3:30 PM - 5:00 PM
[SGD01-P05] Scale factor calibration for spring-type relative gravimeters using Japanese latitudinal survey line
Keywords:gravity, scale factor, relative gravimeter, absolute gravimeter, volcano
Since gravity is expressed by the spatial integral of mass distributions, spatiotemporal mass variations in volcanoes can be monitored from campaign gravity measurements using portable spring-type relative gravimeters. The relative gravity value can be expressed as g = f(x), where x indicates the reading value (corresponding to the elongation of the spring) and f indicates the conversion function provided by the gravimeter's manufacturer. However, if f deviates from a true one, a systematic error is contained in the relative gravity value. To obtain the true relative gravity value g', a scale factor (SF; denoted by S) should be multiplied as g' = S * f(x). The SF value can be determined as the ratio of the absolute gravity difference to the relative one, both of which are measured between two gravity points.
SF has been conventionally regarded as a constant value for each relative gravimeter. However, Onizawa (2019) showed that the SF of a Scintrex CG5 relative gravimeter depended on its reading value linearly (S = S(x)). He also found that the SF changed temporally according to the temporal variation in the reading value associated with instrumental drift (S = S(x(t))). Because Onizawa (2019) discussed the characteristics of SF for only one relative gravimeter, SFs and their reading-value dependencies should be investigated for other relative gravimeters.
We therefore calibrated SFs for five LaCoste gravimeters and one Scintrex gravimeter which have often been used at Japanese volcanoes. We first conducted parallel gravity measurements at several gravity points from Hokkaido to Okinawa from 2020 to 2022, using the six relative gravimeters. We then calculated a relative gravity difference between each of two gravity points for each relative gravimeter, by correcting for several effects. The correction of the hydrological gravity disturbance was also applied for absolute gravity data. We finally determined the SF. As a result, we found that the SFs of the six relative gravimeters varied linearly with the reading values at a maximum rate of 1.003 * 10^(-6) /mGal (for the LaCoste G791 gravimeter). This result means that the SF varies by up to 0.0005 within the reading range of 500 mGal, so if the SF contains the error of 0.0005, the systematic error of up to 100 microGal may be contained in the measured relative gravity difference of 200 mGal.
We confirmed that the relative gravity difference can be obtained most precisely when the SF effect is corrected considering the reading value dependence of the SF. We here prepared the following three models. [1] The SF was assumed to be 1.0 within the entire reading range for all gravimeters. [2] The SF was assumed to be constant within the entire reading range, and the average of the observed SF values was used for the constant SF value for each gravimeter. [3] The reading value dependence of SF, as was found in the previous paragraph, was considered for each gravimeter. As a result, the standard deviations of the gravity difference between the foot and crater of Aso Volcano measured from October 2020 to September 2022 were calculated to be [1] 49, [2] 13, and [3] 10 microGal.
We also confirmed that the absolute value can be determined from the relative gravity data most accurately, when the SF effect is corrected by the [3]'s model. We measured the relative gravity difference in October 2022 between the gravity points in Sakurajima and Aso, and corrected for the SF effect using the three models ([1]-[3]). We then calculated the absolute gravity value at Aso from the relative gravity difference, by adding the absolute gravity value at Sakurajima. We found that the absolute gravity value calculated based on the [3]'s model agreed within 1 microGal with that measured by the FG5 gravimeter independently.
SF has been conventionally regarded as a constant value for each relative gravimeter. However, Onizawa (2019) showed that the SF of a Scintrex CG5 relative gravimeter depended on its reading value linearly (S = S(x)). He also found that the SF changed temporally according to the temporal variation in the reading value associated with instrumental drift (S = S(x(t))). Because Onizawa (2019) discussed the characteristics of SF for only one relative gravimeter, SFs and their reading-value dependencies should be investigated for other relative gravimeters.
We therefore calibrated SFs for five LaCoste gravimeters and one Scintrex gravimeter which have often been used at Japanese volcanoes. We first conducted parallel gravity measurements at several gravity points from Hokkaido to Okinawa from 2020 to 2022, using the six relative gravimeters. We then calculated a relative gravity difference between each of two gravity points for each relative gravimeter, by correcting for several effects. The correction of the hydrological gravity disturbance was also applied for absolute gravity data. We finally determined the SF. As a result, we found that the SFs of the six relative gravimeters varied linearly with the reading values at a maximum rate of 1.003 * 10^(-6) /mGal (for the LaCoste G791 gravimeter). This result means that the SF varies by up to 0.0005 within the reading range of 500 mGal, so if the SF contains the error of 0.0005, the systematic error of up to 100 microGal may be contained in the measured relative gravity difference of 200 mGal.
We confirmed that the relative gravity difference can be obtained most precisely when the SF effect is corrected considering the reading value dependence of the SF. We here prepared the following three models. [1] The SF was assumed to be 1.0 within the entire reading range for all gravimeters. [2] The SF was assumed to be constant within the entire reading range, and the average of the observed SF values was used for the constant SF value for each gravimeter. [3] The reading value dependence of SF, as was found in the previous paragraph, was considered for each gravimeter. As a result, the standard deviations of the gravity difference between the foot and crater of Aso Volcano measured from October 2020 to September 2022 were calculated to be [1] 49, [2] 13, and [3] 10 microGal.
We also confirmed that the absolute value can be determined from the relative gravity data most accurately, when the SF effect is corrected by the [3]'s model. We measured the relative gravity difference in October 2022 between the gravity points in Sakurajima and Aso, and corrected for the SF effect using the three models ([1]-[3]). We then calculated the absolute gravity value at Aso from the relative gravity difference, by adding the absolute gravity value at Sakurajima. We found that the absolute gravity value calculated based on the [3]'s model agreed within 1 microGal with that measured by the FG5 gravimeter independently.