Figure 6: Same as Figure 5 but for COCO17
Apart from the observations made from the island GPS station COCO, we report two typical low elevation cases (IISC26 and HYDE29) observed from inland stations, which are not reported so far, to explain the significance of low elevation observation and critical inter-IPP distance (> 5 km) of strong aliasing (Fig. 7 and 8). The ionospheric perturbations coinciding with the tsunami propagating at 70oE were observed by two GPS receivers located around 800 km away from the location of tsunami waves, that is IISC, Bangalore (13.0o N, 77.5oE) (Fig. 7a) and HYDE, Hyderabad (17.4o N, 78.5oE) (Fig. 8a), using signals transmitted by PRN26 and PRN29 respectively. The time series of IISC26 and HYDE29 and the time-distance plots show that the TIPs are detected by all the three methods with the presence of aliasing when the inter-IPP distance is above 5km in the case of dTEC. Further, there is a difference in phase as well as amplitude between rTEC and gROT. But, the difference is only in amplitude between dTEC and gROT (Fig. 7a). The cross correlation of rTEC and gROT ( Table 1) reveals that the signal in rTEC is delayed by 330s in the case of HYDE29 and advanced by 180s in the case of IISC26. These two cases reveal that the TIPs observed by two different satellite-receiver pair at low elevation angle cannot be due to multipath or any other error associated with the observational system or surrounding environment. The coincidence of location and time of TIPs with the simulated tsunami (Fig. 5, 6, 7 and 8), with a delay associated with time required for vertical propagation, further substantiates that the TIPs observed at low elevations are caused by the tsunami. Further, the elevation-distance and azimuth-distance plots (Fig. 7 and 8) show that the combination of low elevation and high azimuth keeps the inter-IPP distance less than 6 km. Hence, the aliasing in dTEC is less compared to COCO07 (Fig. 5) and COCO17 (Fig. 6) though the elevation of the observation is less than 20°. These results reveal that merely discarding the low elevation observation will not remove the aliasing, rather it will lead to missing potential signals in addition to reducing the area of exploration. The above cases illustrates that accounting the inter-IPP distance, which is a function of both elevation and azimuth (Eq. 7), using SPLA while computing the ionospheric perturbations will not only remove the amplitude of aliasing but also removes the necessity of excluding low-elevation observations. Eventually, SPLA helps to increase the area of ionospheric exploration by a single GPS receiver. Hence, SPLA can be a suitable candidate for tsunami early warning.