In theory, if a beam is completely parallel and fits within the aperture of a sensor, then it should make no difference at all what the distance is; it will be the same number of photons (ignoring absorption by the air, which is negligible except in the UV below 250nm). If, nevertheless, you do see such a distance dependence, there could be one of the following effects happening:
- If you are using a thermal type power sensor, you might actually be measuring heat from the laser itself; when very close to the laser, the thermal sensor might be “feeling” the laser’s own heat. That would not, however, continue to have an effect at more than a few cm distance unless the light source is weak and the heat source is strong.
- Beam geometry – The beam may not be parallel and may be diverging. Often, the lower intensity wings of the beam have greater divergence rate than the main portion of the beam. These may be missing the sensor's aperture as the distance increases. To check that you'd need to use a profiler, or perhaps a BeamTrack PPS (Power/Position/Size) sensor.
- If you are measuring pulse energies with a diffuser-based pyroelectric sensor: Some users find that when they start with the sensor right up close to the laser and move it away, the readings drop sharply (typically by some 6%) over the first few cm. This is likely caused by multiple reflections between the diffuser and the laser device, which at the closest distance might be causing an incorrectly high reading. You should back off from the source by at least some 5cm, more if the beam is not too divergent.
Needless to say, it’s also important to be sure to have a steady setup; a sensor held by hand could easily be moved around involuntarily, which could cause partial or complete missing of the sensor’s aperture at increasing distance, particularly for an invisible beam.