• What FCCT14/EFCC18 is (and why you should use it)
• Using FCCT14/EFCC18 in Find_Orb
• Getting the debiasing data
A lot of astrometry (especially that before about 2016) is referenced to star catalogs that have systematic biases in positions and proper motions. If you know which catalog was used to reduce a given observation, and where that observation was in the sky, it's possible to determine what those biases are, and to remove them. The resulting unbiased positions can get you a considerably better orbit solution.
More recently, much astrometry is being referenced to the Gaia star catalog. Such astrometry is, for practical purposes, bias-free. But that leaves you with a lot of older astrometry reduced to older catalogues, some of which had rather substantial biases.
FCCT14 is described by Farnocchia, Chesley, Chamberlin, and Tholen in Icarus 245 (2015) 94-111, http://adsabs.harvard.edu/abs/2015Icar..245...94F. Essentially, they broke the sky up into 49152 = 3 * 2^14 equal-area, almost-square "pixels", each covering a little under a square degree. The pixels were determined using the HEALPix scheme for dividing the sky into "pixels". Then they determined the biases for 19 different catalogs relative to a subset of stars from the PPMXL catalog, on a pixel-by-pixel basis.
EFCC18 is described by Eggl, Farnocchia, Chamberlin, Chesley, in an arXiv article. This time, the debiasing was done on 26 catalogs, relative to the Gaia-DR2 catalogue which has come out in the interim. EFCC18 should be used in preference to FCCT14. It both corrects for more catalogues and does so to a slightly better catalog.
While the debiasing tables were far from straightforward for the authors to put together, it makes debiasing rather simple for orbit determination software such as Find_Orb. It determines in which of those 49152 "pixels" a given observation falls, gets the 19 or 26 possible catalog biases, and applies the bias for the catalog used for our given observation.
At present, there is no user-friendly method for turning debiasing on. You have to edit the file environ.dat in the Find_Orb folder, and change the line
BIAS_SCHEME=0
to equal 1. You'll know this is working because, the next time you load up observations in Find_Orb, you'll be told that you don't have the file containing the debiasing tables. (And later, you'll see the biases in question shown in the "observation info" bit of the dialog, whenever you click on an observation for which bias data exists.) Which leads us to...
If you don't already have it, Find_Orb will tell you that you need the file bias.dat. The EFCC18 version of this can be extracted from https://ssd.jpl.nasa.gov/ftp/ssd/debias/debias_2018.tgz (about 9.6 MBytes, with bias.dat decompressing to about 37 MBytes). This file contains the needed bias.dat file plus two other files, README.txt and tiles.dat; you can ignore both of these. You just need bias.dat.
Once this file is in your Find_Orb directory (for Windows) or ~/.find_orb directory (for Linux, *BSD, or OS/X), you'll no longer get the warning that Find_Orb can't locate the debiasing data.