Professional Documents
Culture Documents
P
'
Figure 1. Retrogade/progade orbit pairs with close gound tracks. Shown are horizontal
velocity vectors of satellites at the Equator with respect to the Earth's surface (solid) and
inertial space (dashed). Te retrogade satelltes are labeled R, the progade P. The
velocities of the satellites with respect to inertial space are Vp and vR; their radii from the
center
.
of the Earth are rp and rR
and the efective Earth rotation rates for their orbits are 0 _
and OR .
2
Te corresponding equatorially matching progade orbit inclination I
p
i then gven as:
I
p
= 1800 - I
R
- 0
R
- 0p . (2)
In eq. (1) f
P
is the radius to the satellite (retrogade and progade), and v
R
and vp are their
circumferential velocities at the Equator. Te efective rotation rates of the Earth for these
orbits (accounting for the rotation of their orbit planes) are
(3)
where 0 is the precession rate of the orbit's node and e i the actual rotation rate of the
8
Earth.
Tese rates and satellite velocities (for a near cicular orbit) are given by (Wagner 1987, p.
8138):
(4)
where C
20 is the principal geopotential term in the Earth's oblateness, l_ is the Earth's mean
equatorial radius, and a is the mean semimajor axis of the (near circular) orbit,
(5)
where H is the mean motion of the satellite and J is the Gaussian gavitational constant for the
Earth,
V a(n ^ w)
where w is the mean motion of the satellite
'
s perigee,
C
20
=
-1082.63x 10-6
= 398600.5 I 3 /
l_ = 6378.138 I
e 0.7292115 X 10-4 ,a/s
3
(6)
(7)
(8)
(9)
(10)
(11)
To determine a for an Exact Repeat Mission of R (integer) revolutions in D (integer) snodic
(or nodal) days, where one nodal day i
the algebraic equation
n * 0
21
. .
6 - 0 e
R
. .
= 21
D
T
= [R(6e - O)/D] ,
N
(R/D) = [+
]
[6 e - 0]
(12)
or
(13)
must be solved for a, given the inclination I of the orbit and the mean motions, eqs. (4), (5), and
(7), which are accurate to first order in the oblateness, or 10- (see e.g., Cutting et al. 1978 or
Stewart 1985).
In practice, the initial spherical Earth approximation:
yielding
. .
(R/D) = (n/6e) = [/a3r / 6e
may be used to start an iterative solution of eg. (13), namely
(14)
(15)
However, igoring terms in Cio and higher powers permits an estimate of the mean orbit a to
order 100 m:
(16)
In any case, formal convergence of a to 1 m fom eqs. (14) and (15) is obtained by the third
iteration over all inclinations and ERM specifcations RID.
In particular, for the Geosat ERM with I " 1080 , R " 244 and D " 17, the mean semimajor
axis a i found from eq. (15) to be 7162.578 k (784.440 k altitude). Te ERM ccle time,
D
TN
, from eq. (12) is found to be 17.0505 days.
4
860
840
S
820
. .
800
T
780
.~
... 760
-
~
.~
740
"
+
I
C
720
700
680
100 120 140 160 180
Retrograde Orbit Inclination: Degrees
Figure 2a. Orbit altitude for progade/retrogade ERM pairs. The Exact Repeat Mission is
244 revolutions in 17 days. The upper cure is for the retrogade orbit, the lower cure for
the equatorially matching prograde orbit.
But because of the effects of oblateness on the mean motions [eqs. (4), (5), and (7)], the
mean semimajor axis of the frozen Earth prograde Geosat ERM orbit (at 1800 - 1080 720 )
is [from eqs. (14) and (15)] only 7107.716 k (729.578 k altitude). Thus the progade Geosat
not only has the launch advantage over the retrogade ERM of gaining a measure of the Earth's
rotation velocity, but the required altitude to achieve the 244 revolutions in a 17-day orbit is
also significantly less than for the retrogade mission. (See fig. 2a.)
Continuing with the estimation of the equatorially matching progade ERM orbit (both its
mean semimajor axis and its inclination), a geat simplification of the nonlinear system
determining these parameters [namely eqs. (1), (2), and (13)] i effected by assuming initially a
spherical Earth as well as a near circular orbit, implying that:
(17)
In addition, for a close Earth satellite PR 1 and a
p
,R 1. (Note that for a 24-hour orbit, P
" 1 and a can be large). With these assumptions, the solution of eg. (1) is simply
(18)
5
and the progade inclination from eq. (2) is
(19)
With these simplifications and using aR " 7163 k (which would b available i any cse from
the well-tracked Geosat orbit), the initial estimate of the progade Geosat ERM's iclination,
from eq. (19) is 64.40 (See fgs. 2b, c.) A much more preise value, based on eq. (15) or (16)
is 64.580, sigificantly less than the frozen Earth estimate of 720
Unfortunately the bias from the frozen Earth value of iclation, due to the Earth's rotation,
means that the progade track will depart significantly from the retrogade at high latitudes.
Thus for the Geosat-pair, the difference is more than 70 at maximum latitude. But the
equatorially matched tracks are still tangent to one another (geogaphically) at the Equator and
stay close for a considerable distance before the departure becomes excessive.
In the following discussion, I use an arbitrary reference figure of 10 k to define acceptable
closeness to the retrogade track. It is about the maximum deviation of the 17-day Seasat orbit
from the Geosat ERM. Thus a track within these lmits would satisf militar security
considerations since all Seasat and Geosat ERM altietry has been released publicly. Closeness
requirements from a scientific standpoint are more diffcult to define sice they depend on cross
track sea surface gradient errors that are highly variable. Certainly with such a mission it will be
necessary when comparing sea surface heights over "close" tracks to remove a detailed
"permanent" surface (predominantly the geoid) from the height in both tracks before assessing
oceanographic variability.
Nevertheless, since the prograde orbit always has a smaller maum latitude than the
retrogade, it is always possible to "fne tune" its bias by raising its inclination somewhat and
still maintain acceptable closeness to the retrogade track to geater latitudes than the simple
equatorially matching track.
ORBIT BIAS ESTMATION FOR ACCEPABLE DEPARTURES
Figure 3 shows the evolution of equatorially matching track-pairs to high latitude for the
Geosat ERM, as generated by numerical integation fom a tpical node point. On this and
subsequent fgures, I show a departure zone surrounding the retrogade track. Evidently the
progade track stays "close" to the retrogade reference for more than :20 in latitude. While
it is difficult to determine analytically the departure (or geogaphic separation) of the tracks as a
function of the latitude, it is fairly straightforard to do this by an iterative proedure outlined
in fgure 4.
Let _ be the time forard in the progade track from the node N (or N' to reach latitude c.
In this time, f
r a circular orbit, the argument of latitude will b fp _ (np p), where the
two mean motIOns are computed from eqs. (5) and (7) for iclination Ip and mean semimajor
axis ap [determined for the ERM by eq. (15) or (16)].
Te inclation Ip will b close to the equatorially matchig value found previously. Here we
seek to find an optimum change from this equatorially matchg track which maxiizes the time
spent within a specied acceptable departure fom the reference retrogade track.
6
N
w
20
K
M
w
O
__ ____ ___------
120 140 160 180 100
Retrograde Orbit InClination Degrees
Figure 2b. Inclination of matching ERM orbit pairs. The ERM completes 244 revolutions in 17
days. The solid cure is for the rotating Earth, the dashed (line) for a "frozen" Earth.
67.0
0
66.5
w
O
66.0
|
O
~
65.5
,
~
~
. 65.0
~
H
: 64.5
64.0
N
6
3
.5
6
3
.0
106.0 107.0 108.0 109.0
Retrograde Orbit Inclination: Degrees
Figure 2c. Inclnation of matching ERM orbit pairs. The solid cure refers to the estiate
based on a rotating, oblate Earth. The dashed cure refers to the estiate based on a
rotatig spherical Earth.
7
7
W
W
W
W 6
W
-
4
.~
4
d
1
SOJ1 I&CX 1DC1UBS
1 K BIU1BS
d
LOT@1TM6' 6@T66S
B
.
+
Figure 3. Geosat gound tracks for progade/retrogade orbit pairs. Te solid cure shows the
track near a node for the retrogade ERM orbit of 244 revolutions in 17 days at 1080
inclination as determined by numerical integation. Also shown are 10 k parallel tracks.
Te dashed track arises from a 64.60 inclied orbit of 244 revolutions in 17 days (ERM), also
determined from numerical integation.
Including the small rotation of the orbit plane in this short time, the geogaphic longitude in
the prograde track (with respect to N' at
1
is
' = I A p - 6 pt
l
= cos-
l
[(1 - sin2e /sin2 I p) / cos el ]
- [6psin
-
l(sinel / sin
Ip)] /(
np
w p) .
(20)
Similarly, the geographic longitude back along the retrogade track (with respect to N' at _
before reaching N' (at el) is
Finally, the geogaphic distance, d, between these poits of the track (at el ) is
d = coS-l[sin2e cos2ecos(l - l)].
(22)
Evidently, from the curatures of the tracks, i 1 R' P 1 p' the crossover poit has not ben
reached and the minimum distance (or departure) of the progade track at el , l can be found
by decreasinfR until that minimum distance i reached. Alterately, i AR' A
P
' the
crossover has been passed and the departure can b found by increasin fRo In either case, for
8
t
c
/t.
N' long. A--
RETROGRADE TRACK
PROGRADE TRACK
Figure 4a. Geographic tracks of a progade/retrogade orbit pair.
N
Figre 4b. Iertial track of orbit.
9
1' ` jcos c
f= sin ` sin c/sin I
f= argument of latitude
EQUATOR
each trial f; beyond fR
'
the latitude in the retrogade track i determined from
sin $ R = sinfR'
sin IR
Then the new retrogade longitude is found from eq. (21) with tR replacing $, and the new
distance between the track points from
(23)
(24)
Figure 5 shows the results of these departure calculations for fine-tuned Geosat ERM pairs.
(See also figs. 6a, b, c.) For example, the departure can b kept to within 10 k as far as
latitude 27.6 i the biased progade orbit has an inclination of 65.07. The mean altitude of
such an orbit (from fig. 2a) is 722 k. The ERM ccle time for 244 revolutions i also reduced
from that of the retrogade Geosat, from 17.0505 days to only 16.8193 days.
AIS ED TDA PERODS
Since the oblateness of the Earth is of order 10-3, the rotation of the orbit plane of a close
Earth satellite is of order 10-3 times its mean motion, or about 5 a day. Polar orbiting satellites
experience no plane rotation and those that are slightly retrograde have positive rotation rates
near 1 per day, which can make them nearly synchronous with tides that depend on the Sun.
l
O
w
8
O
~
.~
20
18
16
14
12
10
8
6
4
2
o 0 5
1.
10 15 20 25
I
I
t
30
Latitude in Prograde Track: Degrees North or South
Figure 5. Departure of progade track from the Geosat ERM retrogade reference. Shown are
the track separations for non matched progade orbits away from the Equator as determined
by iteration from eqs. (20) through (24). The matching progade orbit (at the Equator) has
an inclination of 64.6. Notice these overbiased tracks cross over the retrogade track
between 20 - 25 latitude. The matched track has a uniormly increasing departure to high
latitude.
10
<.
B
1.U
1<.
W
1<.U
l
W
11.
b
7
D
+
-~
+
11.U
G
>
<7.U
<U.
W
<U.U
|
C
c
<.
1
<.U
>
<+.
.UU
.
.
.
.
1.UU
o
.
.
#
/
"
-
.
.
.
.
.
1.1U
.
.
.
.
.
.
.
.
.
.
.
I.Z
-
.
.
1U.+ 1U.U 1U. I1. 11.< 11.+
L0Dg1IMU6. 6gT66S
Figures 6a, b, c. Geosat gound tracks for retrogade/progade orbit pais. Te ER for the
pairs are 244 revolutions in 17 days, but an oblate Earth i assumed and the tracks are
determied by numerical integration. The black solid cures are tracks for the retrogade
orbit (I 1080) and 10 k departures.
11
For example, table 1 shows the aliased periods (in ERM ccles) for the 11 principal lunisolar
tides on the retrogade Geosat A orbit. These are the periods as seen by the altimeter
returning ever 17.0505 days (for the retrogade mission) and 16.8193 days (for the prograde
mission) to the same geogaphic poit over the Earth. Tus, i the retrogade mission a single
ERM ccle of 17.0505 days corresponds to 32.9462 M2 ccles. T ERM samples an M2 wave
movig at -0.0538 M2 ccles every ERM ccle. Te sampling is complete only afer 1/0.0538
18.6 ERM ccles. (See table 1.)
Table 1.--Aliased tide periods for Geosat ERM pairs [repeat ccle: 244 revolutions/17 days.
ccle times: I 1080, 17.0505d; I 65.10, 16.8193d]
Tide name
M2
Kl
S2
0
1
P
I
N2
M
f
K2
M
m
SSa
01
Equilibrium
amplitude
(cm)
24.2
14.2
11.3
10.1
4.7
4.6
4.2
3.1
2.2
1.9
1.9
Period
(hr)
12.42060
23.93447
12.00000
25.81934
24.06589
12.65835
327.8406
11.96724
661.301
4383.00
26.86816
ERM aliased period
ERM ccles
I 1080 I 65.10
18.60
10.29
9.90
6.62
261.96
3.05
4.03
5.15
1.6
10.7
4.3
2.00
7.42
2.77
2.73
4.41
9.02
4.33
3.71
2.57
10.9
42.04
On the other hand, for the prograde mission (1 ERM cce 16.8193 days) the aliased M2
period as seen by the altimeter is only 2 ERM ccles. But note that for the principal solar
(diurnal) tide P
I
the aliased period in the retrograde mission is 262 ERM ccles. It is essentially
stationary with respect to repetitive passes (of the same kind, asending or desending) over the
same geographic point. Determining PI from overlap diferences of altimetr yields no
information. If differences of crossover altimetry are used, however, ony the same phase
difference is repeatedly sampled and the to harmonic constants of P
I
cannot be distinguished.
On the other hand, in the progade mission a single P
I
wave is completely sampled by one
kind of pass in only 4.41 ERM ccles (about 74 days). Notice that both pricipal solar tides (P
I
and S2) have significantly smaller alia sed periods in the progade mission which should improve
their determination, compared to the retrogade Geosat. In addition, the variety of periods
provided by data from both missions should geatly benefit the separation of al tidal
components from altimeter measurements.
12
SUMMARY
Te feasibility of a progade mission with gound track closely matching the current
retrogade Geosat ERM has been demonstrated. The progade mission permits geater payload
weigt both because the launch is in the direction of the Earth's rotation and the Earth's
oblateness permits the required mean motions to be achieved at sigicantly lower altitude. A
additional beneft of the progade Geosat is its ability to determie al solar tide components.
However, the rotation of the Earth also reduces the zone where the progade track cn b kept
aceptably close to the retrogade. For the Geosat ERM this zone i between :28 latitude for
10 k track departure. In addition, the progade mission does not sample data to as hig a
latitude as the current retrogade Geosat.
REFERENCES
Bor, G. H., Mitchell, J. M., and Heyler, G., 1987: The Geosat-ERM mission desig. J.
Astronaut. Sci., 35(2), 119-134.
Cutting, E., Born, G. H., and Frautnick, J. C., 1978: Orbit analysis for Seasat A. 1 Astonaut.
Sci, 26(4), 315-342.
Stewart, R. H., 1985: Method of Satelte Oceanogaphy. Univ. of California Press, Los Angeles,
CA, 301.
Wager, C. A., 1987: Geopotential orbit variations: Applications to error analysis. J. Geophys.
Res., 92(B8), 8138.
13