6.5 Detecting aliasing and artifact free CIPs
Time-distance plots given in Fig. 10 show that the aliasing is prominent in CIPs obtained using the differential method (dTEC). There are significant differences in amplitude of gROT and dTEC in 61% of the cases when the inter-IPP distances are more than 4 km (Fig. 10), which indicate that unaccounted inter-IPP distances in differential method aliased as ionospheric perturbations in dTEC.
The scenario is completely different when using the residual method (rTEC). The rTEC values show strong signals in many cases at different times compared to dTEC and gROT (Fig. 10). Upon comparing these values with the corresponding time series (Fig. 11), one can easily identify that most of the signals are artifacts associated with data gaps. Other than artifacts, notable CIPs occur prior to the earthquake, for example PYUT16 (Fig. 10). The corresponding time series given in Fig. 11 reveals that these are multiple artifacts occurred in a short time which caused this signal aliasing.
Apart from the artifacts and aliasing, time series of receiver-PRN pairs show that residual method (rTEC) completely failed to detect CIPs in 50% of the cases compared to the differential method and SPLA (Fig. 11 and Table 2). Upon exploring, we found that the poor fits of high order polynomials used to represent the trend of vTEC time series lead to failure in detecting the CIPs (Fig. 12). For example, in the case of TPLJ03 (Fig. 12) it is shown that the polynomial fits well with vTEC only from 4 UT to 5:45 UT. However, the CIP detected by dTEC (Fig. 12c) and gROT (Fig. 12d) at 6:15UT falls under the misfit region in the case of rTEC (Fig. 12b) which leads to failure in detecting the CIP by residual method. Further inspections revealed that the filtering process of obtaining CIPs from rTEC filtered out the high residual appear from 6:15UT to 8:45UT (Fig. 12b) and this lead to the failure of detecting the sharply varying CIP with a frequency range of 1.6 to 16 mHz by the residual method. However, such sharp variations were successfully detected by SPLA.