From Science News:
The astronomical tables that Columbus consulted during his voyage proved useful for determining latitude and, to some degree, longitude. A prediction contained in the tables probably saved his life at a crucial moment during his fourth voyage to the lands he had discovered.
Nearly 2 years after sailing from Cadiz in 1502, Columbus and his restless, disgruntled crew were stranded on the north coast of Jamaica, confined to worm-eaten, leaking ships. The native inhabitants were no longer awed by the newcomers. Annoyed by their voracious appetites and angry at the depredations of crew members, who had plundered several villages, the population was hostile and would no longer supply food.
Weary and ill, Columbus had withdrawn to his ship. There, he pondered his precarious situation. Returning to the stained pages of the Ephemerides, he noted Regiomontanus’s prediction of a total eclipse of the moon on Feb. 29, 1504….
…Columbus had observed a lunar eclipse on an earlier voyage and had noticed discrepancies between the predictions made by Zacuto and those contained in the Ephemerides. Moreover, he had no reliable way of determining the correct local time of this particular projected eclipse. The times provided by Regiomontanus for its start and end were for Nuremberg, Germany.
Despite these uncertainties, Columbus was desperate enough to take a chance. On the day before the predicted eclipse, he summoned the leaders of the native inhabitants and warned them through an interpreter that if they did not cooperate with him, the moon would disappear from the sky on the following night.
The natives for the most part were unimpressed; some even laughed. Columbus nervously awaited the outcome of his gamble. Could he rely on tables that had been compiled several decades earlier and that predicted the positions of celestial bodies only for the years between 1475 and 1506? How large were the errors?