...
The manual doesn't define AGC, but contains this info, on pages 3-50,51, indicating that as an AGC gets value near 100% the data quality decreaseslikely indicates a water droplet remaining in the optical path, but that temporary droplets can effect the measurements without an indication in AGC:
Dew
The LI-7500 can tolerate droplets on the windows to a certain extent; the AGC value will increase, but the calibration is unchanged. If the droplets coalesce and get big enough, the AGC will go to 100%, and eventually the readings will become bad. This can be minimized by sensor orientation, and by coating the windows with a wax such as RainX®.Rain/Snow
Flying droplets and flakes in the optical path will affect the performance of the LI-7500, even if the total light blockage is small enough that the AGC does not reach 100%. The reason is that the objects are moving, and if a droplet or flake is in the path for a sample measurment, but out of the path for a reference measurement (or vice versa), it will influence the resulting reading. Figure 3-4 illustrates a strip chart trace made during a light snow shower. The occasional spikes are due to blockage changes between sample and reference readings. Notice that in some cases both CO2 and H2O are affected, and sometimes just one. Also, the spikes are equally likely to be up or down.