I don't know of too many websites that would educate you on this matter. I learned all I know from my meteorology classes at ISU. I know
www.theweatherprediction.com has a lot to look at. Also, the SPC mesoanalysis page can be really good practice for applying what you learn. I'll give you some tips right here:
Surface
Look for boundaries, either in terms of thermal boundaries, wind shift lines, pressure troughs, or outflow boundaries. These can be locations of strong surface convergence, which can result in upward motion. Look for wind direction and speed. Generally, the more backed (i.e., more SEerly or even ESEerly) the surface winds, the better set up for strong low-level directional shear (assuming the 850 and 700 mb winds veer appropriately).
850 mb
Not much to look at at 850 mb except for wind direction and speed. Determine whether the 850 winds have veered or accelerated much from the surface winds to get a good idea for low-level shear. You could also look for low-level jets. Although LLJs may cause storms to be elevated, they pump in tons of warmth and moisture and can really cause shear to jump up. Some temp data can be useful at 850 mb, but better info is usually given at 700 mb.
700 mb
The biggest thing to look for at 700 mb is a cap. Look at the temperatures. In the spring and summer months, 10 or 11 is getting a little high, whereas 13 to 14 is nearly unbreakable. If you see anything over 14, expect clear skies unless the mother of all convergence boundaries is nearby. I attached a humorous image from a day in which you would
notwant to be chasing in SE SD.
Also continue to see if the winds are changing...shear through this layer is very important.
500 mb
Like the other layers, pay attention to wind speed and direction. They should both continue to change through this level. You can look at temperatures to determine mid-level instability (this is where the LI is determined). Temps below -5 are usually okay, but if you get above that, watch out...you may suffer some lack of instability. Look for troughs and ridges with associated vorticity. PVA at 500 mb is a good indication of where you'll find upward motion. The stronger the better (shortwaves tend to move faster and can have stronger PVA).
300/250/200 mb
I would look at 250 or even 200 mb in the late spring and summer for upper air info because the tropopause will be higher at that time of year. Thus, the jet stream and the fastest winds will probably be at this level instead of at 300 mb. Hopefully the wind speeds have accelerated from the surface up to this point. Supercells need plenty of deep layer wind shear (0-6 or 0-8 km shear). Also look for jet streaks. Upward motion is favored in the right entrance and left exit region of jet streaks (so long as they're linear...if a jet streak is curved, the inside-of-the-curve region has its motions enhanced).
Essentially, you want to look for anything that can cause upward motion in a region of instability and shear. Hopefully those areas are collocated. If you have access to model sounding data, like BUFKIT, I would suggest also looking at model soundings from specific areas to get a full profile of the atmospheric column over a given region.