After Velara III: Yield and Cost of the Non-Carbon Life Screen in Federation Pre-Terraforming Surveys of 81 Candidate Worlds, 2364–2375

1. Introduction

Terraforming is permitted by Federation practice only on worlds that support no life. Before 2364 that condition was checked by surveys designed to find life as the Federation knew it best: carbon-based, cellular, water-dependent. On Velara III the survey found none, and a terraforming station had been working on the planet for some time when one of its engineers was killed in what proved not to be an accident. The planet was inhabited by an inorganic, crystalline life form, intelligent and able to communicate once the Universal Translator was given its patterns, living in a layer of saline water beneath the surface that the terraforming plan would have drained (Federation Terraforming Survey Office, 2364). The project was halted.

Velara III was not the first warning. In 2267 miners on Janus VI had destroyed many of what they took to be silicon nodules in the rock, which were in fact the eggs of the Horta, a silicon-based species that lived in the rock itself; the killing stopped only when communication was established with the surviving adult Horta and the miners agreed to spare its eggs (Tavish-Pell, 2301). In both cases, surveys looking for life in familiar forms had looked straight past life in unfamiliar ones.

After the Velara III report, late in 2364, the Federation terraforming authorities added an interim screen for non-carbon life to every pre-terraforming survey. The screen was formalised in 2365, together with an order for re-survey of worlds already cleared but not yet altered (Federation Terraforming Survey Office, 2365). This paper asks what that screen has found in its first eleven years, from late 2364 to 2375, how often its flags are borne out, and what it costs.

2. Methods

The screen adds three elements to the standard survey: a subsurface scan for structured crystalline or mineral matrices that are not explained by the planet's geology, a search for organised electromagnetic activity at the scales found on Velara III and Janus VI, and a period of observation of any flagged site for change over time (Qasimi, 2366). A world is flagged when any element gives a positive result that the survey team cannot explain geologically. Flagged worlds go to follow-up, in which a specialist team attempts contact with the Universal Translator and conducts close study of the flagged structures. Follow-up concludes that indigenous life is present, absent, or undetermined.

From the survey records we identified every world assessed for terraforming from late 2364 to 2375 (Federation Terraforming Survey Office, 2376). There were 58 new candidates surveyed under the revised protocol. There were also 23 worlds that had been cleared under the old protocol before 2364 but on which no terraforming had begun, and which were re-surveyed under the revised one. The order was limited to unaltered worlds because terraforming changes the subsurface conditions the screen reads, so that a screen of an altered world is hard to interpret. The 81 worlds make up our cohort. For each we recorded whether the screen flagged it, the outcome of follow-up, and the time the screen added to the survey, measured from the start of the screen to its conclusion or to the start of follow-up.

Our outcomes are counts of worlds, and we report proportions with exact binomial confidence intervals. We compared the new candidates with the re-surveyed worlds using Fisher's exact test.

3. Results

The screen flagged 9 of the 81 worlds: 7 of the 58 new candidates (12%) and 2 of the 23 re-surveyed worlds (9%; Fisher's exact test, p = 1.0). Follow-up confirmed indigenous non-carbon life on 3 of the 9, all of them in subsurface habitats, two in mineral matrices and one in a subsurface brine resembling that of Velara III. Two of the 3 were new candidates and 1 was a re-surveyed world, cleared before 2364, on which terraforming would otherwise have begun in 2366. Follow-up found no life on 6 flagged worlds, where the flagged structures proved to be geological. No follow-up ended as undetermined. Table 1 gives the results by group.

Confirmed life was found on 3.7% of all screened worlds (3 of 81; 95% CI 0.8–10.4%). The positive predictive value of a flag was 33% (3 of 9; 95% CI 7–70%). The rate did not differ detectably between new candidates (2 of 58, 3.4%) and re-surveyed worlds (1 of 23, 4.3%; Fisher's exact test, p = 1.0), although both groups are small.

The screen added a median of 41 days to a survey (interquartile range 29–58). For flagged worlds, follow-up added a further median of 7 months. On all three worlds where life was confirmed, the terraforming candidacy was withdrawn and the world placed under protection; on the six false positives, terraforming proceeded after follow-up.

4. Discussion

In eleven years the screen and its follow-up found three inhabited worlds, at least one of which the old survey had in fact cleared and which was due for terraforming. Each of those three worlds, if altered, would have been another Velara III, or worse, since not every such species will be able to make itself understood before its habitat is destroyed. Against that, the screen costs about six weeks per survey and a further seven months or so for each flagged world. Terraforming projects run for decades, and a delay of weeks at the start is small against their length and cost (Ferreira-Oakes, 2370).

Two in three flags were false. That is a tolerable rate for a screen whose purpose is to avoid a grave and irreversible error, but it bears on how flags are handled. Survey teams under pressure to clear a world may be tempted to explain away a flag geologically; the record shows that the correct response to a flag is follow-up, and that follow-up resolves the question in months.

Its sensitivity is the screen's greatest weakness. We know how many flags were right; we do not know how many inhabited worlds the screen cleared. No independent method exists to check a cleared world short of long observation, and the screen was designed around the two forms of non-carbon life found living in planetary crusts, at Janus VI and Velara III. Life in a third form might pass it as easily as the Velaran life passed the old survey (Okoro-Beck, 2372). The screen should therefore be treated as a floor, not a guarantee, and revised as new forms of life are met.

One further finding needs action now. One of the 23 re-surveyed worlds, cleared before 2364, proved to be inhabited. The re-survey order of 2365 covered only worlds not yet altered. Worlds cleared before 2364 and already partly terraformed have never been screened, and if one in twenty-three unaltered worlds was inhabited, some altered worlds may have been too. We recommend extending re-survey to every world cleared before 2364, whatever its state.

5. Limitations

Our cohort is small, and three positive worlds give imprecise estimates; the confidence intervals are wide. The screen's sensitivity cannot be estimated from these data at all. The survey records were kept by the authorities who designed the screen, and one of us designed its elements (Qasimi, 2366), so this is not an evaluation by outsiders. The time added was recorded by survey teams whose own procedures varied. Finally, several surveys were delayed or abandoned during the Dominion War; we included only surveys that were completed.