hey, it really depends on what kind of play you are interested in- if you are delta hedging (delta neutral) the implied vol and the other greeks would be managed in a different way than if you just for example putting on a directional play. ( where you just want to pay a premium for a spread say)

Always remember that implied vol is the premium that people are willing to buy/sell options at and these will differ significantly for at the money options to say puts in equity indices; puts trade at much higher vol due to supply and demand- we are in a bear market and vol explodes when say the dow/ftse/dax touches year/multi year lows. That is all a vol surface is, different implied vols for different strikes (calls, atm, puts)- quadratic and tertiery equations can be used to describe this surface and the parameters I mentioned are such parameters.

This implied vol is different to realised vol which is the average (standard deviation) of the underlying movement on a tick/hourly or minute basis. In a broad sense, they are related in that if the underlying starts moving violantly, up or down or both, then the realised /historical vol is higher and this will start pushing implied vol up through market makers wanting to own options to profit from the gamma of more violant underlying movement.

To answer your question directly about modelling future vol I would say there are 3 approaches you can take, but I would always suggest using the implied market vol as your starting point. The 2 alternatives are:

1) that if you are looking to profit from gamma movement, look at the historical (underlying movement) vol for as you say, a period of observation matching the days to expiry. To extend this, you can look at the expected p&l's from gamma and comapre this to the implied vol and expected theta of the options

2)If you want to look at vol movement as the underlyier moves, map more expected p&l's from scenario analysis (e.g. if you expect vol to double on a move down within 15 days of a 30 day option, plug these parameters into your analysis and look at your Greeks and p&l, then look at worst case scenario where say the underlying goes to your long strike and vol is smashed through the floor).

In short though, there is no accuarte way of predicting future vol. Ironically, the best estimate is implied vol (implied vol is a forward looking parameter that the market assesses)