<!-- ARTICLE_BODY_START --> On October 27, 2026, the European Space Agency published an image announcement with an unusually consequential distinction: Hera could now resolve Didymos and Dimorphos as two separate asteroids. Seeing the smaller member separately was a step toward investigating what NASA’s DART impact had done there in September 2022. ESA still described detailed examination of the altered asteroid as work ahead. An approach image should not be mistaken for a completed reconstruction of the impact site. ESA’s dated Hera update.
The distinction matters because planetary defense has moved beyond a purely hypothetical question. A spacecraft can hit an asteroid and measurably alter its motion. The harder question is how confidently that successful experiment can be translated to a different body, with a different internal structure, on a different trajectory, discovered with a different amount of warning.
An asteroid does not arrive with a specification sheet. Its apparent brightness is not its mass. A picture of a boulder field does not reveal all the voids beneath it. A precise orbit is not a precise prediction of what an impact will accomplish. Even an accurately delivered impulse must be applied in a useful direction and early enough to change the eventual encounter with Earth.
That makes the evidence after a deflection unusually important. The disappearing spacecraft supplies an intervention; telescopes, radar, navigation data and physical models determine its consequences. A complete defense would have to connect the two without allowing an impressive image, or one successful number, to stand in for the entire chain.
This article examines that chain rather than treating DART as a rehearsal for a cinematic last-minute rescue. The dated observational results are real. The simplified calculations below are explicitly illustrative. No currently threatening asteroid is inferred from the examples, and no future Hera measurement is presented as if it has already happened.
Two orbits, two different measurements
Dimorphos travels around Didymos while the pair travels around the Sun. Those motions create two distinct opportunities to measure a change. The local orbit is useful because repeated passages and eclipses provide a comparatively rapid clock. The heliocentric orbit answers a different question: has the path of the system through the Solar System changed?
Thomas and colleagues reported in 2023 that the impact shortened Dimorphos’s orbital period by 33.0 minutes, with a quoted uncertainty of 1.0 minute at three standard deviations. Their result combined Earth-based lightcurve and radar observations, and two analysis approaches agreed. This was a measured change in the moonlet’s period around its companion, not a claim that the entire system had been moved 33 minutes ahead along its solar orbit. Primary orbital-period study.
In March 2026, a further result concerned the larger trajectory. Makadia and colleagues measured an along-track change in the system’s heliocentric velocity of minus 11.7 plus or minus 1.3 micrometers per second, with the uncertainty reported at one standard deviation. Their analysis used stellar occultations and other astrometry, including radar. It estimated a heliocentric momentum enhancement factor of 2.0 plus or minus 0.3. Primary heliocentric-deflection study.
There is no contradiction in a conspicuous change to a small moon’s local orbit and a very small change to a much more massive system’s solar motion. Think in terms of the object whose momentum is being counted. Moving one member relative to the other and changing the motion of their shared center of mass are different measurements. A result can be large relative to one clock and tiny relative to another.
For a future defensive mission, the relevant comparison would be the change in the predicted Earth encounter, not whichever intermediate quantity looks largest in a headline. A shorter local period would be interesting but insufficient by itself. The mission assessment would need to explain how the measured intervention changes the trajectory of all material that could matter at the encounter.
Keeping these two orbits separate also prevents a common numerical mistake: comparing an uncertainty from one paper with the central value from another, as if both were estimating the same thing under the same assumptions. Their reference frames, mass boundaries and observational sensitivities differ. The units alone do not identify the physical question.
An impact is a momentum budget
The simplest picture is a spacecraft striking a body and remaining embedded in it. Before the collision, the spacecraft carries momentum. Afterward, the combined material must carry the corresponding total momentum, once any external forces over the collision interval are accounted for. This gives a useful starting point, but an asteroid impact is not usually a neat sticking experiment.
Material can leave the target. Its momentum matters even when its mass is a small fraction of the asteroid. If ejecta travels preferentially opposite the direction in which the target is pushed, the remaining body receives an additional recoil. Momentum has not been created. It has been divided between the remaining asteroid and departing debris.
Cheng and colleagues’ 2023 analysis inferred a decrease of 2.70 plus or minus 0.10 millimeters per second in Dimorphos’s along-track velocity. Their momentum enhancement estimate depended strongly on assumed mass: a bulk-density range of 1,500 to 3,300 kilograms per cubic meter yielded enhancement values from 2.2 to 4.9. The often quoted value near 3.6 assumed equal densities of 2,400 kilograms per cubic meter for the two bodies. It was a conditional estimate, not a universal property of asteroids. Primary momentum-transfer study.
Consider an invented, deliberately simplified calculation. A 500-kilogram impactor moving at 6,000 meters per second relative to a target has three million kilogram-meters per second of momentum. If a five-billion-kilogram target absorbs that momentum in a single aligned direction, its velocity changes by 0.0006 meters per second, or 0.6 millimeters per second. If recoil doubles the useful momentum, the corresponding change becomes 1.2 millimeters per second. These are arithmetic examples, not DART reconstructions or mission designs.
The denominator explains why mass knowledge matters. Doubling the target mass halves the velocity change for the same delivered momentum. A spacecraft does not become ineffective merely because the target is heavier than expected, but its achieved deflection may fall outside the planning margin. Conversely, underestimating recoil can leave a mission with more trajectory change than anticipated. Either outcome has to be propagated through the encounter geometry rather than judged by impact energy alone.
Energy remains important for excavation, heating, crushing and fragmentation. Yet two collisions with the same energy need not provide the same useful momentum in the desired direction. An energy comparison cannot replace the directional momentum budget that a deflection calculation requires.
Why the enhancement factor needs a definition
A single Greek letter can conceal several physical choices. Enhancement relative to what baseline? Measured along which direction? Counting material that escapes the small moon, or only material that escapes the entire binary? A useful number must answer those questions before it can be transferred into a new model.
The 2026 heliocentric study explicitly distinguishes the momentum associated with the mutual orbit from the momentum associated with material escaping the binary system. It also reports a revised mutual-orbit enhancement estimate of 2.1 plus or minus 0.3. The authors explain that incorporating their lower density estimate makes the new and earlier mutual-orbit results statistically consistent. Reading the newer value simply as a refutation of the old one would discard the earlier density assumption. Definitions and comparison in Makadia and colleagues.
Here is an original bookkeeping illustration. Imagine a stone pushed off a small moon but left orbiting the larger body. It can change the moon’s motion without carrying its momentum permanently out of the whole binary. Now imagine otherwise identical material escaping the entire system. The second case changes the momentum balance of the surviving binary in a different way. Both events involve ejecta; their contribution depends on the boundary drawn around the system.
The same care applies to direction. A large recoil component perpendicular to the desired correction is not equivalent to the same component aligned with it. If a planner records only the magnitude, a favorable-looking estimate may hide an unfavorable orientation. The target surface, the incoming trajectory and the ejecta plume establish different directions that cannot automatically be collapsed into one arrow.
For this reason, a defensible comparison between simulations should preserve the vector information until the final application is specified. A table headed only “efficiency” is too compressed. It should say which component has been extracted and which assumptions connect that component to the observable orbit change.
This is also a lesson in scientific updating. A later result can sharpen mass, geometry or boundary conditions without erasing the earlier demonstration. Progress often means replacing a convenient assumption with an observation. The public account should show what changed in that replacement, rather than setting up a contest between two bare numbers stripped of their definitions.
The target participates in the experiment
The impactor is engineered; the asteroid is not. Its strength, porosity, blocks, cavities, rotation and shape affect what happens when it is struck. A material response inferred for one target should therefore enter future planning as evidence with a domain of applicability, not as a fixed coefficient stamped onto every object.
Raducan and colleagues’ 2024 simulations found that weak, low-cohesion models could reproduce features of the DART observations. Their preferred interpretation suggested global deformation and resurfacing, potentially leaving a reshaped asteroid rather than a simple crater. This is a model-based interpretation constrained by observations, not a direct view inside Dimorphos. Primary simulation study.
To see why the distinction matters, imagine two equally massive targets. One concentrates material in a relatively coherent body. The other distributes it among blocks with open space between them. Even if their outside dimensions are similar, an impact can spend different fractions of its energy crushing voids, accelerating loose material, breaking strong rock or rearranging the whole structure. Equal mass and equal impact speed do not uniquely specify the ejecta.
This illustrative comparison does not establish which target would be easier to deflect. That would require a quantitative material model and a defined impact geometry. Its purpose is to identify why a surface photograph cannot close the calculation. A model that matches one observable may still describe the interior incorrectly.
A particularly useful test asks whether several observables can be matched together. A simulation might reproduce the period change but predict the wrong plume direction. Another might reproduce the bright dust while requiring an implausible bulk density. The best explanation should account for independent aspects of the event with compatible parameters, not tune a separate invisible asteroid for each dataset.
There is also a practical boundary between characterization and delay. Additional information is valuable when it changes a decision or its margin. An unlimited reconnaissance program is impossible during a finite warning period. The difficult engineering question is which unknowns most strongly affect the chosen action, and which measurements can reduce those unknowns soon enough to be used. Studying material response is therefore part of scheduling, as well as part of geology.
The bright plume is not a weighing scale
Dust is easy to see when sunlight illuminates a large total surface area. That does not mean the brightest part of a plume carries most of its mass or most of its momentum. A collection of tiny grains can dominate scattered light while larger, darker or less favorably illuminated pieces carry a different fraction of the mechanical budget.
Li and colleagues followed the ejecta with Hubble from about fifteen minutes after impact to eighteen and a half days afterward. Their observations showed an evolution influenced initially by the binary’s gravity and later by solar radiation pressure. This supplied evidence about how debris develops into an asteroid tail, rather than a direct census of every fragment. Primary Hubble ejecta study.
A simple geometric argument helps explain the inference problem. For identical spherical grains of radius r, mass scales with r cubed while projected area scales with r squared. At a fixed total mass and material density, smaller grains supply more total area. Halving the grain radius doubles the area per unit mass in this idealization. Optical behavior is more complicated than pure geometry, but the example shows why brightness cannot be translated into mass without assumptions about particles.
Momentum adds another variable: velocity. A hypothetical kilogram moving at one meter per second carries the same scalar momentum as ten kilograms moving at a tenth of that speed, if both are aligned. Their brightness could be very different. A bright image alone does not identify the momentum distribution, and a visually dramatic tail can be shaped by forces acting after the initial launch.
Time-resolved observations help because different motions leave different patterns. Material that rapidly leaves the field, material that stays near the system and material gradually sorted by sunlight are not interchangeable samples. However, any observing campaign has finite resolution, exposure time and sensitivity. Invisible material should not be silently assigned zero mass.
For planetary defense, the right question is consequently narrower than “How spectacular was the impact?” It is whether the observed debris, the orbital response and the physical model support a consistent account of momentum. A spectacular plume is valuable evidence. It needs to be combined with the less photogenic measurements that establish what the remaining body actually did.
Measuring an orbit without watching a complete outline
Astronomers can obtain useful orbital information even when an asteroid is unresolved. Repeated changes in brightness, measured positions against background stars and radar observations constrain different aspects of the geometry. None is a magic replacement for the others. Their value comes from how they reduce different ambiguities.
A lightcurve is a record of brightness over time. In a binary, mutual events can create features tied to orbital motion. But brightness also depends on rotation, shape, illumination and viewing direction. A fitting model has to distinguish these contributions. Otherwise a changed appearance could be assigned too confidently to a changed orbit, or a real orbital effect could be absorbed into an overly flexible brightness model.
The general measurement logic can be illustrated without reproducing a particular observing pipeline. Suppose a repeating event is timed to within a minute on one night. If the event is observed again after many cycles, the accumulated phase difference can constrain the average period more tightly than a single cycle does. But the analyst must count the cycles correctly and account for any actual evolution in the period. More elapsed time is helpful only when the model remains connected to the observations.
Stellar occultations offer another kind of constraint. In its March 6, 2026 report, NASA/JPL described the use of twenty-two occultations recorded from October 2022 through March 2025, combined with other observations, to measure the system’s changed solar orbit. These events occur when the asteroid passes in front of a star. The reported solar-period change was 0.15 seconds. NASA/JPL’s dated account.
An occultation is not a photograph of the whole orbit. It anchors a specific geometrical encounter between a line of sight, a star and the asteroid’s silhouette. Its timing, observer location and interpretation must be carried into the orbit solution. Several well-placed constraints can be more informative than many measurements that repeat nearly the same geometry.
This is why follow-up is a designed part of a deflection experiment, not an optional appendix after impact. The intervention produces a changed trajectory. Observation turns that trajectory into a measured result, and a measured result into a claim that can be independently checked.
The orbit is an estimate with structure
NASA/JPL’s Sentry documentation describes orbit determination as fitting observations with a dynamical model. There is a best-fitting solution, but also a region of other acceptable solutions because the observations are imperfect. New observations constrain that region and require an updated fit. The nominal path should therefore never be treated as the only physically plausible path. Sentry’s technical introduction.
For an original analogy, imagine trying to infer a road from two distant photographs of a moving vehicle. The photographs may pin down direction across the images while leaving distance along the line of sight less certain. A third photograph taken later, from a different viewpoint, can resolve an ambiguity that ten nearly simultaneous photographs would not. Asteroid orbit determination is far more sophisticated, but observation geometry remains central.
Uncertainty also evolves. A small range of starting conditions may spread differently in different directions as the trajectories are propagated. Near a strong gravitational encounter, slight differences can produce substantially different later paths. Representing the uncertainty only as a circle around an asteroid icon can hide this geometry.
A hypothetical defense assessment should therefore ask whether the uncertainty region misses Earth with an adequate margin after the intervention, not merely whether the nominal line does. It should also ask which parts of that conclusion come from observation and which come from assumed impact performance. The orbit uncertainty before impact and the uncertainty in the delivered velocity change are different inputs. Combining them does not make either disappear.
There is a corresponding communication challenge when an estimated impact probability rises after additional observations. Better data can concentrate the allowable trajectories nearer a possible collision before later data exclude it. An increase need not mean the asteroid physically changed course toward Earth. It can reflect a sharper description of a path that was previously poorly known.
That statement is a general inference about probability and measurement, not a report of a specific current threat. Any real public warning needs its own date, object designation and responsible assessment. Repeating an old probability without its observational context is especially misleading in a field where updates are the expected behavior of the system.
Warning time is an engineering resource
The familiar claim that a small push can matter if applied early contains a useful intuition. A velocity difference accumulates a position difference over time. However, a straight-line multiplication is only a teaching device for an orbiting body. Gravity, orbital phase and the geometry of the later Earth encounter determine how the actual correction develops.
For a deliberately simplified example, maintain a relative velocity of one millimeter per second for ten years, treating a year as 365.25 days. Multiplying gives about 316 kilometers of displacement. Over one year the corresponding number is about 31.6 kilometers. These are not predicted asteroid miss distances. They ignore orbital dynamics and show only the scale of linear accumulation.
The limitation is essential. A change along an orbital track can change the period and therefore the phase at a later encounter. A change in a different direction can have another effect. The same impulse applied at two orbital locations need not produce the same separation when Earth arrives. A mission planner needs trajectory propagation, not a slogan about how many years remain.
Warning time must also be divided among tasks. Discovery is followed by confirmation, orbit refinement, physical characterization, decision-making, mission preparation, launch opportunities, flight and observation of the outcome. Ten years until a possible encounter does not mean ten years of post-impact drift. If an intervention occurs after a long preparation and cruise, the useful accumulation interval is shorter.
This accounting creates tradeoffs rather than one obvious rule. Waiting for better measurements may avoid an unnecessary mission or improve targeting. Waiting too long may remove a favorable launch opportunity or the possibility of a second attempt. Acting early with broad uncertainty may require a more conservative mission design. The optimal balance depends on information that changes over time.
A useful public timeline would therefore show several clocks separately: time to the encounter, time to the next observing window, time to a decision that preserves launch options and time needed to evaluate a result. Compressing all of these into one countdown conceals the actual choices. Planetary defense is partly the management of those overlapping deadlines.
Why slowing down can shorten a period
The language of orbital change can sound paradoxical. How can slowing an object in its orbit make it complete an orbit sooner? The answer is that a gravitational orbit is not a fixed circular track on which a vehicle simply chooses its speed. A velocity change alters the orbit itself.
In an ideal two-body example, a body on a circular orbit receives a small impulse opposite its direction of travel. Immediately after the impulse, its position is unchanged but its speed is lower. The original circle is no longer the matching orbit. The body enters an ellipse with a smaller semimajor axis, and the period of that ellipse is shorter. Its subsequent speed varies around the ellipse. Describing the entire result as merely moving more slowly around the old circle would impose a track that gravity does not provide.
For a small perturbation, the standard two-body relation between period and semimajor axis gives a useful sensitivity: the fractional period change is approximately one and a half times the fractional semimajor-axis change, if the gravitational parameter remains fixed. Thus an invented one-percent reduction in semimajor axis corresponds, to first order, to a roughly one-and-a-half-percent reduction in period. This is a mathematical illustration, not a fit to the post-impact Didymos system, whose changing shapes and coupled motions require more detailed treatment.
This example explains why an orbital-period measurement is informative but does not directly identify every component of an impulse. Different directions can alter different orbital elements. A period strongly constrains orbital energy under the model, while inclination, orientation and eccentricity require other information. The measured number should be interpreted through the equations that connect it to the intervention.
It also clarifies why a mission cannot be described by a cartoon arrow alone. An arrow needs a position and time on the orbit. The same small velocity change introduced at a different orbital phase starts a different new trajectory. For a defensive application, the important evaluation occurs when the asteroid and Earth could meet, not immediately after the arrow is drawn.
Size uncertainty becomes mass uncertainty quickly
A second original calculation shows why a moderately uncertain diameter can create a much wider mission-performance range. For an ideal sphere of uniform density, mass is density multiplied by four-thirds pi times radius cubed. If the diameter estimate rises by twenty percent while density stays fixed, the mass rises by a factor of 1.2 cubed, or 1.728. That is nearly a seventy-three-percent increase, not a twenty-percent increase.
If the delivered useful momentum is held fixed in the same illustration, the resulting velocity change falls to about fifty-eight percent of its previous value. This is not a special asteroid effect; it follows from geometry and momentum. The importance for mission planning is that a seemingly modest change in a reported size can substantially change the inferred deflection margin.
Real bodies introduce additional uncertainty. An elongated object cannot always be represented adequately by one apparent diameter, especially when observations sample limited rotational phases. Bulk density includes internal void space as well as material composition. A spectrum can help identify surface characteristics without supplying a direct total mass. None of these facts makes characterization futile. They tell the analyst which quantities are actually being measured and which are being inferred.
A useful uncertainty study would vary shape and density together where the evidence makes them dependent, rather than combining every extreme as if all combinations were equally plausible. It would also distinguish uncertainty in the measured size from uncertainty in whether a spherical approximation is adequate. Those are different limitations and may respond to different observations.
For readers, the practical question is simple: when a forecast says a particular impactor can move a particular asteroid, what mass distribution was used? A single diameter in a press release is not enough to reconstruct that assumption. Showing the range turns an otherwise opaque performance claim into a testable statement about the evidence and the model.
Reaching the asteroid is different from pushing it usefully
An impact trajectory must satisfy more than a collision condition. The spacecraft must reach the right object at the right time, with an approach direction that produces a useful change. Payload mass, launch energy, travel time and impact speed are coupled by the mission trajectory. They cannot all be independently increased by choosing more favorable numbers in a spreadsheet.
An original comparison makes the distinction clear. Imagine two candidate missions to the same hypothetical target. Mission A arrives sooner but delivers a smaller useful impulse. Mission B delivers a larger impulse but arrives much later. A ranking based only on impact momentum would select B. A ranking based only on time available afterward would select A. Neither ranking is sufficient without propagating both outcomes to the encounter and considering their uncertainties.
Now add a characterization uncertainty. If the target’s mass could vary by a factor of two across plausible models, the predicted velocity change also varies. If the ejecta direction is uncertain, the useful component varies again. A design with the largest nominal correction may have a weaker lower-bound performance than a less ambitious design whose arrival geometry is better constrained.
This does not imply that every uncertainty must be eliminated. Engineering proceeds with margins. But the margin should be attached to the intended outcome, not just to a component specification. Extra spacecraft mass is not automatically an adequate substitute for uncertainty in target properties, and greater impact speed can affect the material response rather than merely scale the result.
Navigation and impact-site selection add another layer. The center of the visible illuminated shape need not coincide with the center of mass. Local terrain can affect the ejecta direction. An approach image is therefore both a navigational measurement and a physical observation, with limitations imposed by lighting and resolution.
The decision-relevant product is an ensemble of plausible post-impact trajectories tied to explicit target and mission assumptions. It should explain which variations would still produce an acceptable outcome, which would require follow-up and which would leave the mission outside its intended operating range. A nominal success animation cannot supply that information by itself.
A second look can change the next decision
The spacecraft that makes the impact cannot continue observing from the impact point. This basic asymmetry makes independent follow-up central to the design. Some measurements can be obtained remotely; others may benefit from a separate spacecraft. The appropriate mix depends on the target, the warning interval and the uncertainty that matters most.
An observer should be selected for a question, rather than added merely because more cameras sound reassuring. Is the priority the immediate plume direction, the target’s mass, a later orbital solution, the surface shape or the existence of significant fragments? Different measurements become available on different timescales. A camera well placed for the flash may not provide the best long-term trajectory constraint.
Consider a hypothetical mission that clearly hits its target but whose initial measured deflection is uncertain. There are at least three possible situations. The actual change may be too small. The change may be adequate but poorly observed. Or the physical model may be misassigning part of the signal. Those situations call for different next steps, even if all three initially produce an unsatisfactory confidence interval.
The observation plan should be designed before launch to distinguish them. It can specify what measurements would count as evidence of adequate deflection, when they are expected and what decisions remain possible if the measurements are delayed. Cloud cover, solar geometry, radio availability and instrument performance are ordinary constraints, not exceptional excuses discovered afterward.
Independent analysis also matters. Two codes fitting the same measurements can reveal implementation mistakes, but they may still share assumptions about forces, measurement errors or the impact. Agreement is strongest when different observing methods and different model choices converge. A shared hidden bias does not become independent evidence because two computers reproduce it.
The practical aim is not to demand perfect certainty. It is to obtain enough well-characterized evidence to choose the next action. If a second intervention is contemplated, its planning should use the updated trajectory rather than the original nominal mission story. Every observation after the first impact potentially changes both the required correction and the time available to deliver it.
A defense exercise is not a discovery notice
NASA/JPL’s 2025 Planetary Defense Conference exercise presents a hypothetical asteroid and staged updates for training. Its fictional designation and scenario dates describe an exercise, not a real asteroid announcement. The scenario deliberately includes uncertain physical properties, changing impact probabilities and limited observing opportunities. The clearly labeled exercise materials.
Such exercises are valuable because a real decision would not arrive as a finished paper with all error bars settled. Participants have to work with information available at a particular moment, while later observations may change the picture. The exercise can reveal whether organizations understand the difference between an uncertain orbit, an uncertain size and an uncertain intervention outcome.
The boundary between scenario and reality must remain conspicuous when material is quoted outside the room. A date and an asteroid-like designation can look authoritative even when copied from a fictional briefing. Removing the exercise label can transform a planning tool into a false alarm. Responsible coverage should preserve the label in the sentence that introduces the scenario, not hide it in a distant footnote.
Exercises can also examine who needs which information. A mission team may need a covariance matrix and a target-shape model. Public officials may need an explanation of which decisions become irreversible before the next update. Communities may need clear statements about what is known, what is hypothetical and who is responsible for future notices. These are different presentations of the same evidence, not opportunities to simplify away uncertainty.
There is a limit to what an exercise proves. Successful discussion does not demonstrate a ready launcher, a completed spacecraft or a tested command chain under actual conditions. It demonstrates that certain decisions and communications have been rehearsed. Its findings should feed technical preparation without being described as an operational interception capability.
The same distinction applies to DART in the opposite direction. DART supplied real physical evidence, not merely a tabletop discussion. But that evidence still has to be combined with detection, characterization and mission readiness before it amounts to a defense for a specific future case.
What should count as a dependable capability?
A useful assessment begins by separating demonstrated performance from expected performance. Demonstrated performance includes the measured consequences of a specific impact under specific conditions. Expected performance includes what models predict for another target. The bridge between them is validation: testing whether the models explain observations and whether the new target lies within a range where those explanations remain credible.
A single successful experiment can be scientifically transformative without providing a universal guarantee. This is familiar in other engineering settings. Testing one bridge component establishes something real about that component and loading condition. It does not license every bridge geometry made from vaguely similar material. The relevant question is how the experiment constrains the broader design space.
For asteroid deflection, useful validation would preserve uncertainties in target mass, material response and encounter geometry. It would identify which conclusions depend on an assumed ejecta direction, which depend on direct orbital measurement and which could be revised by better shape or gravity information. This makes an assessment more informative than a binary declaration that the technology either works or does not.
The evidence should also distinguish a successful collision from a successful protective outcome. A miss of the intended target is a delivery failure. A hit with less momentum transfer than expected is a performance problem. A well-measured deflection in an unhelpful direction is a planning problem. An apparently adequate deflection that cannot be verified is an information problem. All deserve attention, but they are not the same failure mode.
Finally, capability includes the ability to learn after acting. Observation resources, analysis expertise and a preserved record of assumptions determine whether an unexpected result becomes useful information in time. A mission whose evaluation depends on unavailable measurements may leave decision-makers with less practical confidence than its hardware success suggests.
The October 2026 Hera approach is compelling because it continues that process of learning. It connects a completed intervention with a closer examination still to come, without requiring the unknowns to be pretended away. The defensible conclusion is already substantial: a kinetic impact changed an asteroid’s motion, and multiple kinds of observation have measured the consequences. The next challenge is to turn that knowledge into predictions with enough margin, enough warning and enough follow-up to protect a world rather than merely demonstrate a maneuver. <!-- ARTICLE_BODY_END -->