Circulation theory predict downward airflow that reduce effective AoA,which cause induce drag. To change effective AoA this downward airflow must be infront of wing, to "hit" wing with lower AoA, but not behind wing(in the wake).
In reality I only see upwash infront and downwash behind,so where is this flow?
As the name suggests, circulation theory describes the lift as being produced by the airflow that "circles" around the 2D airfoil i.e. by an airflow going up in front of the airfoil, accelerating on its upper part, going down behind it and decelerating on the lower part.
If the freestream airflow's speed could be subtracted from the picture of your question, this circulation would be really nicely visible:
(Source: Daniel P. Raymer, Aircraft design: a conceptual approach)
There is also downwash along the entire surface of the bottom of the wing. This (action/reaction) lift is what you have left when the upper wing stalls.
"Bottom lift" comes at a higher drag price. More efficient "top lift" can be had at higher airspeeds and lower angles of attack. Essentially, this is why Vbg is faster than Vmin sink.
Regarding "induced drag", one might posit that, since we define Total drag coefficient as Cdo + Cdi, why not call induced drag that which comes from increasing Angle of Attack relative to the freestream. With airplanes, we generally do this to create lift.
So one can conceive induced drag from a 2D diagram either as a single vector resulting from raising AoA, or a decomposition to drag and lift. Further study of Lift and Drag and their relationships at various AoA is highly recommended at airfoil tools on the web.
