Sensor Response Estimate and Cross Calibration of Paleomagnetic Measurements on Pass‐Through Superconducting Rock Magnetometers
Sensor Response Estimate and Cross Calibration of Paleomagnetic Measurements on Pass‐Through Superconducting Rock Magnetometers
复制标题
穿过式超导岩石磁强计古地磁测量的传感器响应估计和交叉校准
DOI：
10.1029/2019gc008597
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Oda Hirokuni
中科院分区：
文献类型：
--
作者：
Xuan Chuang;Oda Hirokuni
直通式超导岩石磁力仪（SRM）能够对连续样品进行快速、精确的剩磁测量，对于古地磁研究至关重要。由于SRM传感器响应的卷积效应，需要对直通测量值进行去卷积，以恢复准确和高分辨率的信号。成功反卷积的关键步骤是SRM传感器响应的可靠估计。在这里，我们提出了新的工具URESPONSE，用于基于校准良好的磁性点源的测量来精确估计SRM传感器响应。URESONSE允许对具有不同横截面几何形状的连续样品进行传感器响应估计。我们估计了南安普顿大学的一个旧的液氦冷却SRM（SRM-old）和一个新的无液氦SRM（SRM-new）的传感器响应，并比较了去卷积前后两个SRM上u通道的剩磁测量。对于每个SRM，基于使用不同的磁点源样本和/或测量过程收集的数据的传感器响应估计通常产生小的差异（标准偏差）。<~1%），而具有不同横截面几何形状的连续样品的传感器响应估计值通常显示出较大的差异（标准差）。高达2%）。与SRM‐old相比，SRM‐new具有更小的交叉轴响应、更少的负区域和更宽的主轴响应。我们证明，使用根据传感器响应估计计算的九元素“有效长度”矩阵对数据进行标准化对于最大限度地减少两个SRM上测量值的差异是必要的。使用准确的传感器响应估计对两个SRM上的测量进行反卷积可产生高度一致和高分辨率的结果，而使用不准确的传感器响应数据进行反卷积可导致显著差异，特别是对于具有大交叉轴响应的SRM旧数据。
Pass‐through superconducting rock magnetometers (SRMs) enable rapid and precise remanence measurement of continuous samples and are essential for paleomagnetic studies. Due to convolution effect of the SRM sensor response, pass‐through measurements need to be deconvolved to restore accurate and high‐resolution signal. A key step toward successful deconvolution is a reliable estimate of the SRM sensor response. Here, we present new tool URESPONSE for accurate SRM sensor response estimate based on measurements of a well‐calibrated magnetic point source. URESONSE allows sensor response to be estimated for continuous samples with different cross‐section geometry. We estimate sensor responses for an old liquid helium‐cooled SRM (SRM‐old) and a new liquid helium‐free SRM (SRM‐new) at the University of Southampton and compare remanence measurement of a u‐channel on both SRMs before and after deconvolution. For each SRM, sensor response estimates based on data collected using different magnetic point source samples and/or measurement procedures generally yield small differences (std. <~1%), while sensor response estimates for continuous samples with different cross‐section geometry often show larger differences (std. up to ~2%). Compared with SRM‐old, SRM‐new has smaller cross‐axis responses, less negative zones, and significantly broader main axis responses. We demonstrate that normalization of data using a nine‐element “effective length” matrix calculated from sensor response estimate is necessary to minimize differences in measurements on two SRMs. Deconvolution of measurements on two SRMs using accurate sensor response estimates yields highly consistent and high‐resolution results, while deconvolution using inaccurate sensor response data can lead to significant differences especially for data from SRM‐old that has large cross‐axis responses.