i i

“main” — 2018/3/27 — 10:29 — page 193 — #1

i i

Revista Brasileira de Geofı́sica (2016) 34(2): 193-206

© 2016 Sociedade Brasileira de Geofı́sica
ISSN 0102-261X
www.scielo.br/rbg

CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

Rogério Rodrigues Amarante and Jorge Luiz Alves Trabanco

ABSTRACT. In gravimetric surveys, the determination of the gravity at a specific point depends on field measurements and on the removal of the influence of the
tide. With the current development of more sensitive gravimeters, a review of the tide correction calculation process applied in gravimetry is necessary because values
that were once considered negligible may represent noise above the resolution of the new equipment. Much used until the present, the formulas shown in the study by
Longman (1959) were elaborated with constants (distances, masses, ellipsoidal parameters, and orbital values) that can now be calculated more accurately, as well as
conveniently rounded values for the time. The deformation of the crust caused by the terrestrial tide is another important factor that significantly affects the tide correction
calculation results, and the Earth’s heterogeneity causes this deformation to differ between regions. Much effort has been made to map different regions with the objective
of regionalizing the gravimetric factor; however, computer programs and even modern automated equipment still use fixed values. Therefore this paper aims to present
a review that quantifies the impacts on the calculation of the tide correction considering parameters not yet used.

Keywords: earth tide correction, gravimetric factor, gravimetry, gravimetric surveys.

RESUMO. Em levantamentos gravimétricos a determinação da gravidade de um ponto depende de medições em campo, bem como da remoção da influência
da maré. Com o desenvolvimento atual de gravı́metros mais sensı́veis, torna-se necessário uma revisão do processo de cálculo da correção da maré aplicado na
gravimetria, pois valores antes considerados desprezı́veis podem representar ruı́dos acima da resolução dos novos equipamentos. Muito utilizada até o presente, às
fórmulas apresentadas no trabalho de Longman (1959), foram elaboradas com constantes (distâncias, massas, parâmetros elipsoidais, valores orbitais) hoje atualmente
reformuladas, bem como valores convenientemente arredondados para época. Outro ponto importante é a deformação causada na crosta pela maré terrestre que afeta
significativamente o resultado do cálculo da correção da maré, e que a heterogeneidade da Terra faz esta deformação ser diferente entre as regiões. Muito esforço tem
sido realizado no sentido de mapear diferentes regiões objetivando uma regionalização do fator gravimétrico, contudo softwares usados para redução gravimétrica e até
mesmos equipamentos automatizados ainda utilizam valores fixos. Portanto este trabalho se propõe apresentar uma revisão, quantificando o impacto no resultado de
fatores e influências ainda não considerados no cálculo da correção da Maré.

Palavras-chave: correção da maré terrestre, fator gravimétrico, gravimetria, levantamento gravimétrico.

Universidade Estadual de Campinas, Department of Transport, School of Civil Engineering, Architecture and Urban Design, Av. Albert Einstein 951, C.P. 6021, 13083-970
Campinas, São Paulo, Brazil. E-mails: [email protected]; [email protected]

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 194 — #2

i i

194 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

INTRODUCTION adopting the values:

The vector of the gravitational force →−
g at a specific point P on ω = 7.292115×10−5 rad s−1 (Petit & Luzum, 2010);
the Earth’s surface is the result of the forces of attraction of the r = R cos ϕ;
→
− →
− R = (a2 b)1/3 ;
terrestrial mass F and the centrifugal force C applied to the
mass of the body stationed at that point. This gravitational accel- ϕ latitude of a point P on the surface of the Earth;
eration is expressed in m/s2 or Newton1 by the SI (International r is the distance P to the center of the Earth;
System of Units). For gravimetrics, however, the Gal (Si, 2006) is a equatorial radius of the Earth;
used in homage to Galileo Galilei2 (Gemael, 1985). b polar radius of the Earth;
The law of gravity proposed by Newton in 1687 states that any ω angular velocity of the Earth.
particle of mass m1 in the universe attracts another mass m2 with Considering an ellipsoidal model, the theoretical gravity,
a force that depends on the product of the two masses and the in- which is the normal gravity at the level of the ellipsoid derived in
verse of the square of the distance d between them, as shown by numerical form from the Somigliana formula, would be (Moritz,
function 1: 1980):
→
− m1 m2 1 + k sin2 ϕ
F =G (1) γ = γe  (4)
d2 1 − e2 sin2 ϕ
where G is gravitational constant.
adopting the values:
It is noteworthy that Newton’s Law is theoretical and refers to
particles. Its elucidation is still unknown, despite the fact that this
γe = 978,032.67715 mGal (Torge, 2001);
law can be observed and proven (Gemael, 2002).
Given that the Earth spins around its axis with an angular ve- γp = 983,218.63685 mGal (Torge, 2001);
→
−
locity ω, there is a centrifugal force C that has a maximum value
of approximately 0.3% of gravity at the equator and that is zero at f = 1/298.257222101 SIRGAS 2000 reference system, GRS80
the poles. Gravity −→g is the resultant vector between the attractive ellipsoid (IBGE, 2005);
and centrifugal forces, according to Eq. (2) (Fig. 1).
where:

bγp
k= −1 (5)
aγe
b = a(1 − f) (6)
a2 − b 2
e2 = (7)
a2

with:
γ normal acceleration of gravity (theoretical gravity);
γe normal acceleration of gravity at the equator line;
γp normal acceleration of gravity at the poles;
f ellipsoidal flattening;
Figure 1 – Vector components of gravity. Figure adapted from Gemael (2002). b minor semiaxis;
e2 square of the primary eccentricity.

−
→ → −
− → The true gravimetric potential at points on the Earth cannot
g =F +C (2)
be determined with a simple mathematical equation because the
where, without considering flattening of the ellipsoid (Gemael, heterogeneity and density variation of the terrestrial body make
2002): the integration impossible, thereby necessitating the measure-
→
−
C = ω2 r (3) ment of values for −
→
g to create models (Gemael, 1985).
1 Issac Newton (1642-1727): English physicist, mathematician, astronomer, and natural philosopher
2 Galileo Galilei (1564-1642): Italian physicist, mathematician, astronomer, and philosopher

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 195 — #3

i i

AMARANTE RR & TRABANCO JLA 195

Gravimetry and equipment t is the period in seconds necessary for the movement
Gravimetry is an area of physical geodesy that has the objective of performed by the pendulum to repeat itself;
measuring the gravitational intensity and/or the gradient of grav- l is the length of the pendulum in meters;
ity with terrestrial methods or methods near the Earth’s surface. α is the angle formed between the vertical and the initial
Gravimetric measurements can be performed using “absolute” or position of the pendulum;
“relative” techniques (Torge, 2001). g is gravitational acceleration.
Determining the value of gravity −
→
g is important because, by The free fall method for determining gravity (Fig. 3) is based
calculating its anomalies and it is possible to determine the un- on the principle of accelerating movement shown in Eq. (9),
dulations N of the geoid using the Stokes’ integral, the terrestrial which can be simplified according to formula 10 (Gemael, 2002).
flattening with the Clairaut3 theorem, and the deviation from the
vertical with the Vening-Meinesz4 formulas (Gemael, 1985). 1
x = x0 + v0 t + gt2 (9)
Gravimetrics can be applied in geophysical investigations to 2
estimate the density of rocks and to determine geoidal undula-  
(x3 − x1 ) (t2 − t1 ) − (x2 − x1 ) (t3 − t1 )
tions; furthermore, gravimetrics can be employed in geodynamic g=2 (10)
(t3 − t1 ) (t3 − t2 ) (t2 − t1 )
studies, in which temporal changes of the terrestrial gravitational
field are monitored. where:

– t1 ; t2 ; t3 are time intervals counted from t = 0;

Absolute methods for measuring gravity
Absolute gravity measurements determine − →
g using the funda- – x1 ; x2; x3 are the distances moved in the intervals
mentals of gravitational acceleration, measurement of time, and t1 ; t2 ; t3 respectively.
displacement. The measurement can be performed using the pen-
dulum method, introduced by Galileo Galilei, which is currently Modern instruments for absolute measurement by the free fall
not used, or by the free fall method (Torge, 2001; Gemael, 2002), method are accurate; however, these machines are costly, com-
which became viable with the technological advance of chronome- plex, difficult to transport, and difficult to set up on irregular or
ters, sensors, and electronics. unstable ground. The basic function of this type of equipment is
The pendulum method presented in Figure 2 can be easily to measure the position at several instants in time for a body in
calculated using Eq. (8). free fall in a vacuum chamber.

Figure 2 – Pendulum determination. Figure adapted from Gemael (2002).

  
l sin α2
t = 2π 1+ (8)
g 16
Figure 3 – Determination with the free fall method. Figure adapted from Gemael
with: (2002).
3 Alexis Claude de Clairaut (1713-1765): French mathematician
4 Felix Andries Vening Meinesz (1887-1966): Dutch geophysicist and geodesist

Brazilian Journal of Geophysics, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 196 — #4

i i

196 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

Relative methods for measuring gravity dynamic. The first occurs when the elastic system is altered in a
Relative methods use equipment to measure the difference be- state of rest; during brief periods, the variation is normally small.
tween the readings at one station versus another. The determina- However, surveying one circuit often takes days; therefore, it is
tion of −
→g at a given point depends on the existence of a gravi- inevitable that the team will break for meals and rest. Dynamic
metric station with a previously known acceleration value, from drift results from deformation of the elastic system, which occurs
which the result can be calculated by summing the difference de- over time and from movements that occur during its displacement
termined between the stations. The functioning principle is based in the circuit in which the work is performed. The time limit is a
on the weight of a body equilibrated by a normally elastic inverse characteristic that varies as a function of the model and brand of
force, provided by a spring in an environment with a constant tem- the equipment and is determined from the period necessary for
perature, to maintain the same coefficient of elasticity between the drift to be treated as linear.
the measurements. The differences in mGal are measured on a
graduated scale, and the results are parameterized according to a
calibration table provided by the manufacturer. One functioning
scheme of the LaCoste and Romberg (L&R) model G gravimeter
can be observed in Figure 4.

Figure 5 – Drift of the relative ET18 and absolute SG gravimeters (Hegewald et

al., 2010).

Some devices (e.g., the L&R gravimeter) depend on more ex-

tensive work to derive the results. Other more modern devices
that have onboard electronics are more precise and handle the
data better, further facilitating their use – as in the CG-5 Scintrex
Autograv gravimeter, which is displayed in Figure 6.

Figure 4 – Schematic section of the LaCoste and Romberg model G&D gravime-
ter (LaCoste & Romberg, 2004).

The emergence of relative measuring equipment revolution-

ized gravimetry, as it permitted the rapid densification of gravi-
metric networks due to the reduced size, ease of use, possibil-
ity of installation in diverse environments, and high productivity
(Gemael, 2002).
In contrast with the absolute equipment, a common charac-
teristic of mechanical elastic equipment and one of its primary in-
conveniences is the existence of a phenomenon known as drift,
which should be considered in gravimetric calculations. To il-
lustrate, a comparison between a piece of equipment based on
free fall and another machine with an elastic system is shown
in Figure 5. There are two different types of drift: statistical and Figure 6 – CG-5 Scintrex Autograv gravimeter.

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 197 — #5

i i

AMARANTE RR & TRABANCO JLA 197

Figure 7 – OSG superconductor gravimeter (GRW, 2011).

The CG-5 is part of a new generation of relative gravimeters Another type of gravimeter is the superconducting gravime-
with microprocessors, which function in the range of 8000 mGals ter, shown in Figure 7. Its functioning principle involves the
with a resolution of 1 μGal, without the need to be reinitiated. Un- levitation of a spherical specimen by a magnetic field generated
like mechanical equipment such as the model G from L&R, the by coils. The sphere moves up and down in response to changes
CG-5 does not record isolated measurements made by one op- in gravity; therefore, the voltage is altered automatically, with the
erator but rather samples in pre-defined intervals or in a con- objective of maintaining it in equilibrium and moving the spec-
tinuous manner at a frequency of 6 Hz. That is, a machine that imen to the correct position. This applied voltage is parameter-
is configured to make three measurements of 1 minute each will ized on a scale to provide the gravity changes. Superconductor
store in its memory (6 Hz) (60 s) (3) = 1080 readings in a series of gravimeters can achieve sensitivities of 1 nGal (nanogal) in their
measurements of one station. The fact that this gravimeter makes measurements (GRW, 2011).
multiple measurements per second permits the incorporation of
a seismic filter. The possibility of storing the collected data in its
12 Megabyte memory, in addition to saving work, avoids many TIDE
→
−
possible errors, such as performing an inaccurate reading of the The value of − →g must be determined from the attractive F and
→
−
graduation from the dial, dictating the wrong value to the recorder, centrifugal C forces, and the other perturbations must be elimi-
writing incorrectly in the field notebook, reading incorrectly from nated, including the tide correction Cg , which is also called the
the field notebook when transcribing the data to a computer pro- gravimetric correction. The essential elements of physics for un-
gram, or even typing incorrectly. To facilitate its use, the equip- derstanding tides are contained in Newton’s laws of dynamics and
ment has a graphic screen, a keyboard, and a user-friendly in- conservation of mass (Pugh, 1987).
terface. The machine also has protection against temperature The total gravitational attraction between two large masses
changes, pressure variations, and variations caused by magnetic such as the Earth and the Moon, as well as their proximity, makes
fields. Another important characteristic of the CG-5 is a small this set a system. All other bodies in space also attract each other;
drift, which results from the operational environment of the elas- however, because the distances are so large, the forces are ex-
tic quartz system in conjunction with the correction by the soft- tremely small, with the exception of the Sun. Despite the distance,
ware that provides results in real time with long-term drift and the Sun has a significantly large mass that makes the Sun-Earth-
0.02 mGals per day (Scintrex, 2008). Moon network a system of large influence between the parts. The

Brazilian Journal of Geophysics, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 198 — #6

i i

198 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

Figure 8 – Tide. Figure adapted from MacMillan (1966).

Earth-Moon system spins around a common center of mass near rection (Gemael, 1985), will be presented to permit the codifica-
the Earth because the mass of the Moon is smaller (approximately tion of a program to calculate the lunar-solar attraction at a spe-
1/81 times the Earth’s mass). Because the Earth-Moon system cific point P on the surface of the Earth at time t with the use of
spins around a common center of mass between the heavenly bod- the formulas presented by Longman (1959) and adjusted using
ies, the centrifugal force is small on the terrestrial face closest to the gravimetric factor. According to Dehlinger (1978), the formu-
the Moon, and it is added to the attractive force; on the opposite las presented by Longman (1959) are essentially identical to those
side of the Earth, the centrifugal force opposes the force of attrac- proposed by Bartels (1957), Schureman (1940), and Pettit (1954).
tion and is equilibrated, also creating a high tide on the opposite The calculation of the tide correction Cg of a point P on the
side of the planet, as observed in Figure 8 (Pugh, 1987). surface of the Earth is given by (Gemael, 2002):
The center of mass, also known as the centroid, is between the
Cg = (g + gL)δ (11)
heavenly bodies at approximately 1,710 km beneath the terrestrial
surface (Xie & Kopeikin, 2010). where:
Tides oscillate on a period of approximately 12 hours and 24 gL is the lunar component of the tide defined by Eq. (14);
minutes. The “12 hours” are due to the rotation of the Earth, and g is the solar component of the tide defined by Eq. (37);
the “24 minutes” are due to the daily delay caused by the lunar δ is the gravimetric factor.
orbit, which has a cycle of 29.5 days. The phenomenon known as The gravimetric factor is used to adjust the gravimetric cor-
the syzygy tide occurs when the Earth, Sun, and Moon are in the rection value as a function of the displacement of the point P that
same alignment (Full Moon and New Moon). During this period, occurs due to deformation of the crust, which is not solid and has
the variation between the high tide and the low tide is the largest. a certain deformation as a function of the lunar-solar attraction.
The quadrature tide occurs when the Moon is waxing or waning, The commonly used programs in Brazil do not allow a user to
phases in which the variations are smoother. The differences in insert the value of the gravimetric factor and adopt fixed factors,
the tide that occur at the syzygy tide and the quadrature tide are as is the case of REDGRAV5 , which always uses δ = 1.20, and
presented in Figure 9. GRAVSYS6, which uses δ = 1.16.
Technological advancements, the creation of more precise in-
CALCULATION OF THE TIDE CORRECTION struments, automated data storage, and the regionalization of the
In this section, the schematization of the operations necessary for gravimetric factor necessitates the construction of a new software
calculating the tide correction, also known as the gravimetric cor- technology to meet the current reality of gravimetric surveys. It is
5 REDGRAV: Software for processing raw gravity data developed by the Universidade de São Paulo, Brazil.
6 GRAVSYS: Computer program for processing and adjustment of gravimetric data developed at the GSC (Geological Survey of Canada) in Ottawa.

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 199 — #7

i i

AMARANTE RR & TRABANCO JLA 199

Figure 9 – Variation of the tide amplitude. Figure adapted from Pugh (1987).

easy to find studies that seek to better determine the gravimetric the trigonometric calculations, angular values were transformed
factor in various regions of the world, such as the study by Has- into radians to make them compatible with the most commonly
san et al. (2010). For a long time, gravimetric campaigns used the used programming languages. The final version tested is used to
value of δ = 1.2 for the gravimetric factor throughout Brazil. How- develop the software GravSur7 (Amarante, 2012).
ever, in a study presented in 1991, regionalization of the gravimet- During the development of this study, some values suggested
ric factor was performed as presented in Table 1. In the present by Longman (1959) were exchanged for constants defined in more
study, tests were conducted with the tide, which are available in recent investigations, such as the mass of the Sun reported by the
section “EVALUATION OF RELEVANCE”. documentation of IERS Conventions (International Earth Rotation
and Reference Systems Service Central Bureau) (Petit & Luzum,
Table 1 – Regionalization of the gravimetric factor (Gemael, 2002).
2010). Some of these constants, despite having their values re-
Station δ
configured, do not significantly alter the result; however, with the
Belém 1.19
advance of equipment that can determine differences in the mea-
Manaus 1.19
surements with more precision, it is also necessary to review val-
Teresina 1.22
ues that may have been calculated with a slide rule.
Caicó 1.23
Salvador 1.19
The time zone is critical for calculating the tide, and much
Goiânia 1.16 confusion can occur when determining the time, which can sig-
Cuiabá 1.16 nificantly alter the result for the tide correction. This confusion of-
Campo Grande 1.17 ten occurs when the difference that is used for UTC (Coordinated
Viçosa 1.20 Universal Time) is not recorded or when the team that makes the
Vassouras 1.19 field measurements is not the same as the one that performs the
Presidente Prudente 1.17 calculations.
Curitiba 1.17 Another important problem is that the official time, which
Santa Maria 1.20
changes as a function of legislation, as is the case of Law
For codification in the computer, the sequence shown in the No. 11,662 (Brasil, 2008) presented in Figure 10, which modi-
mathematical form was altered so that it was possible to imple- fied the boundaries of the regions and the number of time zones
ment it in a programming language because the execution se- in Brazil, as well as the data from the beginning and end of day-
quence of the commands is essential for code functioning. For light savings time. These changes in the rules for calculating the
7 GravSur: Computer program for processing and adjustment of gravimetric data developed at the Universidade Estadual de Campinas and Universidade de São Paulo.

Brazilian Journal of Geophysics, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 200 — #8

i i

200 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

Figure 10 – Change in the time zones in Brazil in 2008 (Worldtimezone, 2011).

official time can cause an incorrect choice when the calculation is given by (Gemael, 2002):
is performed at some time after the fieldwork. To avoid problems GMi ri  
gi = 3
3 cos2 θi − 1 δ (12)
of this type, a protocol can be adopted, as was adopted in civil ς
aviation, which determines the use of the UTC times. adopting the values:
The influence of the Sun can reach approximately 0.063 mGal, G = 6.67428 × 10−8 m3 g−1 s−2 (Petit &
while that of the Moon can reach up to 0.168 mGal (Gemael, Luzum, 2010);
2002); these values must be considered because they are far gi = influence of the heavenly body on the tide de-
above the 0.01 mGal resolution of gravimeters such as the L&R fined by Eq. (12);
model G, which are frequently used in campaigns in Brazil and M♀ = 4.8685 × 1027 g (Venus mass);
Latin America. Currently, these machines are being replaced by M = 1.8987 × 1030 g (Jupiter mass);
more advanced models, such as the CG-5 Autograv gravime- M♂ = 6.4190 × 1026 g (Mars mass);
ter from the Scintrex brand with a resolution of 0.001 mGal or ri = 6.378137 × 108 cm;
1 μGal, which is even more precise. Venus, with a mass of ap- ς♀ = 3.82 × 1012 m (Venus);
proximately 4.8685 ×1024 kg, can reach a minimum distance of ς = 5.8850 × 1013 m (Jupiter);
38,200,000 km (Williams, 2010) and thus affects the tide by up to ς♂ = 5.5760 × 1012 m (Mars);
0.00862 μGal. Jupiter, with its large mass of 1.8987 × 1030 kg, θi = 0◦ to simulate the largest influence;
reaches a nearest distance of 588,500,000 km (Williams, 2012), δ = 1.17 for the region of Presidente Prudente
which has an influence of up to 0.00092 μGal. Mars, with its ap- (Gemael, 2002);
proximate mass of 6.419 × 1023 kg (Beatty et al., 1999), can with:
reach a minimum distance of 55,760,000 km (McKim, 2003) and G is Newton’s gravitational constant;
can thus cause an influence of up to 0.00037 μGal, as demon- Mi is the mass of the perturbing heavenly body in grams
strated in Eqs. (12) and (13). Therefore, it is still not necessary to (M♀ or M or M♂ );
consider the most significant close heavenly bodies such as the ri is the distance P in cm to the center of the Earth;
planets Venus, Jupiter, and Mars when calculating the tide be- ς is the maximum approximation of the perturbing
cause these planets’ influences can achieve values well below the heavenly body (ς♀ or ς or ς♂ );
resolutions of the (relative) pieces of equipment, as CG-5, cur- θi is the zenith angle of the perturbing heavenly body;
rently in use. δ is the gravimetric factor.
The influence on the tidal correction due to a heavenly body Considering the example of the perihelion opposition of Mars

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 201 — #9

i i

AMARANTE RR & TRABANCO JLA 201

in August 2003, when it is very close to the Earth, the influence of Hough ellipsoid, while GRAVSYS uses a = 6, 378, 388 from
the planet in the tide can reach 0.37 nGal. Applying similar con- Hayforf (1924). Only updating this value for the GRS80 ellipsoid
ditions to Jupiter and Venus has respectively values of 0.92 nGal in a future program is not interesting because the same applica-
and 8.62 nGal. tion can be used in other seasons and regions of the world, where
The calculation of the tide correction Cg of a point P on the these values can be different. Equation 16 for the calculation of C,
surface of the Earth, considering next most significant planets is Longman (1959), uses the constant 0.006738 default in place of
given by: e2
e , which is only appropriate for the Hough Datum (1906).

Cg = (g + gL + g♀ + g + g♂ + · · · )δ (13) C = 1/(1 + e2 2
e sin ϕ) (16)

with: where:
g♀ is the Venus component of the tide Eq. (12);
g is the Jupiter component of the tide Eq. (12); a2 − b 2
e2
e = (17)
g♂ is the Mars component of the tide Eq. (12). b2
b = a − (af) (18)
Lunar Component
adopting the value:
The lunar component gL is giving by:
f = 1/298.257222101 SIRGAS 2000 reference system
GML r GRS80 ellipsoid (IBGE, 2005);
gL = (3 cos2 θL − 1)
d3L with:
(14) ϕ is the latitude from point P in decimal degrees;
3 GMLr 2
+ (5 cos3 θL − 3 cos θL ) e2 is the square of the second eccentricity of the ellipsoid;
2 d4L e
b is the minor semiaxis;
adopting the values: f is the flattening of the ellipsoid.
ML = 7.34581119761 × 1025 g, source: IERS
Calculation of the distance dL between the centers of the
data (Petit & Luzum, 2010).
where: Earth and the Moon is obtained with:
G is the Newton’s gravitional constant;
d−1
L = 1/c + a e cos(s − p) + a e2 cos 2(s − p)
ML is the Mass of the Moon in grams;
+ (15/8)a me cos(s − 2h1 + p) (19)
r is the distance P to the center of the Earth defined  2
+ a m cos 2(s − h1 )
in the Eq. (15);
dL is the distance between the centers of the Earth and adopting the values:
the Moon, as defined by the Eq. (19); c = 3.844031 × 1010 cm;
θL is the zenith angle of the Moon, shown in Eq. (24). e = 0.05490 (Schureman, 1940);
The distance P to the center of the Earth is given by: m = 0.074804 (Schureman, 1940);
with:
r = Ca + 100H (15) c is the average distance between the centers of the
Earth and the Moon;
adopting the value:
e the eccentricity of the orbit of the Moon;
a = 6.378137×108 cm SIRGAS 2000 reference sys-
a is defined by Eq. (20);
tem GRS80 ellipsoid (IBGE, 2005);
where: s is the average longitude of the Moon referring to the
C is the value defined in Eq. (16); equinox (Bartels, 1957);
a is the equatorial radius of the Earth; p is the average longitude of the perigee of the Moon
H is the orthometric height (multiplication by 100 is for defined by Eq. (23);
the transformation into cm). m is the average rate of movement of the Sun to the Moon

The Earth’s equatorial radius of a = 6, 378, 270, shown where:

in Longman (1959) and used by REDGRAV, refers to the 1906 a = 1/ c(1 − e2 ) (20)

Brazilian Journal of Geophysics, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 202 — #10

i i

202 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

To determine s and p, it is first necessary to know the num- l is the longitude of the Moon in its orbit counted from
ber of Julian “centuries” Tj since December 31, 1899, which is the crossing of the ascension with the equator line de-
calculated with: fined by Eq. (31);
ϕ is the latitude of point P ;
Tj = t/52, 596, 000 (21) A is the ascending intersection of the lunar orbit with the
equator;
with: χ is the ascension to the right of the meridian of observa-
t is the instant of determination of the tide in the Julian tion counted from A (Fig. 11), obtained using Eq. (27);
calendar; i is the constant given by Schureman (1940) used in
Tj adopting a Julian century as (100 years · 365.25 days Eq. (28) referring to the angle between the lunar orbit
· 24 hr · 60 min) = 52,596,000 minutes. and the plane of the ecliptic (Fig. 11).
The average longitude of the Moon referring to the equinox Determination of the angle I between the lunar orbit and
is given by: the celestial equator is given by:

s = 270◦26 11.72 cos I = cos ωt cos i − sin ωt sin i cos Nλ (25)


+ (1336rev. + 1, 108, 406.05 )Tj (22)
adopting the value:
+ 7.128Tj2 + 0.0072Tj3
i = 5◦ 08 43.3546
For the calculation of the average longitude of lunar perigee, The longitude Nλ from the ascending node to the orbit of
the following equation is used: the Moon is defined by:

p = 334◦19 46.42 Nλ = 259◦10 57.12

+ (11rev. + 392, 522.51)Tj (23) − (5rev. + 482, 912.63)Tj (26)
− 37.15Tj2 − 0.036Tj3 + 7.58Tj2 + 0.008Tj3

The REDGRAV program uses the eccentricity of the lunar orbit and χ is given by:
suggested by Shureman (1924); however, this study uses a new
χ = ta + h  − v (27)
value given by the 1940 revision of the book (Schureman, 1940).
Despite this difference, the calculation does not yield significant where:
results. One purpose of this paper is to help in the implementa-
v = arcsin (sin i sin N/ sin I) (28)
tion of algorithms in future gravimetric data processing programs.
Therefore, the values of s, p, h and all other variables that rep- The angular time ta of the average measurement of the Sun
resent angles from equations suggested in Longman (1959) must to the west of the location of observation is defined by:
be transformed into radians to permit the use of the computational
functions. The abbreviation rev., shown in Eqs. (22) and (23), ta = 15 (t0 − 12) − λ (29)
refers to revolutions; that is, each unit must be multiplied by 360◦ .
Average longitude h from the Sun is given by:
The zenith angle of the Moon θL is determined by:
h = 279◦ 41 48.05
cos θL = sin ϕ sin I sin l (30)
  + 129, 602, 768.11Tj + 1.080Tj2
2 I 2 I
(24)
+ cos ϕ cos cos(l − χ) + sin cos(l + χ)
2 2 where:
t0 is UTC time in whole Julian hours;
with: λ is longitude from the point P ;
I is the angle between the orbit of the Moon and the ta is the angular time of the average measurement of
celestial equator according to (Fig. 11) defined by Eq. (25); the Sun to the west of the location of observation.

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 203 — #11

i i

AMARANTE RR & TRABANCO JLA 203

Figure 11 – Orbital parameters (Schureman, 1940).

The longitude of the Moon l is given by: adopting the values:

5
l = σ + 2e sin(s − p) + e2 sin 2(s − p)
4 M = 1.9884158 × 1033g, source: IERS data (Petit &
15 (31) Luzum, 2010);
+ me sin(s − 2h1 + p)
4
+ 1.375m2 sin 2 (s − h1 ) with:
M is the mass of the Sun in grams;
with:
Υ is the vernal equinox (Fig. 11); d is the distance between the centers of the Earth and
Υ is the related equinox (Fig. 11); the Sun, as defined by Eq. (38);
Ω is the node from the beginning of the rising of the θ is the zenith angle of the Sun, as defined by Eq. (40).
Moon (Fig. 11);
Calculation of the distance d between the centers of the
The average longitude of the Moon σ given by the relation-
Earth and the Sun is given by:
ship:
σ = s−ξ (32)
1
d = (38)
and the longitude ξ in the intersection of the orbit with the celes- c + a1 e1 cos (h − p )
tial equator is defined by:  

a1 = 1/ c 1 − e21 (39)

ξ = Nλ − α (33)
where: adopting the value:

α = 2 arctan [sin α/ (1 + cos α)] (34) c = 1.49597870691 × 1013 cm] (Seidelmann,

sin α = sin ωt sin Nλ sin I (35) 1992);

cos α = cos Nλ cos v + sin Nλ sin v cos ωt (36) with:

c is the average distance between the centers of the
Solar component Earth and the Sun;
The influence of the solar component g on the measurement of a1 is defined by Eq. (39);
the acceleration of a point is determined by Longman (1959): e1 is the eccentricity of the orbit of the Earth;
GM r   h is the average longitude of the Sun;
g = 3 cos2 θ − 1 (37)
3
d p is the average longitude of solar perigee8 .
8 perigee: point in the orbit of an object in which it is closest to the Earth.

Brazilian Journal of Geophysics, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 204 — #12

i i

204 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

The zenith angle of the Sun θ is defined by Equation: The differences were compared to the resolution of modern
devices such as the Scintrex CG-5 and the OSG superconductor
cos θ = sin ϕ sin ωt sin l
 gravimeters to calculate the significance of the above-calculated
ωt
+ cos λ cos2 cos (l − χ ) factors.
2 (40)
 By comparing the results from the tide calculations gener-
2 ωt
+ sin cos (l − χ ) ated by the GRAVSYS and REDGRAV with the approach suggested
2
in this study implemented in the GravSur computer program
adopting the value: (Amarante, 2012), it was possible to note differences. The RED-
GRAV program is important because it is used in Brazil and in
ωt = 23◦ 27’08.26” (Schureman, 1940);
certain Latin American countries by the EPUSP (Escola Politéc-
where: nica, Universidade de São Paulo) and IBGE (Instituto Brasileiro
ωt is the obliqueness of the ecliptic on January 1, 1900; de Geografia e Estatı́stica) to calculate the gravimetric reductions.
l is the longitude of the Sun in the elliptic counted from Table 2 contains some points distributed in the four quadrants
the Vernal Equinox; of the globe in UCT time; some observations merit consideration.
χ is the ascension to the right of the meridian of obser- First, GRAVSYS does not allow the input of the year with four dig-
vation counted from the Vernal Equinox; its, which does not permit its use after 1999. A difference is also
ta is the angular time of the average measurement of the noted that apparently is not large but becomes evident when the
Sun to the west of the location of observation. amplitudes of the tides are larger. Notably, GravSur has a con-
The longitude l of the Sun in the elliptic counted from the figurable gravimetric factor, and the calculation was performed
Vernal Equinox is given by: with δ = 1.17; by contrast, REDGRAV and GRAVSYS have fixed
values of 1.20 and 1.16, respectively. Another important point
l = h + 2e1 sin (h − p ) (41)
is that the CG-5 can perform the tide correction automatically;
The ascension χ to the right of the meridian of observation however, it always uses a gravimetric factor of 1.16, because this
counted from the Vernal Equinox is obtained by: value cannot be altered (Scintrex, 2008).
To best illustrate the difference, the example below was cal-
χ = t a + h  (42) culated at the amplitude peak in a spring (syzygy) tide day:
The eccentricity of the Earth’s orbit is calculated by: Location: Santos SP, Brazil

e1 = 1.675104 × 10−2 • Position: 23◦ 57’S 046◦ 18’W

− 4.180 × 10−5 Tj (43) • Date: FEB/18/2011 – 15:20 UCT

− 1.26 × 10−7 Tj2 • GravSur: 0.188 mGals
The average longitude of solar perigee p is defined by: • REDGRAV: 0.193 mGals
p = 281◦ 13 15 • GRAVSYS: 0.186 mGals

+ 6, 189.03 Tj (44)
A difference of up to 0.007 mGals was calculated, which is a

+ 1.63 Tj2 + 0.012 
Tj3 significant value because currently used equipment, such as the
CG-5, has a resolution of 0.001 mGals. This difference tends to
METHOD FOR EVALUATION OF RELEVANCE zero when the same gravimetric factor is used; it is emphasized
(GRAVIMETRIC FACTOR) that the value of δ = 1.17, referring to the region of Presidente
This study aimed to evaluate the extent to which the calculation of Prudente, is the most appropriate for the city of Santos accord-
the tide correction with different gravimetric factors is representa- ing to a publication about regionalization of the gravimetric factor
tive. For the gravimetric factor, from a geographical position, the from Gemael (2002).
differences between the corrections values of the tide calculated Another important fact related to REDGRAV and GRAVSYS
were determined by various programs/machines using the most is that neither system considers seconds when recording the time
recommended value for the region, as indicated in (Table 1). of the reading. Differences of up to 0.001 mGals are observed in

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 205 — #13

i i

AMARANTE RR & TRABANCO JLA 205

Table 2 – Comparison of the tide calculation by different programs and local time (mGals).
Date/Time Location GravSur REDGRAV GRAVSYS
◦ ◦
2010/10/31 8:10 22 44’S 90 30’W –0.055 –0.056 –
2010/10/31 8:10 22◦44’S 90◦30’E –0.003 –0.003 –
2010/10/31 8:10 22◦44’N 90◦ 30’W –0.007 –0.007 –
2010/10/31 8:10 22◦44’N 90◦30’E –0.052 –0.053 –
1996/10/31 8:10 22◦44’S 90◦ 30’W 0.028 0.029 0.028
1996/10/31 8:10 22◦44’S 90◦30’E 0.120 0.123 0.119
1996/10/31 8:10 22◦44’N 90◦ 30’W 0.123 0.126 0.122
1996/10/31 8:10 22◦44’N 90◦30’E 0.029 0.030 –0.029

20-second cycles, which was not very important when using me- BARTELS J. 1957. Geophysik II/ Geophysics II. Handbuch Der Physik
chanical equipment. Presently, the new types of equipment per- Encyclopedia of Physics/ Geophysik/ Ge. Springer-Verlag.
mit the storage of the time with the assistance of a GPS receiver, BEATTY JK, PETERSEN CC & CHAIKIN A. 1999. The New Solar System.
which guarantees good quality. Finally, the CG-5 gravimeter also 4th ed., Sky Publishing Corporation, Cambridge, Massachusetts.
has a fixed gravimetric factor and, as long as the firmware is
BRASIL. 2008. Lei no. 11.662, 24 de dez. de 2008, Brası́lia, DF, Brazil.
not updated to permit configuration of this parameter, in studies
DEHLINGER P. 1978. Marine Gravity. Vol. 22 of Elsevier Oceanography
where accurate data are sought, the correct approach would be to
Series. Elsevier.
import data into programs that permit the treatment of the data, as
is the case for GravSur. GEMAEL C. 1985. Determinação da gravidade geodésica. Imprensa Uni-
versitária – UFPR, Universidade Federal do Paraná, Curitiba, Brazil.
CONCLUSIONS GEMAEL C. 2002. Introdução à Geodésica Fı́sica. 2nd ed., Editora da
Gravimetry adopts the formulas suggested in the study by Long- UFPR, Curitiba, Paraná, Brazil.
man (1959) for the gravimetric correction, which was calculated GRW. 2011. GWR Observatory Superconducting Gravimeter. Available
with convenient and known orbital and ellipsoidal parameters for on: <http://www.gwrinstruments.com>.
the time. Despite this fact, some computer programs adopt equiv- HASSAN RM, ABDELRAHMAN EM, TEALEB A, ZAHRAN K & JENTZSCH
alent values, which do not yield significant results where, for ex- G. 2010. Gravimetric Tide observation at Lake Nasser Region, Aswan,
ample, the ellipsoidal parameters are slightly different. Egypt. Marees Terrestres Bulletin d’informations, 146: 11797–11805.
Superconducting gravimeters can reach sensitivities of 1 nGal
HEGEWALD A, JENTZSCH G & JAHR T. 2010. Comparison of the La-
(nanogal). The gravimetric correction applied to the processing of Coste & Romberg gravimeter ET18 with the superconducting gravimeter
data generated from equipment like this should consider the in- CD-034 at the Geodynamic Observatory Moxa (Germany). Marees Ter-
fluence of the bodies with significant attractive force. When Venus restres Bulletin d’informations, 146: 11781–11788.
is in opposition with the Earth can modify the result of gravimetric
IBGE – Instituto Brasileiro de Geografia e Estatı́stica. 2005. Resolução
correction until 8.62 nGal. 1/2005. Altera a caracterização do Sistema Geodésico Brasileiro, Brazil.
The gravimetric factor, which can vary according to the re-
LACOSTE & ROMBERG. 2004. Instruction Manual Model G&D Gravity
gion, can represent differences of approximately 0.01 mGal and
Meters, Austin, 127 pp.
even so, computer programs such as REDGRAV, GRAVSYS, and
current equipment maintain this value as a constant and does not LONGMAN IM. 1959. Formulas for computing the tidal accelerations
allow set a different factor. due to the Moon and the Sun. Journal of Geophysical Research, 64(12):
2351–2355.

REFERENCES MacMILLAN DH. 1966. Tides. C R Books Limited, London, UK.

McKIM R. 2003. The great perihelic opposition of mars, 2003: Part 1.
AMARANTE RR. 2012. Systematization of gravimetric data process-
Journal of the British Astronomical Association, 120: 280–295.
ing applied to geoid model determination. Ph.D. Thesis, Universidade
Estadual de Campinas, School of Civil Engineering, Architecture and MORITZ H. 1980. Geodetic reference system 1980. Bulletin Géodésique,
Urban Design, Campinas, São Paulo, Brazil. 54: 395–405.

Brazilian Journal of Geophysics, Vol. 34(2), 2016

i i

i i
 i i

“main” — 2018/3/27 — 10:29 — page 206 — #14

i i

206 CALCULATION OF THE TIDE CORRECTION USED IN GRAVIMETRY

PETTIT JT. 1954. Tables for the computation of the tidal accelerations of Poids et Mesures, 8th ed.
the Sun and Moon. Vol. 35. Transactions American Geophysical Union.
TORGE W. 2001. Geodesy. 3rd ed., Walter de Gruyter.
PETIT G & LUZUM B. 2010. IERS Conventions (2010). Verlag, Germany.
WILLIAMS DR. 2010. Nasa Venus Fact Sheet. Available on: <http://
PUGH DT. 1987. Tides, Surges and Mean Sea-Level. John Wiley & nssdc.gsfc.nasa.gov/planetary/factsheet/venusfact.html>.
Sons Ltd., Natural Environment Research Council Swindon, UK ISBN
WILLIAMS DR. 2012. Nasa Jupiter Fact Sheet. Available on: <http://
0 471 91505 X.
nssdc.gsfc.nasa.gov/planetary/factsheet/jupiterfact.html>.
SCHUREMAN P. 1940. Manual of harmonic analysis and prediction of
WORLDTIMEZONE. 2011. New Brazil time zones map and old Brazil
tides. No. 98. U.S. Department of Commerce. Special Publication.
time zones map. Available on: <http://www.worldtimezone.com/brazil-
SCINTREX. 2008. CG-5 Scintrex Autograv System operation Manual. time-new-old.php>.
4th ed.
XIE Y & KOPEIKIN S. 2010. Post-Newtonian Reference Frames for
SEIDELMANN PK. 1992. Explanatory supplement to the Astronomical Advanced Theory of the Lunar Motion and for a New Generation of
Almanac. University Science Books.
Lunar Laser Ranging. Acta Physica Slovaca. Reviews and Tutorials, 60:
SI. 2006. The International System of Units. Bureau International des 393–495.

Recebido em 23 junho, 2014 / Aceito em 3 novembro, 2016

Received on June 23, 2014 / Accepted on November 3, 2016

Revista Brasileira de Geofı́sica, Vol. 34(2), 2016

i i

i i