“Raw data denoising and assessment of records from Misallat and Abu Simbel geomagnetic observatories, Egypt”

Abstract

The prompt access to accurate ground-based geomagnetic data is becoming a more crucial requirement. Misallat (29° 30′ 52″ N, 30° 53′ 22″ E) and Abu Simbel (22° 29′ 22″ N, 31° 32′ 41″ E) magnetic observatories of Egypt are filling a wide regional gap in the global distribution of geomagnetic observatories. However, both observatories suffer from continuous sources of noise that lead to a minimization of their contribution to geomagnetism. The industrial zone in the north of Misallat observatory resulted in convoluting the observatory records with high-frequency noise that appear, due to aliasing, as low-frequency signals imposed on the natural events of the magnetosphere. On the other side, spikes are frequently occurring in the raw data at Abu Simbel observatory. The goal of this research is to provide a variety of selected filtering approaches for denoising data in both Egyptian geomagnetic observatories and to evaluate their reliability after processing. The applied methods comprise frequency and time domain filters, in addition to the wavelet transform denoising procedure. We applied all of the methods to a selected 1-day sample of typical data from both observatories. The noise is efficiently reduced. The filtered data from the different denoising techniques were ranked according to the root mean square deviation (RMSD) and signal-to-noise ratio (SNR) to evaluate the quality of the filtered data. Furthermore, one-day data samples of a quiet day and a stormy day are processed and then compared with simultaneous INTERMAGNET records from Tamanrasset, Algeria and Pedeli, Greece. So, we are currently able to process the raw data from both magnetic observatories while keeping them as useful as possible for all research and application aspects.