In observational astronomy, the observation arc (or arc length) of a Solar System body is the time period between its earliest and latest observations, used for tracing the body's path. It is usually given in days or years. The term is mostly used in the discovery and tracking of asteroids and comets. Arc length has the greatest influence on the accuracy of an orbit. The number, spacing of intermediate observations, and timestamps have a lesser effect. A very short arc leaves a high uncertainty parameter. The object might be in one of many different orbits, at many distances from Earth. In some cases, the initial arc was too short to determine if the object was in orbit around the Earth, or orbiting out in the asteroid belt. With a 1-day observation arc, was thought to be a trans-Neptunian dwarf planet, but is now known to be a 1 km main-belt asteroid. With an observation arc of 3 days, was thought to be a Mars-crossing asteroid that could be a threat to Earth, but was later found to be another main-belt asteroid. A relatively modest observation arc may allow finding an older "precovery" photo, providing a much longer arc and a more precise orbit. An observation arc less than 30 days can make it difficult to recover an Inner Solar System object more than a year after the last observation, and may result in a lost minor planet. Due to their greater distance from the Sun and slow movement across the sky, trans-Neptunian objects with observation arcs less than several years often have poorly constrained orbits. As a general rule objects discovered when they are currently farther from the Sun will have greater uncertainties in their initial orbits if the observation arcs are short. which was discovered when 100+ AU from the Sun and has only been observed 9 times over 2 years will require an observation arc of several years to refine the uncertainties in the orbital period and aphelion (farthest distance from the Sun). with only 4 observations over 1 day has uncertainties so large that the error bars are not really meaningful and just show that the uncertainties are very large.