I purchased and installed a Davis Vantage Pro2 system in May 2011, then added solar and UV sensors by separate purchase in January 2012.
When I occasionally look at nearby stations reporting UV to Weather Underground, I notice rather wide discrepancies on cloudless days, suggesting that the sensors/reporting stations are not operating equivalently. More recently I noticed (with the aid of a neighbor who looks at my UV graph daily in order to control his intentional sun exposure) that my system UV sensitivity had declined substantially. We used historic comparison using cloudless (as readily observed from the shape of the solar curve) days at the same time of year across the years in service, and also cross-checking with the EPA forecast for my position. As I use Cumulus to generate both my Wunderground report and my personal web page, I used the facility in Cumulus to activate a multiplier of 1.40 last June. But I consider this not highly accurate, and suspected that whatever had degraded would continue to degrade.
While Davis suggests frequent calibration for stations desiring high accuracy, I did not pursue that option not wishing for my reporting of UV to be out of service for weeks, and wondering what the calibration price might be. At Christmastime I hit on the idea of buying a new sensor, swapping it for the degraded one, then sending the degraded one to Davis for calibration/refurbishment. I contacted Davis support by email inquiring as to the proper procedure and the price. They replied a couple of days ago, advising that I contact them by phone to initiate things, and, more importantly, suggesting that I request "advanced replacement". In my case this means I gave them my credit card number, they are sending me a previous customer's refurbished sensor. I'll swap the sensors, then ship back my degraded sensor in the package used.
They did ask how old my sensor was before setting this up. Possibly that affected the price. In any case the price I have paid is about a quarter of a good price for a new sensor. So my initial out-of-pocket expense is a fifth of what my original plan contemplated.
Regarding exposure and cleaning. I live in Albuquerque, NM, so my sensor sees more UV and more general sunshine than is probably typical. While in many respects our air is unusually clean, it is possible we have some undesirable content in our dust or air which perhaps may accelerate UV sensor degradation (my solar sensor had degraded little if any). Very roughly once per year I have cleaned both the solar and the UV sensors, always using Everclear, as I have believed it strongly equivalent to the prescribed methanol. However, recently I have read (here, I think) that some customer was specifically advised against Everclear by a Davis support rep. While I am not at all sure that was good advice, I did again seek ethanol for sale to a non-institutional buyer, and found an eBay seller selling from a far place. I'm not confident that product will be in fact superior to EverClear for this purpose, but I'll be adhering to the letter of Davis prescribed maintenance.
I don't know whether the "advanced replacement" option is available on other components of Davis weather hardware (or on whole systems), but if, as I was, you are troubled by the idea of a long in-service gap, perhaps you might inquire when you call Davis support.