Processing of VLF amplitude measurements: Deduction of a quiet time seasonal variation

The amplitude of Very Low Frequency (VLF) transmissions propagating from transmitter to receiver between the Earth's surface and the ionospheric D-region is a useful measurement to detect changes in the ionization within the D-region ranging from 60 to 90 km. The VLF signal amplitude is disturbed by geomagnetic, solar, and atmospheric phenomena. To be able to identify perturbations in the VLF signal amplitude, we determine its averaged seasonal variation under quiet solar and geomagnetic conditions. Here it is challenging, that long time series of the VLF signal amplitude show significant jumps and outliers, which are caused artificially by technical adjustments/maintenance work. This paper presents a new approach for processing long VLF data time series over multiple years resulting in level 2 data. The new level 2 data enables the consideration of time series with artificial jumps since the jumps are leveled. Moreover, the outliers are removed by a robust and systematic 2-step outlier filtering. The average seasonal and diurnal variation for different transmitter-receiver combinations can be computed with the new level 2 data by applying a composite analysis. A subsequently applied polynomial fit obtains the quiet time lines for daytime and nighttime, representing the typical seasonal variation under undisturbed conditions of the VLF signal amplitude for each considered link. The developed quiet time lines may serve as a tool to determine perturbations of the VLF signal amplitude with solar and geomagnetic as well as atmospheric origin. Also, they allow comparison of the VLF signal amplitude variation for different transmitter-receiver links.