1. Gravity anomalies result from local variations in topography and subsurface density and have amplitudes of several mgal.
2. Various corrections must be applied to gravity measurements including tidal, Eötvös, latitude, terrain, Bouguer plate, and free-air corrections.
3. Gravity anomalies can provide information about subsurface density variations and structures. Common models used to interpret anomalies include those of a uniform sphere, infinite line, infinite cylinder, and semi-infinite horizontal sheet.
1. Gravity anomalies result from local variations in topography and subsurface density and have amplitudes of several mgal.
2. Various corrections must be applied to gravity measurements including tidal, Eötvös, latitude, terrain, Bouguer plate, and free-air corrections.
3. Gravity anomalies can provide information about subsurface density variations and structures. Common models used to interpret anomalies include those of a uniform sphere, infinite line, infinite cylinder, and semi-infinite horizontal sheet.
1. Gravity anomalies result from local variations in topography and subsurface density and have amplitudes of several mgal.
2. Various corrections must be applied to gravity measurements including tidal, Eötvös, latitude, terrain, Bouguer plate, and free-air corrections.
3. Gravity anomalies can provide information about subsurface density variations and structures. Common models used to interpret anomalies include those of a uniform sphere, infinite line, infinite cylinder, and semi-infinite horizontal sheet.
1. Gravity anomalies result from local variations in topography and subsurface density and have amplitudes of several mgal.
2. Various corrections must be applied to gravity measurements including tidal, Eötvös, latitude, terrain, Bouguer plate, and free-air corrections.
3. Gravity anomalies can provide information about subsurface density variations and structures. Common models used to interpret anomalies include those of a uniform sphere, infinite line, infinite cylinder, and semi-infinite horizontal sheet.
Average gravity on the Earths surface is about g=9.8 m/s 2 , and varies by ~5300 mgal (about 0.5% of g) from pole to equator. (1 mgal=10 -5 m/s 2 ) Gravity anomalies are local variations in gravity that result from topographic and subsurface density variations, and have amplitudes of several mgal and smaller.
Measurement of Absol ute Gravi ty: Pendul um Method: Measure the period
! T = 2" I mgh = 2" L g
To measure 1 mgal variation, the period must be measured to within 1s. Free-fal l Method: Measure the fall of a mass:
! z = z 0 +ut + gt 2 2 To measure 1 gal variation, time must be measured to within 1ns. Ri se-and-fal l Method: Measure time T for a thrown ball to rise and fall a height z:
! z = g T 2 ( ) 2 2. Then
! g = 8 z 1 " z 2 ( ) T 1 2 "T 2 2 ( ) . gal precision; not portable. Measurement of Rel ati ve Gravi ty: Stabl e Gravi meter: Measure !s, the change in a springs length:
! "g = k m "s Unstabl e Gravi meter: Use a spring with built-in tension, so:
! "g = k m s (LaCost-Romberg gravimeter) Usage: Adjust the spring length to zero using a calibrated screw. Sensi ti vi ty: 0.01 mgal for a portable device. Superconducti ng Gravi meter: Suspend a niobium sphere in a stable magnetic field of variable strength. Sensitivity: 1 ngal GG612 Lecture 4 2/3/11 2 Clint Conrad 4-2 University of Hawaii Gravi ty Correcti ons
Dri ft Correcti on: In relative gravity surveys, instrument drift can be corrected by making periodic measurements at a base station with known gravity.
Ti dal Correcti on: Gravity changes during the day due to the tides in a known way. Tidal corrections can be computed precisely if time is known. For example, if the moon is directly overhead, the tidal correction would be:
! "g T = G M L r L 2 2R E r L +3 R E r L # $ % & ' ( 2 +... # $ % % & ' ( ( This should be added to measured gravity.
Etvs Correcti on: Moving eastward at v E , your angular velocity increases by:
! "# = v E R E cos$ ( ) . This change increases the centrigugal acceleration:
! "a C = da C d# $ % & ' ( ) "# = 2#R E cos* ( ) v E R E cos* $ % & ' ( ) = 2#v E . Downward gravity changes by:
! "g = #2$v E cos%. The Etvs effect decreases gravity when moving east.
Lati tude Correcti on: Absolute gravity is corrected by subtracting normal gravity on the reference ellipsoid:
! g n = g e 1+ " 1 sin 2 # + " 2 sin 4 2# ( )
where
! g e = 9.780327 m/s 2 ,
! " 1 = 5.30244#10 $3 , and
! " 2 = #5.8$10 #6 . Relative gravity is corrected by differentiating g n with respect to !:
! "g lat = 0.8140sin2# mgal per km north-south displacement. This correction is subtracted from stations closer to the pole than the base station.
Terrai n Correcti on: Nearby topography perturbs gravity measurements upward due to mass mass excess above the station (nearby hills) or due to mass deficiency below the station (nearby valleys). The terrain correction is computed using: GG612 Lecture 4 2/3/11 3 Clint Conrad 4-3 University of Hawaii h 2. remove layer!"g BP
1. Drop to ellipsoid!"g FA
! "g = G dmcos# ( ) r 2 + z 2 ( )
where r and z are the horizontal and vertical distances to dm, and " is the angle to the vertical. The terrain correction is always positive. Integrating over a sector gives:
! "g T = G#$ r 2 + h 2 % r 1 & ' ( ) * + % r 2 + h 2 % r 2 & ' ( ) * + & ' ( ) * + r 1 and r 2 are the inner and out radii, h is the height, # is the sector angle.
Bouguer Pl ate Correcti on: This correction compensates for a rock layer of thickness h between the measurement elevation level and the reference level. For a solid disk of density $ and radius r, the terrain correction is:
! "g T = 2#G$ h % r 2 % h 2 % r & ' ( ) * + & ' ( ) * + . Allowing r to become infinite, we obtain:
! "g BP = 2#G$h = 0.0419%10 &3 $ mgal/m if $ is in kg/m 3 . This correction must be subtracted, unless the station is below sea level in which case a layer of rock must be added to reach the reference level. For gravity measured over water, water must be replaced with rock by assigning a slab with density
! " rock - " water ( ) .
Free-ai r Correcti on: This correction compensates for gravitys decrease with distance from the Earths surface. It is determined by differentiating g:
! "g FA = # #r $G M E r 2 % & ' ( ) * = +2G M E r 3 = $ 2 r g = 0.3086 mgal/m This correction must be added (for stations above sea level). GG612 Lecture 4 2/3/11 4 Clint Conrad 4-4 University of Hawaii
Combi ned Correcti on: Free air and Bouguer corrections are often combined:
! "g FA + "g BP = 0.3086# 0.0419$ %10 #3 ( ) mgal/m= 0.197 mgal/m assuming a crustal density of 2670 kg/m 3 . To obtain 0.01 mgal accuracy: -- location must be known to within 10 m (for latitude correction) -- elevation must be known to within 5 cm (for combined correction)
Geoi d Correcti on: For long wavelength surveys, station heights must be corrected for the difference in gravity between the geoid height and the reference ellipsoid, which can vary spatially.
Densi ty determi nati on Knowledge of the density of subsurface rocks is essential for the Bouguer and terrain corrections. Density can be measured in several ways: ! By measuring the density of rocks on the surface ! Using seismic velocity measurements (velocity increases with density) ! By applying the combined correction to depth variations in gravity measurements in a borehole. Assuming two measurements are separated by a height "h and using the lower station as a reference level, The gravity correction at the upper borehole (free-air decreases gravity and Bouguer slab between the stations increases gravity) is:
! "g upper = "g FA + "g BP = 0.3086# 0.0419$ %10 #3 ( ) "h The gravity correction at the lower borehole (Bouguer slab between the stations decreases gravity) is:
! "g lower = "g FA + "g BP = 0+0.0419# $10 %3 ( ) "h Subtracting the two and solving for density $ gives:
! " = 3.683#11.93$g $h ( ) %10 #3 km/m 3
GG612 Lecture 4 2/3/11 5 Clint Conrad 4-5 University of Hawaii Gravi ty Anomal i es
After the appropriate corrections are applied, gravity data reveal information about subsurface density heterogeneity. How should this data be interpreted?
Gravi ty over a Uni form Sphere Gravity for a sphere is the same as for a point mass. The z-component:
! "g z = "gsin# = G M r 2 z r where
! M = 4" 3 R 3 #$ and ! r 2 = z 2 + x 2 giving:
! "g z = 4# 3 G "$R 3 z 2 % & ' ( ) * z 2 z 2 + x 2 % & ' ( ) * 3/ 2
The maximum is at x=0, where:
! "g zmax = 4# 3 G "$R 3 z 2 % & ' ( ) * Rule of thumb: ! z = 0.65w where w is the width at half height of the anomaly.
Gravi ty over an I nfi ni te Li ne An infinitely long line of mass m per unit length produces a gravity anomaly:
! "g z = 2Gmz z 2 + x 2 where z and x are the vertical and horizontal distances to the line.
Gravi ty over an I nfi ni te Cyl i nder An infinitely long cylinder is a useful analogue for a buried syncline or anticline.
! "g z = 2#G "$R 2 z % & ' ( ) * z 2 z 2 + x 2 % & ' ( ) * and
! "g zmax = 2#G "$R 2 z % & ' ( ) * Rule of thumb: ! z = 0.5w. A structure must be more than 20% longer than it is wide or deep for the infinite approximation to be valid (ignore edge effects). GG612 Lecture 4 2/3/11 6 Clint Conrad 4-6 University of Hawaii x 3/4
#g/4
Gravi ty over a Semi -I nfi ni te Hori zontal Sheet A horizontally truncated thin sheet can be used to approximate a bedded formation offset by a fault. If the fault is centered at x=0, z 0 =0, then the gravity anomaly is:
! "g z = 2G"#h $ 2 + tan %1 x z 0 & ' ( ) * + & ' ( ) * + Rule of thumb: z 0 ~x 1/4 ~x 3/4
Where x 1/4 and x 3/4 are the positions where the gravity anomaly is " and # its max value. Note that as x&$,
! "g z = 2G"#h, which is the solution for a Bouguer Plate anomaly.
Gravi ty anomal y of arbi trary shape Any shape can be approximated as an n-sided polygon, the gravity anomaly of which can be computed using Talwanis algorithm. This algorithm estimates gravity by computing a line integral around the perimeter:
! "g z = 2G"# zd$ %
I sostasy Long wavelength variations in topography are isostatically compensated at depth. This means that the excess mass in positive topography is compensated by a mass deficiency at depth. There are three types of isostasy.
GG612 Lecture 4 2/3/11 7 Clint Conrad 4-7 University of Hawaii Airy Isostasy Pratt Isostasy $ c
$ m
$ m
$ c
C C
Ai ry I sostasy: Lateral variations in crustal thickness allow surface topography to be compensated by a deep crustal root. The thickness of this root is determined by requiring the mass in columns above the compensation depth (C) to be equal:
! r 1 = " c " m # " c h 1 or, if the topography is under water,
! r 1 = " c # " w " m # " c h 1
Pratt I sostasy: Lateral variations in crustal density compensate topography, so again the mass in columns above the compensation depth (C) are equal. The density is:
! " 1 = D h 1 +D " c or, if the column is a depth d under water,
! " 0 = " c D# " w d D# d
Veni ng Mei nesz I sostasy: In this type of isostasy, short-wavelength topography is supported by the elastic strength of the crustal rocks. The load is instead distributed by the bent plate over a broad area. This distributed load is compensated. GG612 Lecture 4 2/3/11 8 Clint Conrad 4-8 University of Hawaii Gravi ty Anomal i es over Topography Uncompensated topography (Short-wavelengths) Free-air anomaly (apply the free-air correction only): "g+"g FA >> 0 because of the topographys excess mass Bouguer anomaly (apply both free-air and Bouguer plate corrections): "g+"g FA -"g BP ~ 0 because Bouguer corrects for excess mass. Compensated topography (Long-wavelengths) Free-air anomaly (apply the free-air correction only): "g+"g FA ~ 0 because topography is compensated (no excess mass) Bouguer anomaly (apply both free-air and Bouguer plate corrections): "g+"g FA -"g BP << 0 because Bouguer removes additional mass. Undercompenstated topography: A too-shallow root, yields "g+"g FA >0 Overcompensated topography: A too-shallow root, yields "g+"g FA <0
Geoi d Anomal i es over Topography The gravitational acceleration can be approximated as:
! g = " #U #r $U = $U $N where "N is the change in geoid height We can approximate the change in potential as:
! g"N ~ "U ~ # "gdz 0 z c $ = # 2%Gz"& z ( ) dz 0 z c $ where z is positive downwards from the surface to the compensation depth z c
Then the geoid height can be written as:
! "N ~ # 2$G g 0 z"% z ( ) dz 0 z c & This is non-zero because
! "# z ( ) dz 0 h $ = 0 for isostasy. Thus, the geoid anomaly should be positive over compensated positive topography (e.g., the continental lithosphere, mid-ocean ridges, Tibet, Andes) The geoid gives better constraints on the depth-distribution of mass than does gravity.