I'd say it's extremely difficult to obtain the sound pressure from a WAV file.
Consider what happens to the sound signal between the microphone and the WAV file. Pressure is applied to the mike. It is converted to an electrical signal, with power that depends on the properties of the mike transducer. Then it is attenuated on its way to the sound card, where it is amplified, filtered, sampled, quantized, encoded, and further massaged by the sound card driver and the OS sound system.
(Furthermore, if the signal is ultrasonic, the frequency response of your sound-card front-end may very well come into play too).
Then, in software, the samples are further processed to fit a particular WAV format (the details of which are amply documented in Wikipedia and other sources).
One possible way to proceed is to obtain experimental readings with your particular set-up: generate a set of known, controlled sound pressures on your microphone, and correlate them with the numbers you get in the WAV file.