GEOL 4714/5714 Field Geophysics

 

Gravity Data Interpretation

Due 10/17/02 (field day 10/8/02, lab help 10/15/02 — mandatory for ugrads)

 

Gravity data processing involves several steps, outlined below. We will first use Microsoft Excel to make a spreadsheet of our gravity data and correct for instrument drift, latitude, and elevation. We will then use the program GravModel to model the gravity data.

Steps of gravity data processing:

1. In Excel, Make a table of gravity readings (use a site average), including station name, dial reading, time (in minutes), survey information (azimuth, horizontal distance, vertical distance).

1b. Use azimuth and horizontal distance to determine meters north or south of the survey station (use a little trig.). Make sure your answers make sense. To do cosine of 16 degrees using Excel, for example, you have to convert to radians. So you would type cos (16*PI()/180) in Excel.

2. Drift correction: make a plot of readings at the base station as a function of time, linearly interpolate between the points, and subtract this value from subsequent readings as a function of time. The form of the drift correction is:

drift corrected reading = ri - ti (bend - bbeginning) / (end time - beginning time)

where ri is reading at time i (can be at any station), ti is time i (in minutes), bend is base station reading at end of study and bbeginning is base station reading at beginning of study. Since we will have several different base station readings we will come up with several different slopes, so readings from the beginning until the 2nd reading will have a different correction from readings from the 2nd reading to the 3rd reading, and so on.

A check on the drift correction is that all drift corrected base station readings should be identical.

3. Convert the data from dial readings to milligals. The conversion factor for our gravimeter is 1.0544.

milligals = drift corrected reading x 1.0554

4. Latitude correction. Gravity varies with latitude because the Earth is not a perfect sphere and the polar radius is smaller than the equitorial radius, also, the effect of centrifugal acceleration is different at the poles versus the equator. Thus gravitational acceleration is larger at the poles than at the equator. At the latitude of Colorado, the variation is 88 mgal per degree, which is 0.000786483 mgal per meter. Thus the latitude correction is

latitude corrected gravity = drift corrected reading in mgals - (dist north of base station in meters) x 0.000786483

(make sure you get the sign right, if a station is SOUTH of the base station it is negative (-) NORTH)

5. Free air correction: a high station has gravity that is too low (farther from center of Earth, so gravity is lower) , so free air correction is added to data

free air corrected reading = latitude corrected reading in milligals + 0.3086 x height (in meters)

 

6. Bouguer correction: corrects for excess mass between the reading elevation and your datum elevation (excess mass makes the gravity reading higher), is subtracted from data

Bouguer corrected reading = free air corrected reading - 0.04193 x 2.67 x height (in meters)

7. Terrain correction: we will skip this. The Bouguer correction assumes a simple slab of density 2.67 g/cm3 between the observation point and the base station. The terrain correction uses the actual topography to better model the mass distribution.

8. Demean: subtract out the average value of the reading, since all we care about is relative variations, and if we subtract the mean out it is easier to see the data when we plot it with the Burger software.

demeaned reading = Bouguer corrected reading - average reading of all stations

 

 

Now you have data that you can interpret! The next step is to model the data using GravModel (next handout). It can deal with uneven station spacing, station spacing of fractions of a kilometer, and allows you to use multiple polygons to fit the data. One bug - the units for the density contrast are g/cm3, not kg/m as stated in the program (check this for yourself). Also, it is a good idea to have some space at both ends of your profile to avoid edge effects.