You may want to reconsider your choice of sensor. Although a decent permanent magnet (PM) DC motor will produce an output voltage, which in turn can produce a current in a resistive load, that is more or less a linear function of the motor shaft speed, it may not be the "best" choice based on your limited experience.
Given very good, low-friction, bearings and a strong PM field, a three-cup anemometer vane attached to the motor shaft should produce a measurable output signal, even at very small wind velocities. How small depends on how "easy" it is to turn the motor shaft when wind blows on the anemometer cups. At some low wind velocity insufficient torque will be developed by the anemometer vanes to turn the motor shaft.
Of course there is also the possibility that what little signal does exist, prior to the motor shaft ceasing its rotation, is buried so far into the noise floor that it is not measurable. All sensors exhibit this problem... there is always a minimal detectable signal. If that minimum isn't small enough (insufficient sensor sensitivity), then either a different sensor is required, or more signal amplification may be necessary. Unfortunately, for signals below the noise floor, amplification amplifies the noise too, so there is no net gain in signal-to-noise ratio. It is beyond the scope of this comment to explain how to improve this situation by adding intelligence (modulation) to the signal.
There are many devices that respond to changes in temperature by changing an electrical characteristic. Tungsten wire, for example, increases in resistance as it becomes hotter. This attribute can be used to sense wind velocity because a heated wire loses heat at a rate that is a function of the velocity of air moving across the wire. Circuits that implement wind velocity measurements based on the rate of energy loss in a heated wire are called hot wire anemometers. I suggest that you research this method of measuring wind velocity.
Here is a link to get you started.