Astronomical Easter Comparison & Calculation by OP Armstrong

10.25/15

Table-A Calculation Formula for Astronomical Easter Calculation, Gregorian Year JD# Method
A
B
C
D
E
F
YR pEpact JD#1Jan JD#JanNM JD#PFM1 JD#Eqnx
-4000 24.3 260090.45 260113.8 260187.8 260170.2
2019 6.0 2458484.8 2458489.8 2458563.8 2458563.4
Yr

G
JD#PFM
260187.8
2458563.8

H
WkDy
7
5

pEPCT JD#Jan1.3 (B-1)+C 74.02+D F.3 or F.7 IF(E>=F,E,E+29.5) f.6(G)

JD# Easter
260188.8
2458566.8
IF(H=1,G+7,8INT(H)+G)

J
D#ck
1
1
f.6(i)

The above Table-A is a tabulation of calculation steps for finding Astronomical date of Easter. The
example calculations use formula-2 of Table-B to determine moment of January New Moon for
selected year, expressed as astronomical Julian Day number, JD#. Alternatively one could use
any number of resources to arrive at a value for column B. Just be sure to express the moment as
an Astronomical formatted Julian Day Number, JD#. Julian calendar or Gregorian calendar dates
may be verified by the day of week. The numbers in column F, JD#.Equinox, are for equinox
moment by formula 4 of Table B. The month number and day-of-month may be determined in a
spreadsheet by adding two more columns and using formula 8 and 9,Table B, given that Year, Yr,
is stated as input in column-A. Alternatively, the Easter JD# may be converted back to Gregorian
Date by several free programs. This routine was compared by 70 dates. If using JMT in place of
GMT (UT), then add an offset of 0.098 days to step I and J.
A check was made against the WCC Easter dates table for years 2025 to 2001. Against that Table
this method using formula 2 and 4 reproduced their result. Other checked instances returned
results that matched either the Catholic Easter dates or alternative astronomical calculation
results. However the data of Ovidiu Vaduvescu did not confirm the astronomical values set forth in
the WCC document nor results of this calculation. If Equinox and PFM dates are closer than one
1
(1) day, verification by a more precise routine is advised on slide-9. NEXT N3

Astronomical Easter Comparison & Calculation by OP Armstrong

9.5/15

Name (Nu)

TABLE-B Excel Astronomical Name Formula: Yr-year; JD#-Julian Day

pEpact.Cassidy.f1

29.09-MOD(MOD(Yr,19)*11-INT((Yr-1502.57-12*MOD(Yr,19))/228),29.983)

pEpact.Cassidy.0.f1b

29.5-MOD(MOD(Yr,19)*11-INT((Yr-1584-12*MOD(Yr,19))/228),30)

pEpact.Lunation#.f2

IF((1+MOD((365.242454*(-4006-Yr)),29.5306))>=30,((1+MOD((365.242454*(-4006Yr)),29.5306))-30),(1+MOD((365.242454*(-4006-Yr)),29.5306)))

JD# Jan1.f3

257898.52-365.242454*(-4006-Yr)

JD# Equinox.f4
1st Page

(2457102.448+(Yr-2015)*365.2422)+((-0.0005947871)*((Yr-2015)/1000)^4+(0.00392591)*((Yr-2015)/1000)^3+(0.013808751)*((Yr2015)/1000)^2+(0.1590901)*((Yr-2015)/1000))

March 1st Moon.f5

JD#.Jan1 + pEpact + 59

Day of Week.f6

(1+INT(MOD((1.5+JD#),7)))

one is Sunday and 7 is Saturday, etc

JD#21March.f7

257978.00-365.242454*(-4006-Yr)

Day of Month.f8

1+INT(MOD(((INT(MOD(((INT(JD#+0.5)+(37+INT(0.5+0.75*INT((INT(JD#+0.5)-4479.5)/36524.25))))59.25),365.25)))+0.5),30.6))

Month Number.f9 ,
3=March, 4=April

1+MOD((2+INT(((INT(MOD(((INT(JD#+0.5)+(37+INT(0.5+0.75*INT((INT(JD#+0.5)-4479.5)/36524.25))))59.25),365.25)))+0.5)/30.6)),12)

Next page

Astronomical Easter Comparison & Calculation by OP Armstrong

Why the Sunday after the Paschal Full Moon?
Take a look at either the proleptic Catholic or
Astronomical Easters for years 30 and 33AD. For these
dates, the proleptic Gregorian Easter dates are April 7
and April 3, respectively. Next determine the Lunar
Conjunction (astronomical new moon) age of 17.9 and
17.2 days, respectively, at 4pm Sunday. Given that Jesus
died before Friday sunset and was buried before sunset,
so 1 day back is, Saturday 4pm and another day back is
Friday 4pm. Thus at time of Jesus death, the Lunar days
are 15.9 and 15.2 respectively. Given that the visible new
crescent or sighted moon is some where about 1 day
after conjunction. So on Good Friday the lunar sighted
moon of these years is 14 days in age. Thus Good Friday
of years 30AD and 33AD would either be on Nissan 14 or
15. However, if this skip week, after were not used, then
sometimes the Easter would come early to the lunar
calendar.
Why the Postponement Rule?
From this it is seen that reckoning Easter by these
methods speaks of "Christ our Passover", Jn18:39, 19:14,
1Cor5:7, Heb11:28, Num9:13, Ex12:14.
Since the Catholic Easter concurs the Astronomical
Easter in about than 90% of the dates, keeping Good
Friday and Easter are keeping "Christ our Passover" as
"a memorial; and .... a feast to the LORD throughout your
generations; ... feast by an ordinance for ever." Ex12:14.
The postponement rule importance is illustrated by
calculation. NEXT

9.5/15

Look at 135 different years between 4000BC and

2038AD. The average lunar conjunction age at
4pm Easter Sunday was about 19. days, with a
minimum of 15.5 days and a max of 22.2 days.
Without the postponement rule, then some Easter
Sundays would land before a lunar age of 13.5 day
This gives a very early Lunar Good Friday of just
12.5 days. Application of the rule "Sunday
following the Paschal Moon" keeps Good Friday in
better alignment with the 14th lunar day. For at the
minimum found above, Good Friday falls more
closely on the 14th or 15th lunar day.
Recapping the "raised up on 3rd day", Good Friday
was day one by inclusive counting as Jesus was
laid in grave before sunset on the Sabbath. Then
day 2 was Friday after sunset to sunset Saturday,
and day 3 inclusive count was Saturday after
sunset unto about sunrise Sunday. Thus the
reason for sunrise Easter service is to memorialize
"Christ our Passover", "a feast by an ordinance for
ever." To the end that others may wish to abolish
this ".- feast by an ordinance for ever", they neglect
the blessings of the Almighty and sadly invite the
fires of judgment to blot their memory from off the
earth!
UTC to Jerusalem Meridian Time
Adjustment to Jerusalem time can be
accomplished by adding 2.33 hours or 0.098 days
to the if statements and day of week calculations.

Astronomical Easter Comparison & Calculation by OP Armstrong 10.25/15

Frequency of Variation
For 24 years, from 2001 to 2025, only once did the
Astronomical Easter differ from the Catholic Easter. That
being 2019, with Astro Easter on 24March, and Catholic
Easter on 21April. Many of the tabulated dates of the
table were evaluated for known difficult dates likely to
offer discrepancy between the two methods. Even so, only
11 of 60 years show variance between the methods. So
most likely, 90% of years will show the Catholic Easter and
Lunar Easter are same.
Other Feasts: Pentecost, Tabernacles, and The Lord's Supper

The other two of the three main feasts were Pentecost

and Tabernacles. For Christians, Pentecost (7 weeks after
Easter) celebrates the Holy Spirit. The feast of Tabernacles
celebrates being freed from bondages of the flesh, the
world, and eternal death, thru election by God the Father.
By celebrating these 3 feasts, Passover (hope of a more
better resurrection thru Jesus Christ), Pentecost (hope of
continual renewal by Holy Spirit), and Tabernacles (hope
of election and freedom from bondage thru grace of
Father God), the Holy Trinity can be honored by Christians.
Because Jesus initiated the Holy Communion, then
communion on these days should be held in high esteem.
The 1st of the first and seventh lunar months are seen as
special days or Sabbath's. Also, the middle of these
months are feasts of remembrance.
Pentecost is by the definition, always on a Sunday by
virtue of the week count, 7th Sunday after Easter.

"And ye shall count unto you from the morrow after the
Sabbath, from the day that ye brought the sheaf of the wave
offering; seven Sabbaths shall be complete: Even unto the
morrow after the seventh Sabbath shall ye number fifty
days; and ye shall offer a new meat offering unto the LORD."
Lev 23: 15/16. The Fifty days were numbered from a
Saturday, so the 50th day falls always on a Sunday, 49days or
7 weeks after Easter Sunday
For Christians, the Easter Sunday sets the precedent to find
the Sunday in the 7th month. The earliest Easter being 22
March. By this the 7th lunar month starts around the last
week of September. The latest Easter falls on 25 April, by
which the 7th lunar month looks to fall on last week of
October. The 1st Sunday after and a 3rd Sunday would be a
logical memorial for the Autumn Feasts.
Effect of using Jerusalem as Principle Meridian
Exercise Caution when the Universal Time new moon
moment is on a Saturday night after 21:30 hours. As in 1998,
when the Full Moon moment was Saturday at 22:23. The
+2h21m offset between GMT and Jerusalem, gives a Paschal
full moon on 00:44 Sunday 12 Apr. This astronomical Easter
is delayed unto the following Sunday. Compared to a
Catholic and/or Universal Time Easter of 12April. This shows
the principle difference in the two methods. The Catholic
calculation is stepped in days and weeks. The Astronomical
method is a moment defined calculation. Thus it depends
upon the details of complex astronomical calculations.
NEXT
4

Astronomical Easter Comparison & Calculation by OP Armstrong

st

th

10.25/15

The 3 Mosaic Feasts of 1 month, 7 Month Ex23:14, and Pentecost

Comparison of Three Major Hebrew Feasts
Pentecost,
The Table at right compares the three feasts set forth by Reference Nissan, Lev23, Nu28:16
Trumpets -Tishari
Du16:16-17
Moses. Two of these feast weeks are routinely celebrated
in most churches. The first is Easter Holy week. That starts MonthDay/
March/April
May/June Ex23:14 September/Oct.
date+\with Palm Sunday, followed by Good Friday, and lastly,
Sabbath & blow
Resurrection Sunday or Easter Sunday. The first of these 1st Day-A Sabbath, 1st Day of 1st 15th Nissan + 49D, Trumpets 1st of 7th,
New Moon & sin
two Holy Days of Palm Sunday and Good Friday, correspond Sabbath
Month
Easter + 49 days
offering
approximately to the 10th and 15th days of a lunar month.
A Sabbath Lev23:8,
This provisioned that the 1st day of the lunar month falls 7th Day
A Sabbath
Ex12:6
upon a Friday, the 8th day being a Friday also, the 10th day is
pick passover Lamb, 9th evening to 10th day of atonement,
then a Sunday, and the 15th day being a Friday. It is
10th
Psalm Sunday
evening
fasting
provisioned that Jesus died on a Friday and His empty
Passover: sunset
tomb found early Sunday; thus the term Holy Week.
th
14
Deut16:6, kill lamb,
The feast of Pentecost is celebrated in most churches 7
Ex12:6
weeks after Easter. These first two Mosaic Feasts being
unleven bread 7 days,
Tabernacles for 7
th
th
celebrated, begs the question? If two are celebrated, then
15
15 is Good Friday &
days, start with a
17th Easter
Sabbath
why not the 3rd and if so, then how so? If one remembers
st
21
evening end unleven
the words of the blessed Savior, When you Fast they
nd
22
offering for 7 day
were not in the permissive sense but in the imperative
rd
instance. The 40 days of Lent could be taken as something
23
Assembly
dealing with fasting. However the Master was noted to
Taking out the two and a fraction of weeks elapsed unto
have kept the Feast of Tabernacles. So then does it not
Easter, then 22 weeks after Easter corresponds to nearly the
seem proper for churches to also remember this 3rd Feast?
start of the seventh lunar month. A comparison of dates 22
The Sunday following Passover is typically Catholic Easter
weeks after Easter to 1 Tishrei shows the two dates agree
Sunday. As Easter is the Sunday following the first full moon
within a few days. The odd exceptions being when Hebrew
after Spring Equinox, then the start of the lunar month is
calendar postponement rules are applied. One possible
about 2 weeks prior.
church memorial to this time would be Eucharist 22 and 24
A lunar month is about 28.5 days in length. Six lunar
weeks after Easter.
BACK
NEXT
months are then 171 days or 24.4 weeks.
5

Astronomical Easter Comparison & Calculation by OP Armstrong

Only one application offers astronomical or
uniform Easter calculation. Kalendis has not
been ported to mobile devices. This Excel
spreadsheet will find the Astronomical Easter
for years between 4007BC and 3027AD. In Table
to right, are differences between this calc and
others Astro-Easter. For these, exact values of
full Moon and Equinox are input. The evaluated
Uniform Easter dates by this excel sheet were
the same day as those from Kalendis. The below
chart shows variations of my Full Moon Date to
Kalendis. The normal range is +/- 1 hour for
about 7000 years. Beware, not all
implementations of mobile Excel have the
needed numerical accuracy to evaluate these
complex expressions. Only IOS-Numbers and MSExcel routines could accurately execute these
complex calculations.

97 YR's Astro v Catholic Easter Variants

2038AD-1AD

YR

2038
2019
1998
1967
1962
1954
1943
1927
1876
1845
1829
1825
1818
1805
1802
1744
1724
1700
550
-61

Astro
28Mr
24Mr
19Ap
02Ap
25Mr
25Ap
28Mr
24Ap
16Ap
23Mr
19Ap
03Ap
22Mr
14Ap
18Ap
29Mr
09Ap
04Ap
26Ap
me26Mr

First. Page

Cath
25Ap
21Ap
12Ap
26Mr
22Ap
18Ap
25Ap
17Ap
09Ap
30Mr
26Ap
10Ap
29Mr
21Ap
25Ap
05Ap
16Ap
11Ap
19Ap
NoCalc

10.25/15

Easter
by
UTC
UTC
this astro Ful l Moon Equi nox
28-Mar 2465503.59 2465503.03
24-Mar 2458563.57 2458563.42
19-Apr 2450915.43 2450893.33
2-Apr 2439575.64 2439570.82
25-Mar
25-Apr
28-Mar
24-Apr
9-Apr
30-Mar
26-Apr
10-Apr
29-Mar
21-Apr
25-Apr
29-Mar
9-Apr
4-Apr
26-Apr
26-Mar

2437744.83

2437744.60

2430805.42

2430805.00

2406353.32

2406333.76

2395014.35

2395011.24

2389167.08

2389167.36

2387719.77

2387706.39

2385151.09

2385149.70

2380425.49

2380401.54

2379333.61

2379305.83

1922022.20

1922022.43

1698859.401 1698859.395

Astronomical Easter Comparison & Calculation by OP Armstrong

10.25/15

Excel Formula for Catholic Easter by Julian Day Number for Calendar Converter or direct
=MOD(-8-11*MOD(Yr,19)+INT((Yr-1600)/100)-INT((Yr-1600)/400)-INT((8*INT((Yr-1400)/100))/25),30)

Excel Cell
30
14

=P'-IF(P'=29,1,IF((1+MOD(Yr,19))>11,IF(P'=28,1,0),0))

14

.f12 D.1

=118+INT(365.25*(Yr+4712))-INT(0.75*INT(((Yr)/100)+49))

1732097

.f13 D.2

=D.1+P

1732111

.f6 D.3

=1+INT(MOD((1.5+D.2),7))

.f14 D.4

=IF(D.3=1,D.2+7,8-D.3+D.2)&"Catholic Easter as Julian Day Number"

1732114

end Easter

use D.4 to get Day of Month, f.9, &Month#, f.8, &day-of-Week,f.6 as =D&f.n&

D7M4wkd1

C var
.f10 P'
.f11

!. Single Input of Gregorian Year, if BC then 1-Yr.BC & find Easter in 7 steps, i.e. 30AD=>

Above is Catholic Easter formulation using Gregorian Year as

input to find Easter Date as a Julian Day Number. This
simplifies several steps as compared to other methods. The
resultant JD# may be used to rapidly find Pentecost Sunday
or any number of Easter dependant days. The JD# can be
rapidly changed to calendar dates via formula f.6, f.8, & f.9
for most, if not all, dates. Beware excel dates do not display
for years prior to 1900. Thus the above work around. The
Easter JD# can be ported to excel date system for years 1900
forward. Simply determine the offset to Excel date number
and JD#, then subtract offset to other dates. Apply this
method when a wide range of dates are to be reviewed.

Next page

Most Easter routines, but not this one, are valid for a few
hundred years after 1901. This is substantial since Christ
our Passover was slain from the foundation of the world.
Thus there has been a perpetual Easter since the day this
world was founded. That date was when Adam sinned,
4000 years prior to Jesus Baptism at river Jordan by Saint
John. The day of resurrection was hidden by God and
revealed by Christ earthly Easter day in 30AD.
The following list dates when Astronomical Easter is not
same date as the Catholic Easter, about 1 in 12 years.

257973-365.242454*(-4006-Y)+MOD((365.242454*(-4006-Y)),29.5306)+CF
CF= -0.40614*SIN(l')+ 0.01614*SIN(2l) + 0.17302*SIN(l) - 0.17+ CFt
CFt = - C^3/999999.45 - C^2/4028.335 - C/64.259 + 1/ 547.41

Another estmate of astronomical PFM is

given at left. This uses Delaunay arguments
for lunar and sun anomalies, expressed in
Julian century, C of J2000.
7

Astronomical Easter Comparison & Calculation by OP Armstrong

Variant Astro-Easter Years to Catholic
2049 1700 1457 1206 915 543 254
2045 1693 1453 1199 895 536 242
2038 1685 1446 1182 881 519 235
2019 1666 1429 1162 861 516 218
1981 1629 1427 1155 854 509 216
1974 1622 1419 1142 846 496 191
1967 1609 1409 1135 837 489 188
1962 1598 1408 1128 827 482 171
1954 1590 1402 1111 817 474 168
1943 1582 1389 1108 810 469 144
1927 1578 1375 1101 783 438 137
1924 1571 1370 1088 736 411 124
1923
1362 1081 729
This tabulation covers from years 2050 back to 60BC. A
span of about 2100 years. The above Tabulation shows
that some centuries have less discrepancy than others.
The cycle of 391-19 or 372 years can be seen in some
skip sequences. The average is about 11 of 12 years
agree and 1 in 12 years are not in agreement between
the two Easter Methods. However the pattern is not
uniform. Some centuries have fewer concurrences than
others.
Next page

10.25/15

Variant Uniform Easter Years to Catholic

1903 1558 1351 1074 716 384 73
1900 1552 1332 1061 685 367 66
1876 1551 1331 1030 634 343 42
1873 1527 1328 1003 614 323 22
1845 1514 1313 998 607 320 15
1829 1507 1311 979 590 313 -5
1825 1503 1308 960 588 303 -29
1818 1487 1294 959 587 292 -49
1805 1484 1284 941 570 289 -53
1802 1483 1277 935 569 273 -56
1778 1473 1237 932 563 269 1998
1744 1465 1226 922 550 262
1724 1463 1218
Given the primitive nature of physics and astronomy at
the time of the original formulation of the Easter
Calendar, the Catholic method mostly agrees with the
Astronomical formulation. That being stated, the
Paschal Moon of the Catholic Easter is not a true
moon. In fact there is similarity to Molad of the
calculated Hebrew Calendar. The Molad event is timed
to start around the autumn equinox vs. around or
about the Spring Equinox for Catholic Easter.
8

Astronomical Easter Comparison & Calculation by OP Armstrong

Uniform Easter Spreadsheet Method: Find full moon

using longitude routine and: Moon age = (Solar less

Moon Longitudes)*29.5306/360
Since Full moon is 180 degree, then full moon age is
half of 29.530 Three approximations (February,
March, April) are used to find Paschal full moon date.
When solved yield average days between spring full
moons for the year. The first guess when corrected,
may not meet Paschal Criteria after correction. Thus
2 checks are used. Given the Equinox for a year, as
EQ.jd, then the logics are: 1) IF(guess < EQ.jd, guess,
then new is (old guess+29.5), Guess-1 is

257972.7365.242454*(-4006-Yr)+MOD((365.242454*(-4006-Yr)),29.5306)

from which guess, an age is determied. This AGE is

converted to full moon correction, CF, for an improved guess
as follows: Cfi =29.5306/2 - IF(AGE>0, AGE, 29.5306+AGE)

Guess2 = Guess1 + CF1, other iterations are made

as Guess3 = Guess2 +CF2, the 4th and final value,
Guess4 = Guess3 + CF3 is taken as the Full Moon
date. A final check is used to ensure the Full Moon
date falls after the Spring Equinox. Macro loops are
avoided by using successive Full Moon dates:
estimate, estimate +/- 29.5 days and logically select
PFM. The day, 257973, is March FM of Gregorian
year -4006. The file is embedded on page 6 of the
MS document file. This file may be used to better
see the method. A simplified method screened the
prior table for validation by 70 term method. FM JD#
Constants: April 258002.07, March257972.70, February 257943.37

10.25/15

The Paschal Full Moon is defined as first moon, after

the Spring or Vernal Equinox. This spreadsheet
method of longitude was corrected and adapted after
T. Alonso Albis adaptation of P.Duffet but with
additional terms. T. Albis original lambda terms were
only about 24. Accuracy was improved by using more
Lambda terms. The total Sine terms comes to about
70. Evaluation of Delaunays arguments from Chapront
& Chapront ELP-S2001 was used to improve
accuracy. The ELP-S2001 4th order terms are reported
to be within 10 arc seconds down to 1500BC, if using
the full series of 100s of adjustments. Here only about
70 terms are employed. The original method does not
account for Tidal Braking. Thus, an Equation of Time is
used to extend the solution validity. The extended
solution calculates Paschal Full moon dates to within
an hour of NASA or Kalendis for years between
4000BC to 3000AD. Typically less than 60 minutes
difference is found. The graph on page 6 shows the
variation for 110 points over a 7100 year span.
It is incorrect to apply angle reduction, Mod360, for
values less than zero. The spreadsheet solution is as
follows: reduce angle by A/360=rA, then find
IF(ABS(rA)>1, 360*(ABS(rA)-INT(ABS(rA)))*SIGN(rA), 360*rA).
This yields a correctly reduced angle for either positive
or negative input values. It is possible to improve
accuracy by limiting the year range, to just a few century.
Next page

Astronomical Easter Comparison & Calculation by OP Armstrong

New Moon Moment by Spreadsheet
The prior calculations were for finding a Full Moon moment
by longitudes. Here a new moon moment is also calculated
by longitudes. This improved the accuracy of moon age
cell in spreadsheet Gregorian date to/from Julian Day. The
prior calc used lunation number. The Solar to Lunar
longitudes method improves accuracy when used with
adaption. The adaptation was to tailor an EOT specific for
this data set. In the first instance, Broombergs 4th order
lunation number correction was adjusted to minimum
error around J2000. The results of that fit are given in
graph of this page. This gave +/- one hour error around the
mean between 4000BCE and 3000AD. Next a 3rd order
supplemental correction was fitted in terms of Julian
Millennia or c/10. Where c is (JD#- 2451550.098)/36525.

Nov7/15

The new moon moment was found by setting the longitude

For 128 points, the Standard difference to zero. The EOT is required because Longitude
deviation was about 3 minutes at methods do not account for tidal braking of earth rotation.
a mean of about zero. The RMS This method is known as a truncated series of the VSOP
mean was 2 minutes. The analytical series, as compared to the gold standard direct
maximum/minimum differences integration of DE200, as discussed in USNO write ups. Take
were 10 and -7 minutes. This as example March 30 AD: guess M15 and JD# is 1732091,
nearly corresponds to the 12 min and EOT date is 1732095.950, and is used for second guess,
gap of 2 times SD. This method the third and final calc gives a JD# of 1732096.244 for the
explains 98% of determined EOT corrected date when moon age is zero, this is Mar20
17:51HR vs. NASA table of 17:47 or 4 minute later. Likewise
variance.
Because the original equation is in hours, divide by 24 to for April 457BC , start midmonth at 1554614.00 to arrive at
give days. This correction is then added to the Broomberg EOT JD# 1554617.311 or 19:28HR vs.19:25 NASA.
NEXT
term. The sum is then subtracted to yield a new moon BACK
10
time, +/- 6minute with 97.5% confidence.