I’ve been doing a fair amount of web searching on this topic. Much of the literature is too technical for me. The best I found was this on-line ET calculator:
http://www.tfrec.wsu.edu/orchard/pet/petcalc.html
The Javascript or Java source is available and the author has made a number of helpful simplifications to the Penman equation. 3 of the 4 inputs (average temperature, relative humidity and wind speed) are fairly easy to find on the internet (for example, wunderground.com). But solar radiation is tricky. I’ve found some data for my area, but not for the entire US. The calculator offers an approximation using time of year and sky conditions (clear, partly cloudy, etc.), but I couldn’t find a data source for sky conditions.
Could UV index be used? I searched the web for some correlation but couldn’t find any. It may be possible to compare historical solar radiation and UV index data and look for correlation.
Could we just drop solar radiation? Maybe, but as I vary the solar radiation values it makes a pretty big change to the computed ET. Would ignoring solar radiation produce an unacceptable amount of error (and over/under watering)? I suspect it would but I’m really not sure.