practical while creating summaries, quality flags and prioritized products for decision-making. Compression or down-selection must not erase the context needed to reinterpret a surprising result later.

Data handling also includes checksums, redundant storage and version control for derived products. A scientific archive must survive equipment failures and software updates just as a life-support log does. The difference is that lost science data may be irreplaceable even when no immediate alarm sounds.

12. Interpretation must label uncertainty and alternatives

Field teams are vulnerable to confirmation bias because they see the campaign unfold and may become invested in an early hypothesis. Reports should separate observation, inference and speculation, and record plausible alternatives. A strong scientific result can remain valuable even when the favored interpretation later changes.

Uncertainty is not weakness. It tells later researchers which measurements are robust and where new evidence could change the conclusion. The objective is to produce a traceable argument whose assumptions can be tested, not a dramatic headline from every sortie.

Field-science calculation laboratory: EVA time and station capacity

Formula 1 — available scientific station time inside an EVA

Formula 2 — simple sampling allocation across stations

Campaign cases: protect evidence while operating under Mars constraints

A surprising organic signal appears in the first sample

The crew should preserve the excitement without immediately declaring discovery. They review blanks, standards, tool history, cleaning chemicals and sample handling, then repeat the measurement if enough material remains. A second sample from related context can test whether the signal is spatially coherent. The team records the instrument configuration and preserves an archive fraction before performing more destructive tests.

If controls are clean and independent evidence supports the observation, the finding becomes stronger. If a blank shows the same signal, the campaign has learned about contamination rather than Mars chemistry. Either outcome is scientifically useful when provenance is complete. The dangerous outcome is a dramatic number with missing controls that no one can later interpret.

A rover wheel issue cuts the traverse range in half

The science lead re-ranks stations by information value and by the questions they discriminate. A distant attractive site may be dropped in favor of nearby paired observations that preserve the core hypothesis test. The crew should not simply visit the first half of the original route, because that may destroy the sampling design.

Route change is documented so later analysts know why some planned samples do not exist. If the missing station creates a major comparison gap, a future EVA can target it specifically. Operational adaptation is part of scientific method when the change is transparent.

A sample label is unreadable after return to the habitat

Identity should be reconstructed only from independent records such as container sequence, timestamped imagery, tool logs and digital station identifiers. If more than one origin remains plausible, the sample is marked uncertain rather than assigned the most convenient identity. Scientific databases should preserve that uncertainty.

The incident then drives a procedural correction: redundant human-readable and machine-readable identifiers, protected labels, verbal confirmation or digital scans before containers leave the field. A chain-of-custody failure is a systems problem, not just an operator mistake.

A laboratory instrument fails calibration during a critical campaign

Repeating the same calibration indefinitely can consume standards and time without improving confidence. The procedure defines how many attempts are reasonable, which diagnostic checks follow and whether a backup or alternate measurement can answer the key question. Samples may be preserved rather than forced through a questionable instrument.

Earth specialists receive the calibration history and raw diagnostics, but local operators retain authority to protect sample integrity. The right decision may be to delay analysis until repair rather than produce numbers whose traceability is already compromised.

Maintenance cancels two planned science days

The campaign is replanned around scientific dependencies. Samples requiring a particular seasonal condition may outrank routine monitoring, while laboratory work on already collected material can fill time when EVA personnel are unavailable. The science team also checks whether maintenance activities themselves generate dust, vibration or chemical contamination that affect experiments.

This avoids treating science and settlement operations as rival departments. The same planning board can expose shared resources and opportunities. Reliable habitation is a prerequisite for sustained science, while good science can improve site knowledge and resource decisions.

A sample may contain a volatile component

Handling should minimize unplanned heating, venting and time at conditions that alter the target. The team chooses a compatible sealed container and records temperature and transfer history. Laboratory sequence prioritizes measurements that can characterize the volatile fraction before destructive preparation changes it.

The key principle is preservation of scientific state. A perfectly labeled sample can still lose its value if the material evolves during storage. Chain of custody therefore includes environmental history, not just possession.

Field-science practice with reasoned solutions

Exercise 1 — Turn a broad goal into a testable question

Rewrite “study the old river channel” as a question that guides observations and samples.

Exercise 2 — Budget station time

An EVA is 420 min. Travel takes 130 min, setup/pack-down 50 min and protected reserve 70 min. How much science time remains?

Exercise 3 — Choose samples across a gradient

You have six containers for a contact between altered and unaltered rock. How would you avoid selecting only the most visually dramatic pieces?

Exercise 4 — Diagnose a contaminated blank

A procedural blank shows the same trace signal as several samples. What should happen before claiming the signal is Martian?

Exercise 5 — Protect sample identity

A sample is split into three laboratory subsamples. What metadata relationship should be preserved?

Exercise 6 — Decide whether to repeat a measurement

Two replicate measurements disagree strongly. Why is taking their average not the first response?

Exercise 7 — Plan local autonomy

Earth cannot answer for twenty minutes, and the crew must decide whether to take an extra sample before leaving. What preplanned rule could help?

Exercise 8 — Separate observation from interpretation

Write one observation and one interpretation for a layered rock face.

Field campaign cases: evidence, contamination and autonomy

A field team discovers an unexpected stratigraphic contact late in the EVA

The crew first captures context imagery and a concise description before deciding whether to sample. The pre-agreed decision rule can permit one opportunistic sample if reserve time, container capacity and contamination controls remain above thresholds. If not, the location is marked precisely for a later sortie. This preserves the discovery without allowing excitement to consume return margin.

Back in the habitat, the contact is placed against the original hypothesis and traverse map. The next campaign can be redesigned deliberately rather than treating the opportunistic observation as an isolated curiosity.

A drill bit previously used on one sample touches another target

The contamination risk depends on the scientific question. For bulk mineralogy the carryover may be negligible after cleaning; for trace organics it may invalidate the measurement. The crew records the contact event and decides whether to collect a new sample with a clean tool, downgrade the sample to another analysis, or preserve it with a contamination flag.

Traceability protects value even after a mistake. Hiding the event would create a stronger-looking but less trustworthy dataset. Scientific operations reward accurate provenance over perfect appearance.

A sample container seal is found partly open after transport

Identity can remain secure while environmental integrity is uncertain. The sample is isolated, the condition documented, and analyses sensitive to atmosphere, moisture or volatile loss are flagged. Other properties may still be usable. The team should not discard all scientific value nor pretend the seal failure had no consequence.

The container lot and closure procedure are reviewed because one partial opening may indicate a mechanical or human-factor pattern. Corrective action can include tactile confirmation, secondary indicators or improved glove-compatible closure geometry.

Two instruments give different mineral identifications

The team examines what each instrument actually measures and the uncertainty of the algorithms rather than choosing the preferred answer. Sample heterogeneity, calibration, spectral overlap or preparation can explain disagreement. A third method or targeted subsample may discriminate the alternatives.

The report preserves both results and the reasoning used to reconcile them. Scientific confidence should grow from converging independent evidence, not from suppressing inconvenient measurements.

A dust storm cancels exterior work but the campaign clock continues

Laboratory priorities can shift toward processing existing samples, reviewing metadata, maintaining instruments and preparing future routes. The team identifies analyses whose results can improve the next field decision once EVA resumes. This turns weather delay into useful campaign work rather than simply compressing all remaining field stations into fewer days.

If seasonal access is being lost, the science lead may drop lower-value targets. The revised plan records what question can no longer be answered and what minimum dataset still supports a useful conclusion.

A promising sample is too large for the available archive container

The crew should not fracture it casually if spatial texture or internal relationships are important. They can document the intact specimen, choose a representative subsample with orientation information, or defer collection for a larger container. The trade depends on what information is preserved by geometry versus material chemistry.

This is a reminder that sample hardware constrains science. Container sizes, cutting tools and mass limits should be reviewed against expected target types before the campaign.

A weekly campaign is consuming more crew-hours than planned

The science team decomposes the overrun into EVA preparation, field tasks, laboratory handling, data processing and maintenance. Some steps may be automated or simplified without reducing quality; others are irreducible controls that should remain. Removing blanks or metadata to save time would create apparent productivity at the cost of scientific validity.

Priority can shift toward fewer well-documented stations. Information per crew-hour is a legitimate planning measure when habitability and maintenance compete for the same people.

Earth requests an additional measurement after the crew has already left the site

The team determines whether archived samples or recorded imagery can answer the request. If not, the request becomes a candidate for a future traverse and is weighed against other priorities. Communication delay means Earth cannot assume a field site remains instantly accessible.

This encourages specialists to design decision trees before an EVA: what result would trigger which follow-up measurement? Preplanned branches let the local crew act while the site is still in front of them.

A sample shows large internal heterogeneity

One analytical value may not represent the whole specimen. The team documents visible domains, takes controlled subsamples and treats variability as part of the scientific result. Averaging without recording structure can erase evidence of multiple formation processes.

Sampling strategy can then expand to the outcrop scale: is the heterogeneity unique to the specimen or systematic across the unit? Field and laboratory observations inform each other iteratively.

A software update changes the processing pipeline mid-campaign

Derived results before and after the update may not be directly comparable. Version control records which algorithm produced each product, and representative raw data can be reprocessed through both versions to quantify the change. The original raw measurements remain preserved so interpretation does not depend permanently on one software state.

Operational science therefore treats code configuration as measurement metadata. A silent algorithm change can mimic a geological trend if provenance is weak.

A field note contains an interpretation later shown to be wrong

The original note should remain in the record with a later annotation rather than being rewritten as if the team always knew the correct explanation. Preserving the sequence of reasoning helps reviewers understand why samples were chosen and prevents hindsight from altering provenance.

Scientific archives distinguish observation from interpretation partly for this reason. A mistaken interpretation can be corrected; a lost original observation cannot be reconstructed.

A rare sample must be divided between immediate analysis and long-term archive

The allocation is based on which measurements are necessary to answer the current question and how much material future methods may need. Non-destructive characterization can guide where to cut, and subsample identifiers preserve orientation and parent relationship. The team explicitly records what fraction remains untouched.

This discipline protects against short-term pressure for complete analysis. On Mars, preserving material for a future instrument or Earth laboratory may be one of the highest-value scientific decisions the local crew makes.

Science campaign design dossier: traceability and laboratory continuity

Design dossier — design the campaign around discriminating evidence

A traverse should not maximize the number of stops; it should maximize the evidence that separates competing explanations. Before departure, the science team can create a matrix linking each station and sample to the hypothesis it tests. If two stations provide nearly identical evidence while another critical comparison is absent, the plan is unbalanced even if total sample count is high.

This matrix becomes valuable when operations force cuts. The crew can drop redundancy where it adds little while preserving the minimum set needed to discriminate the main alternatives. Scientific priority becomes explicit rather than based on whichever target appears most exciting in the moment.

Design dossier — preserve orientation and geometry when they carry information

Some samples are valuable not only for composition but for direction: bedding orientation, fracture fill, magnetic fabric or a contact between two units. Containers and labels should preserve the orientation marker or photographic relationship needed for later analysis. Breaking a sample into convenient pieces can erase the structure that made it scientifically important.

Field protocols therefore specify when orientation, top/bottom relationship or exact position must be recorded before removal. The extra seconds are justified when the information cannot be reconstructed after the specimen reaches the laboratory.

Design dossier — keep sample preparation zones logically separated

Cutting, crushing, sieving and chemical preparation can generate dust or residues that contaminate trace analyses. Laboratory layout should separate dirty preparation from sensitive measurement and clean storage. Tools can be dedicated by method or cleaned through validated procedures, and airflow or containment should prevent fine particles migrating between benches.

The laboratory chain is only as clean as its weakest transition. A pristine field sample can be compromised by a shared crusher or mislabeled vial. Facility design supports scientific integrity in the same way clean-room zoning supports hardware assembly.

Design dossier — use reference materials to track instrument health over months

Calibration is not only a pre-use action. Repeated measurements of stable reference materials can reveal gradual drift, detector aging or software changes across a long campaign. Plotting reference response over time allows the team to detect deterioration before it becomes obvious in unknown samples.

When an instrument is repaired or updated, the same reference set helps establish continuity between old and new configurations. Long-term science benefits from this metrological memory, especially when no Earth laboratory is immediately available to arbitrate discrepancies.

Design dossier — plan destructive analysis as a decision tree

Instead of assigning every sample the same sequence, the laboratory can use non-destructive results to decide which destructive test is justified next. A sample that already answers the campaign question may be archived; another with an unusual phase may be split for additional work. The tree protects mass while directing effort toward the highest information gain.

The decision and consumed quantity are recorded. Future researchers can then understand why material was used and what remains. This is particularly important for rare samples that may never be collected again.

Design dossier — reconcile local autonomy with Earth expertise

The crew needs enough scientific authority to act during communication delay without severing collaboration with specialists on Earth. Shared decision rules, campaign objectives and preplanned branches provide that balance. Earth can define what evidence matters; the local team decides how to obtain it safely under actual conditions.

After each sol, compressed summaries and raw data priorities can be transmitted for deeper review. The next cycle incorporates Earth feedback. Autonomy becomes a time-scale division of responsibility rather than a competition between local and remote scientists.

Design dossier — protect scientific clocks and timestamps

Correlating samples, imagery, environmental data and instrument logs requires consistent time references. Clock drift or mixed time standards can make it difficult to reconstruct whether an atmospheric change preceded a measurement anomaly. Systems should record the time basis and synchronize or cross-calibrate clocks used in campaign data.

When delay-tolerant communications reorder files, embedded timestamps and stable identifiers preserve chronology. Scientific operations depend on time traceability even when the underlying question is geological rather than navigational.

Design dossier — verify the campaign with an end-to-end mock sortie

Before a high-value field campaign, crews can rehearse the sequence using representative terrain or a local test area: planning, labeling, imaging, sample handling, transfer, laboratory receipt and metadata ingestion. The exercise often reveals practical gaps such as inaccessible labels, overloaded gloves, software fields that take too long or ambiguous container numbering.

Correcting these issues before the real traverse protects both EVA efficiency and scientific integrity. Verification should cover the whole evidence chain, not just whether the instrument produces a reading.

Final synthesis for mission qualification

Final synthesis — a sample is evidence plus context plus history

A sample has scientific value only to the extent that researchers can reconstruct what it is, where it came from, how it was selected, what happened to it and which measurements were performed. Field context, custody, preparation and instrument configuration therefore belong to the evidence itself. The most impressive analytical precision cannot repair a broken identity chain. Campaign quality should be reviewed from the final conclusion backward to the station, asking whether every important inference can be traced to preserved observations and controls.

Final synthesis — local scientific autonomy is a designed operating mode

Autonomy does not mean the Mars crew ignores Earth expertise. It means the mission has already agreed which decisions must be made locally before Earth can respond, which thresholds authorize extra sampling or repetition, and which uncertainties require conservative preservation instead of immediate interpretation. Earth specialists then work on a slower cycle using transmitted data and can redirect later traverses. This division of time scales allows science to remain both collaborative and operationally executable.

Final synthesis — campaign success is information gained without consuming future options unnecessarily

A good campaign leaves the settlement with stronger explanations, well-documented samples, preserved archive material and a clear record of what remains uncertain. It does not maximize specimen count or instrument hours. The team should be willing to stop collecting redundant material, postpone destructive tests and protect future opportunities when current evidence is already sufficient. Scientific discipline on Mars is partly the art of deciding what not to consume today.

Qualification notes

Qualification note — preserve negative results

A campaign should retain well-controlled negative measurements instead of discarding them because they appear uninteresting. A non-detection with known sensitivity can constrain hypotheses, guide future sampling and prevent later teams from repeating the same work. The archive should record detection limits, controls and sample context so absence of evidence can be interpreted correctly.

Qualification note — distinguish environmental change from instrument change

When a time series shifts, investigators should ask whether the environment changed, the sampling location moved, the instrument drifted or the processing pipeline changed. Reference materials, metadata and version control provide the evidence to separate those possibilities. This discipline is essential when long campaigns compare measurements across seasons and hardware maintenance.

Qualification note — close each campaign with unresolved questions

The final report should not merely summarize successes. It should list uncertainties that remain open, samples that deserve future analysis, missing comparisons and any operational limits that constrained interpretation. These unresolved questions become deliberate inputs to later traverses rather than being rediscovered from scratch.

Final evidence note

Scientific qualification should include enough reserve in time, containers, calibration material and archive capacity that an unexpected result can be checked rather than forcing the crew to choose between abandoning the finding and consuming the protected material budget. A campaign with no slack cannot investigate surprise, and surprise is often where the most valuable science begins.

Final archive note

Every campaign closure should verify that raw data, metadata, photographs, custody records and derived products are stored redundantly and linked by stable identifiers. The scientific result is not complete until another team can reconstruct the evidence chain without depending on the memory of the people who collected it.

Completion note

A final readiness review should confirm that the field plan, laboratory sequence and archive strategy still answer the same scientific question after all operational compromises have been incorporated. If route changes, instrument outages or lost samples have removed a critical comparison, the report should state that limitation explicitly rather than allowing the original hypothesis test to appear complete.

Interactive beginner glossary

These definitions emphasize traceability and decision use rather than dictionary wording.

• hypothesis — A proposed explanation that can be tested against observations or measurements.
• field station — A defined location where a planned set of observations or sampling tasks is performed.
• sampling plan — A documented strategy describing what will be collected, where, why and in what quantity.
• representative sample — Material selected to characterize a defined population, unit or condition rather than only an unusual feature.
• replicate — A repeated sample or measurement used to evaluate variability or repeatability.
• calibration — The process of relating an instrument response to a known reference.
• blank — A control expected to contain none of the target signal, used to detect contamination or background from the method.
• control — A comparison condition used to interpret the effect or signal being studied.
• standard — A reference material or value with known properties used to check measurement performance.
• metadata — Information describing the circumstances, configuration and provenance of a sample or data product.
• provenance — The documented origin and history of a sample or dataset.
• chain of custody — The traceable record of sample possession, handling, transfers and relevant conditions.
• subsample — A portion taken from a parent sample for a specific analysis or archive.
• archive fraction — Material deliberately preserved for future analysis rather than consumed immediately.
• non-destructive analysis — A measurement intended to preserve the sample substantially intact for later work.
• destructive analysis — A procedure that consumes or irreversibly alters some or all of the sample.
• instrument drift — A gradual change in measurement response unrelated to the true sample value.
• quality control — Checks used to detect measurement, handling or process errors and to judge data reliability.
• uncertainty — A quantified or described range of doubt about a measurement or interpretation.
• selection bias — Distortion caused by choosing samples in a way that does not fairly represent the question being studied.
• context image — A photograph showing a sample or observation in relation to its surrounding geological setting.
• station identifier — A stable code linking observations, images, samples and notes to one field location.
• trace contamination — A small introduced substance capable of confusing a sensitive measurement.
• planetary protection — Policies and practices intended to limit harmful biological contamination between planetary environments.
• sample integrity — The degree to which a sample preserves the identity and condition needed for its scientific purpose.
• field notebook — A structured record of observations, decisions, anomalies and interpretations made during field operations.
• raw data — Original measurement output before interpretation or derived processing.
• derived product — A result created by processing or interpreting raw measurements.
• campaign — A coordinated sequence of field and laboratory activities designed to answer a scientific objective.
• decision rule — A pre-agreed condition that authorizes or directs an operational choice when time or communication is limited.

Operational review checklist

• State a testable question before defining the traverse.
• Record context before disturbing or sampling a target.
• Make every planned sample serve an explicit comparison or hypothesis.
• Use blanks, standards, controls and replicates for different quality questions.
• Assign stable identifiers from field station through laboratory subsamples.
• Record instrument configuration and method version with every important result.
• Preserve chain of custody and environmental condition for sensitive samples.
• Sequence laboratory work to protect material needed for later or less reversible analyses.
• Define when surprising results require repeat measurement or an independent method.
• Give the local crew pre-agreed authority for time-sensitive field decisions.
• Protect archive fractions when future analysis may be more valuable than one extra current test.
• Plan crew-hours and EVA access around maintenance and safety obligations.
• Store raw data redundantly and preserve processing provenance.
• Label observation, inference and speculation separately in reports.
• Treat contamination control as both a scientific and operational design problem.