NavList:

 I'm practising sight reductions using GPS Anti Spoof Pro as a source for synthetic real time altitudes when it's cloudy.  When using stars, a calculator and my known position, I get Ho = Hc to the nearest 0.1', as expected for a known position.  When using the moon, I'm finding Ho and Hc differ.  Here are some examples:

UTC date: 02AUG22, GPS position: 42° 49.4'S 147° 15.7'E.

GPS Anti Spoof Pro (ASP) settings: Body: Moon, LL, sea horizon, no delay, T=11-14°C, P=1007 hPa/mb. Index correction -1.0'.  Height of eye 2.9m.

01:02:06 UTC ASP Hs=12° 53.1',  Ho 13°54.5',  Hc 13°54.2'

01:55:07 UTC ASP Hs=21°39.3',  Ho 22°39.9',  Hc 22°39.5'

04:12:17 UTC ASP Hs=39°18.3',  Ho 40°11.6',  Hc 40°11.0'

05:55:58 UTC ASP Hs=43°07.4',  Ho 43°58.5',  Hc 43°57.9'

Hc was taken from Frank Reed's USNO web app clone.  Ho was calclated from Hs using the nautical almanac.  I don't know why the difference grows from 0.3' at low elevations to 0.6' in the last two "sightings".  I think the differences are irrelevant to the designed function of the anti spoof program - I would expect much larger differences if the GPS was being jammed - but it is possible that I'm doing the lunar altitude corrections wrongly. 

Perhaps somebody can spot my error.  I subtract 4.0' from Hs for combined index error and dip to get Ha.  I then enter the lunar altitude correction tables at the back of the almanac.  For the 05:55:58 "sighting", I get Ha = 43° 03.4', so I enter the 40° - 44° column of the 35° to 90° table to find 51.8' as my first correction.  Remaining in the 40° - 44° column, I then look down to the second correction columns below, choosing the left hand column (headed 'L') with the moon's horizontal parallax of 56.0'.  I interpolate between HP= 55.8 and 56.1 to yield a correction of 3.3' (this table is not a critical table, so I interpolate.)  The total correction is 51.8' + 3.3' = 55.1'.  Adding this to Ha = 43° 03.4', I get Ho 43°58.5'.

I originally posted this on the Practical Celestial Navigation Facebook page, but administrator Kurt Wege suggested I should post it here.  (I have edited and reworded my original post for clarity.  I removed Zn from the table because in each case Zn was identical to that on the USNO web clone.  I don't have a version number in the copy of Anti Spoof Pro on my phone.)