nuArc and other graphic arts integrators aren't calibrated in any sense other than they are set to count down at around 1 second/unit with a fully warmed new lamp. A 1KW unit counts down at 1 second/unit and a 6KW unit counts down at 1 second/unit - a 6:1 variation in what a unit is equal to. However, given that lamps are consistent when new it is possible to transfer exposure information between like models.
Sunlight is very constant - 1 KW/square meter is known as 'the solar constant'. Of course it changes with time of day - but between, say, 10:00 am and 4:00 pm [more in summer] exposure times will be very consistent. Be sure to keep the printing frame pointing directly at the sun. Clouds attenuate UV less than they do visible light so exposures will be about the same with hazy clouds. Moving clouds and heavy cloud cover can be problematic. You can, as Bill mentioned, correlate exposure meter readings to UV exposure times for cloudy days even though an exposure meter doesn't read UV.
UV monitors for sunbathing aren't very accurate - they pretty much tell you the UV is high, medium or low, something one can see for oneself. I tried using one for cyanotypes and the results were pretty miserable.
Interestingly the ratio of UV to visible light is highest when the sun is just below the horizon - that isn't to say there is a lot of UV, just as there isn't a lot of visible light. The UV is scattered from the sky overhead and it is also polarized. The combination makes for very saturated colors when photographing flowers.