Search for a command to run...
Based on 5298 positional observations, we determined the radial, transversal, and normal components of the nongravitational accelerations A1, A2, and A3 within the framework of the Marsden model and we also estimated the corresponding quantities for the alternative dependences of the acceleration on heliocentric distance: 1/r0, 1/r2, 1/r3, and 1/r4. We have shown that all the nongravitational acceleration models considered, when determined simultaneously with the coordinates and velocities at the chosen epoch, yield almost the same level of agreement with observations—the only exception being the constant-acceleration model (1/r0). Notably, the derived values of nongravitational accelerations exceed those characteristics of both Solar System comets and the first known interstellar comet, 2I/Borisov. In every model considered, the relationship |A1|/|A2| < 1 holds, meaning that the total nongravitational acceleration deviates markedly from the sunward direction, since the radial and transverse components are mutually perpendicular. This can be attributed either to rapid rotation of the nucleus or to pronounced inhomogeneity of its surface. For the dependencies proportional to 1/r0, 1/r3, and 1/r4, as well as in the Marsden model, the value of the parameter A1 proves to be negative. An analysis of the O–C values (differences between the observed and calculated positions) revealed no statistically significant periodic variations. Using the inject-and-recover test, we estimated the upper limit on the amplitude of hypothetical periodic O–C variations, at which their period could be reliably recovered from the available observational data.