D region effects of the 17–19 March 2015, a St Patrick's Day super geomagnetic storm (Dst = −223 nT), using a navigational transmitter very low frequency (VLF) signal (NWC, 19.8 kHz) recorded at a low‐latitude Indian station, Allahabad (geomag. lat., 16.45°N), have been analyzed and compared with similar strength of the 22–25 June 2015 storm (Dst = −204 nT). During the March storm, NWC signal amplitude decreased on 17 March (main phase of the storm) and recovered on 27 March, which is 1 day after the recovery of the storm, whereas for the June storm, VLF amplitude decreased for 2 days only during its recovery phase. The decrease in the amplitude was pronounced during evening terminator for both the storms. The modeling of VLF signal anomaly on 17 March and on 25 June using Long‐Wave Propagation Capability code shows an increase in the D region reference height (h′) by ~2.6 km and ~2.5 km, for March and June storms, respectively. The D region electron density (N_e) determined using storm time h′ and sharpness factor β shows a decrease in the N_e during the main phase followed by a slow recovery during the recovery phase of the March storm, whereas June 2015 storm showed a decrease in the N_e only on 25 and 26 June. Morlet Wavelet analysis of the amplitude for both the storms shows a presence of strong wave‐like signatures, suggesting propagation of atmospheric gravity waves/traveling ionosphere disturbances to the low latitude D region due to the Joule heating at high latitudes.

The response of the D region low‐latitude ionosphere has been examined for extreme space weather event of 14–16 December 2006 associated with a X1.5 solar flare and an intense geomagnetic storm (Dst = −146 nT) using VLF signals from Northwest Cape, Australia (NWC) (19.8 kHz) and Lualualei, Hawaii (callsign NPM) (21.4 kHz) transmitters monitored at Suva (Geographic Coordinates, 18.10°S, 178.40°E), Fiji. Modeling of flare associated amplitude and phase enhancements of NWC (3.6 dB, 223°) and NPM (5 dB, 153°) using Long‐Wave Propagation Capability code shows reduction in the D region reflection height (H′) by 11.1 km and 9.4 km, and enhancement in ionization gradients described by increases in the exponential sharpness factor (β) by 0.122 and 0.126 km⁻¹, for the NWC and NPM paths, respectively. During the storm the daytime signal strengths of the NWC and NPM signals were reduced by 3.2 dB on 15 and 16 December (for about 46 h) and recovered by 17 December. Modeling for the NWC path shows that storm time values of H′ and β were reduced by 1.2 km and 0.06 km⁻¹, respectively. Morlet wavelet analysis of signal amplitudes shows no clearly strong signatures of gravity wave propagation to low latitudes during the main and recovery phases. The reduction in VLF signal strength is due to increased signal attenuation and absorption by the Earth‐ionosphere waveguide due to storm‐induced D region ionization changes and hence changes in D region parameters. The long duration of the storm effect results from the slow diffusion of changed composition/ionization at D region altitudes compared with higher altitudes in the ionosphere.

The changes in critical frequency of the F₂ layer (f_oF₂) and f_oF₂ deviation (Δf_oF₂) have been determined for three geomagnetic storms in March of the years 2012, 2013, and 2015 at low‐latitude stations, Darwin (geomag. lat. 21.96°S) and Townsville (28.95°S), and midlatitude stations, Brisbane (36.73°S), Canberra (45.65°S), and Hobart (54.17°S). The moderate storm during 15–16 March 2012 (Dst = −87 nT) showed a decrease in f_oF₂ at midlatitude and no effect at low‐latitude stations. For the intense storm of 17–18 March 2013 (Dst = −132 nT) and the super storm of 17–18 March 2015 (Dst = −222 nT), some middle‐ to low‐latitude stations showed a short‐duration increase in f_oF₂, but all stations showed a long‐duration decrease in f_oF₂ during the recovery phases with Δf_oF₂% varying from 26% (Darwin) to 36.6% at Hobart for the March 2013 storm and above 40% for the March 2015 storm at all of the stations. Short‐duration (~2–4 hr) increase in f_oF₂ seems to be associated with the prompt penetrating electric fields. Long‐duration (>6 hr) decrease in f_oF₂ is mainly accounted to the decrease in thermospheric O/N₂ density ratio because of storm‐induced high‐latitude circulation of gas with depleted O/N₂ density ratio to lower latitudes and partly due to disturbance dynamo electric fields. A comparison of ionosonde given f_oF₂ for equinoctial storms (March 2013 and 2015) with similar strength Southern Hemisphere winter storms (July 2012 and June 2015) has been made with the IRI‐2016 model f_oF₂ for Darwin, Brisbane, and Canberra stations.

Very low frequency (VLF) electromagnetic signals from navigational transmitters propagate worldwide in the Earth‐ionosphere waveguide formed by the Earth and the electrically conducting lower ionosphere. Changes in the signal properties are signatures of variations in the conductivity of the reflecting boundary of the lower ionosphere which is located in the mesosphere and lower thermosphere, and their analysis is, therefore, a way to study processes in these remote regions. Here we present a study on amplitude perturbations of local origin on the VLF transmitter signals (NPM, NLK, NAA, and JJI) observed during tropical cyclone (TC) Evan, 9–16 December 2012 when TC was in the proximity of the transmitter‐receiver links. We observed a maximum amplitude perturbation of 5.7 dB on JJI transmitter during 16 December event. From Long Wave Propagation Capability model applied to three selected events we estimate a maximum decrease in the nighttime D region reference height (H′) by ~5.2 km (13 December, NPM) and maximum increase in the daytime D region H′ by 6.1 km and 7.5 km (14 and 16 December, JJI). The results suggest that the TC caused the neutral densities of the mesosphere and lower thermosphere to lift and sink (bringing the lower ionosphere with it), an effect that may be mediated by gravity waves generated by the TC. The perturbations were observed before the storm was classified as a TC, at a time when it was a tropical depression, suggesting the broader conclusion that severe convective storms, in general, perturb the mesosphere and the stratosphere through which the perturbations propagate.

Lightning strokes are known to cause direct heating and ionization of the D region, some of which are detected via scattering of VLF transmitter signals and are known as Early VLF events. The disturbed ionosphere typically recovers in many tens of seconds. New experimental evidence is presented demonstrating that the scattering pattern and onset amplitude of Early VLF events are strongly related to both the magnitude and polarity of causative lightning peak current. Observations of Early VLF events at nine Stanford VLF receiver sites across the continental United States are combined with lightning geolocation data from the National Lightning Detection Network (NLDN). During January and March 2011, NLDN recorded 7769 intense lightning discharges with high peak currents (>100 kA) generating 1250 detected Early VLF events. We show that the size of the scattered field due to the ionospheric disturbance increases with the peak current intensity of the causative lightning discharge. The most intense peak currents of >+200 and <−250 kA disturb VLF transmitter signals as far as ∼400 km away from the lightning stroke. Early VLF event detection probability also increases rapidly with peak current intensity. On the other hand, the observed VLF amplitude change is not significantly dependent on the peak current intensity. Stroke polarity is also important, with positive strokes being ∼5 times more likely to generate Early VLF disturbances than negative strokes of the same intensity. Intense positive cloud‐to‐ground lightning discharges, especially when occurring over the sea, are also more likely to produce Early VLF events with long recovery (many minutes).

We estimate D region changes due to 22 July 2009 total solar eclipse (SE), 13–14 November 2012 total SE, and 9–10 May 2013 annular SE, using VLF navigational transmitters signal observations at Suva, Fiji. The North West Cape (NWC) signal (19.8 kHz) showed an amplitude and phase decrease of 0.70 dB and 23° during November SE and 2.0 dB and 90° during May SE. The modeling using Long Wave Propagation Capability code for NWC‐Suva path during November and May SEs showed an increase in average D region reflection height (H′) and sharpness factor (β) by 0.6 and 0.5 km and 0.012 and 0.015 km⁻¹, respectively. The July total SE for JJI‐Suva path showed an increase in H′ of 1.5 km and a decrease in β of 0.055 km⁻¹. The decrease in the electron density calculated using SE time H′ and β is maximum for July total SE and minimum for May annular SE. The effective recombination coefficient estimated from the decay and recovery of signal phase associated with May annular SE was higher (27%) than normal daytime value 5.0 × 10⁻⁷ cm⁻³ s⁻¹ and varied between 1.47 × 10⁻⁶ and 1.15 × 10⁻⁷ cm⁻³ s⁻¹ in the altitude 70 to 80 km. Morlet wavelet analysis of signals amplitude shows strong wave‐like signatures (WLS) associated with three SEs with period ranging 24–66 min, but the intensity and duration of WLS show no clear dependence on SE magnitude and type. Apart from the cooling spot, the eclipse shadow can also generate WLS associated with atmospheric gravity waves.