The standard processing of airborne gamma-ray spectrometry (AGRS) data provides useful preliminary information to interpretation in several contexts, such as environmental studies, geological mapping, and analysis of mineral deposits. For optimal results, the acquisition conditions should be nearly constant, and the flight height should be uniform. However, abrupt changes in flight height (often originated from mountainous terrain) are common and lead to spurious variations and incorrect anomaly interpretation. Moreover, the commonly used corrections applied to radiometric data do not consider the effective sampling area of a survey, especially the overlap between successive fields of view. As a consequence, the concentrations estimates of potassium (K), uranium (eU), and thorium (eTh) are not sharp. A solution to deal with this problem is to compute the concentration from the AGRS data through an inversion algorithm. Inversion of AGRS data has shown to be an effective approach to suppress the significant overlap between successive fields of view. We introduce a logarithmic barrier approach for the radiometric inversion to avoid spurious negative values in the radioelement concentration models. We choose calibration ranges from Brazil and Canada to test the proposed methodology. We produced concentration models for K, eU and eTh and compared the standard approach model with the recovered model from both ranges. The predicted data from inversion of both calibration ranges are mostly consistent with the survey data, not over smoothing the data or fitting noise. The predicted data from inversion showed more consistency with the observed data than the one predicted from the standard procedure.

机载伽玛射线能谱（AGRS）数据的标准处理为一些环境下的解释提供了有用的初步信息，例如环境研究，地质制图和矿床分析。为了获得最佳结果，采集条件应几乎恒定，并且飞行高度应均匀。但是，飞行高度的突然变化（通常起源于山区）是常见的，并且会导致虚假变化和错误的异常解释。此外，应用于辐射数据的常用校正方法并未考虑勘测的有效采样区域，尤其是连续视场之间的重叠。因此，钾（K），铀（eU）和th（eTh）的浓度估算值并不精确。解决此问题的方法是通过反演算法从AGRS数据中计算浓度。AGRS数据反演已被证明是抑制连续视场之间明显重叠的有效方法。我们为辐射反演引入对数势垒方法，以避免在放射性元素浓度模型中出现虚假负值。我们从巴西和加拿大选择校准范围，以测试建议的方法。我们制作了K，eU和eTh的浓度模型，并将标准方法模型与两个范围内的回收模型进行了比较。来自两个校准范围的反转的预测数据大部分与测量数据一致，而不是过度平滑数据或拟合噪声。