All UKIRT imagers are equipped with filters from the Mauna Kea Consortium (MKO-NIR) filter set, see "The Mauna Kea Observatories Near-Infrared Filter Set. I. Defining Optimal 1-5 Micron Bandpasses" by Simons & Tokunaga 2002 PASP 114, 169 and "The Mauna Kea Observatories Near-Infrared Filter Set. II. Specifications for a New JHKL'M' Filter Set for Infrared Astronomy" by Tokunaga et al. 2002 PASP, 114, 180. These filters have been designed to be a better match to the atmospheric windows, offering maximum sensitivity together with minimum dependance on altitude and atmospheric water-vapour content, and are also of excellent optical quality.

The MKO-NIR filter set has been in UIST and UFTI since commissioning in October 2002 and October 1998 respectively, and replaced the old JHK set in IRCAM/TUFTI prior to commissioning of its new plate scale in September 1999. Both the J and H filters are significantly different from the old filters and hence the magnitudes for the Faint Standards on the old IRCAM3 system are no longer on the natural systems of the imagers. Colour transformations have been derived in two ways: one empirically based on UFTI photometry and the other calculated by convolving the known filter profiles with spectroscopic data for a representative set of red stars. The two determinations agree well. Note that the transformations for IRCAM/TUFTI with the same filter set should be effectively identical.

The transformations between the old IRCAM3 system and the new MKO-NIR system at H and K are well behaved and single-valued. However for the J filter different terms have to be applied depending whether or not the standard star has intrinsic water absorption features. This is due to the fact that the new J filter cuts off shorter than the old filter, specifically to avoid water absorption in the terrestrial atmosphere. We have determined that for stars with no intrinsic water features the colour transformations are:

K(old) = K(new) + 0.020[+/-0.005](J-K)(new)
(J-H)(old) = 1.040[+/-0.010](J-H)(new)
(H-K)(old) = 0.830[+/-0.010](H-K)(new)
(J-K)(old) = 0.960[+/-0.010](J-K)(new)


K(new) = K(old) - 0.020[+/-0.005](J-K)(old)
(J-H)(new) = 0.960[+/-0.010](J-H)(old)
(H-K)(new) = 1.205[+/-0.010](H-K)(old)
(J-K)(new) = 1.040[+/-0.010](J-K)(old)

K(new) = K(old) - 0.020[+/-0.005](J-K)(old)
(J-H)(new) = 0.870[+/-0.010](J-H)(old)
(H-K)(new) = 1.205[+/-0.010](H-K)(old)
(J-K)(new) = 0.980[+/-0.010](J-K)(old)

The UFTI I-band system has been compared to Landolt Cousins-I standards (Landolt 1992 AJ 104, 340), and the Z system to the Sloan Sky Survey standards (Krisciunas et al. 1998 PASP 110, 1342). The Sloan Z values have been converted from an AB-system to our Vega=0mag system by subtracting 0.572mag from the Krisciunas et al. values. The following transformations were measured during three engineering nights in June 1999, August 1999, January 2000, March 2000 and December 2000.

I(UFTI) = I(C)    - 0.16[+/-0.01](V-I([C])
I(C)    = I(UFTI) + 0.14[+/-0.01](V-I[UFTI])

(I-Z)(UFTI) = 0.62[+/-0.05](I[C]-Z[S])
(I[C]-Z[S]) = 1.61[+/-0.10](I-Z)(UFTI)
Z(UFTI) = Z(S) - 0.34[+/-0.03](I[C]-Z[S])
Z(UFTI) = Z(S) - 0.21[+/-0.03](Z[S]-J[UKIRT])
I(UFTI) = I(C) - 0.72[+/-0.03](I[C]-Z[S])
I(UFTI) = I(C) - 0.27[+/-0.03](I[C]-J[UKIRT])

Z(S) = Z(UFTI) + 0.55[+/-0.03](I-Z)(UFTI)
Z(S) = Z(UFTI) + 0.27[+/-0.03](Z-J)(UKIRT)
I(C) = I(UFTI) + 1.16[+/-0.10](I-Z)(UFTI)
I(C) = I(UFTI) + 0.37[+/-0.04](I-J)(UKIRT)