No Earth-shadow deficit in 107,608 POSS-I transients (VASCO-ML 8.5° anti-Sun cone; Palomar umbra at 15,000–60,000 km) when per-plate transient counts set the expected rate

Two studies report fewer POSS-I transients in the Earth's shadow than expected. They interpret this as evidence for sunlit objects in orbit before 1957. We tested this claim with the public VASCO-ML catalog of 107,875 candidates. We assigned 107,608 transients to 635 POSS-I red plates. For each plate, the expected count in the shadow is the number of transients on the plate multiplied by the fraction of the plate footprint that the shadow covers. We used five footprint models and a plate bootstrap. In the VASCO-ML 8.5° anti-Sun cone, we observe 501 transients and expect 510.8 (O/E = 0.981, 95% interval 0.80–1.23). The published rate (0.692%) gives each plate the same weight. This overestimates the expected count by 42–52%. Placebo cones show that the equal-weight statistic is not calibrated (placebo O/E 0.37–2.79). In the topocentric umbra seen from Palomar at 15,000–60,000 km, O/E is 0.94–1.09 for all footprint models. At 42,164 km, O/E = 1.07 (95% interval 0.96–1.54). A within-plate permutation test of the highest-score decile (D9) gives contrast ratios of 0.89–1.03 (one-sided p ≥ 0.13). The released catalog does not reproduce the published decile table (D9 in the cone: 22, not 31). We also found that the POSS-I DATE-OBS field 'HH:XX' holds decimal hours. The STScI full-plate headers read it as minutes, which gives time errors of up to 39 min. Limitations: few plates contribute at large distances. At 42,164 km, the 95% intervals include deficits of up to 4% (empirical footprint) or 13% (other footprints). We used the released catalog and did not examine the plate images. The supporting ZIP holds all code, derived data, run logs and figures.

POSS-I · Palomar Observatory Sky Survey · Earth shadow · umbra · transients · vanishing sources · VASCO · photographic plates · Digitized Sky Survey · DATE-OBS · plate observation time · geosynchronous distance · glints · placebo test · plate bootstrap · permutation test · reproduction · null result · pre-Sputnik