I, too, had just returned from a training flight, and on final approach I noted that the windsock at the north end of the airport indicated a slight right crosswind; the windsock at the south end indicated a left crosswind; and a flag atop one of the tall buildings to the east indicated a left crosswind. The ATIS reported a slight right crosswind.
In my opinion, the upset was caused by orographic turbulence generated by the buildings to the east. Anyone familiar with mountain flying knows the perils of flying on the leeward side of major obstructions such as mountains, ridge lines, hills, tall tree lines, or large structures.
Too many pilots rely solely on the tower, ATIS, AWOS, or ASOS reported winds to tell them what the wind is doing. And I always wonder why some pilots on final approach will ask the tower for a wind check. These wind reports are historical snapshots, good only for general knowledge at the location of the wind sensor.
On-the-scene indications are far more meaningful. Windsocks, smoke, steam, dust plumes, flags, and the smoke generated by a jet's tires smacking the runway are all good indicators. The most important consideration is whether or not your airplane drifts to the left or right after you roll out on final approach.
Fortunately, the tall buildings on the east side of our airport have wind indicators - the flag I mentioned and a windsock on another building's heliport. During the winter is when we encounter our most treacherous wake turbulence conditions. It's at that time of year that a large, stationary high pressure system is often located over the region. In the morning the ATIS will report calm or light southwesterly winds on the ground, but the wind indicators atop the buildings may show northerly winds. Within one or two hours, those northerly winds will reach the surface with wind gusts that exceed 30 to 40 knots.
Wake vortices from jets on approach, therefore, are moving not only laterally with the crosswind component, they are also moving forward with the tailwind component. Good luck to the unobservant pilot who relies strictly on reported winds and thinks all is well because his landing is planned beyond the touchdown point of a jet landing on the parallel runway.
Another tailwind-aloft clue is available for pilots who fly a standard visual descent from pattern altitude to the runway. They never start their descent until they intercept a normal visual glide path. If the airplane keeps going above the desired visual glide path when normal power and flap deployment are used, groundspeed is probably higher than normal because of a tailwind.
I did not learn to appreciate the fickleness of wind until I learned to fly gliders. Watch the water swirl in a creek. That's what wind does.
Three or four times while flying at low altitude, I have been enjoying a perfectly smooth flight when wham, I've received a severe jolt. In one case, my head slammed against the cabin roof and I saw stars momentarily. (Yes, my seatbelt was securely fastened.) This situation, which occurs frequently at high altitude, is called clear air turbulence. (Which is why only a fool would sit in an aircraft seat during flight with an unfastened seatbelt - and that includes airline flights.)
A good pilot always knows his or her location, the wind, and where to land if the engine quits. Make wind awareness a high priority on your next flight and every flight thereafter. It can void the general rules for wake turbulence, and it will help you anticipate orographic turbulence.