Morphological Development and Spectral Features

[7] Between July 2004 and April 2009 ISUAL imager registered 32 GJs. Among them, 20 GJ events taken through Table 1. Band Passes of the ISUAL Imager and Spectrophotometer

and the Major Emissions in the Passing Bands

Filter Wavelength Band (nm) Emission Band System Imager

6 no filter full wavelength observation

Spectrophotometer

SP1 150–280 N2Lyman‐Birge‐Hopfield (LBH)

SP2 337 (bandwidth: 5.6) N22P(0,0) SP3 391.4 (bandwidth: 4.2) N2+

1N(0,0)

SP4 624–750 N21P

SP5 777.4 (773.6–783.4) OI(1) in lightning

SP6 244–392 mid‐UV band

CHOU ET AL.: NEGATIVE AND POSITIVE GIGANTIC JETS A00E45 A00E45

2 of 13

the N21P filter that were not contaminated by lightning or other types of TLEs are selected for analysis. We classify these 20 GJs into three types according to their morphological development and spectral properties.

3.1. Type I GJs (Eight Events)

[8] The spatial‐temporal evolution of this type of GJs is similar to those observed in ground observations. This type of ISUAL GJs triggered on the fully developed jet, as shown in Figure 1a. Hence only the fully developed jet and the trailing jet stage were recorded, and the leading jet was not observed.

In the fully developed jet stage, both SP2 and SP6 showed double photometric peaks which separated from each other by∼1.5 ms. The AP data provided auxiliary information that helps to understand the SP reading. From the AP blue module data (Figure 1b), the double photometric peaks in SP2 and SP6 were identified as peaks associated with upward dis-charge and the ensuing downward return‐stroke‐like process [Kuo et al., 2009]. The AP red module registered photometric traces with very low S/N ratio for the GJ emissions from

∼45 km altitude or lower, thus only the first upward propa-gating peak was unambiguously resolved. The first photo-metric peak was seen by all the SP channels except SP5. SP2, SP3, SP4, and SP6 data also show slowly varying continuous emissions following the distinct peaks that are associated with continuing current flow from the cloud to the iono-sphere [Kuo et al., 2009; Cummer et al., 2009]. The duration of the continuous current varies from event to event and differs for different event distances. From the AP data, one can clearly discern that the continuous emissions recorded by the SP are emitted by electric current flowing inside the thundercloud [Kuo et al., 2009].

3.2. Type II GJs (Seven Events)

[9] Image sequence in Figure 2a suggests that this type II GJ starts with a jet‐like event (frames 2–4) and then slowly propagates upward and finally develops into a fully devel-oped jet (frame 5). Figure 2a indicates that signals in SP2, SP3, and SP6 are only clearly discernible within ∼1 ms around the trigger time with sharp rise and slow decay fea-tures. The AP blue module registers a jet‐like signal that cross

∼2–3 channels, while the red module signals always have low S/N ratios. All the initiating jets share the same photometric features as those shown in Figure 3 for a typical ISUAL blue jet; the image sequence was also taken through the same N21P filter. Therefore it can be concluded with high confidence that the type II GJ starts with a slow upward propagating blue jet and later develops into a fully developed jet. This process typically takes ∼110 ms. The fully developed jet appeared in the fifth image frame, and the last image frame seems to contain a trailing‐jet‐like luminous column that radiated continuously and rose up from the cloud to∼30 km altitude.

However, the morphology of this trailing‐jet‐like column is

very different from that for the type I GJs, and there is no detectable radiation at the cloud deck level. Hence whether the trailing jet feature exists in the type II GJs is an unsettled issue. The corresponding SP emissions for the fully devel-oped jet of the type II GJs are full of noises and cannot be properly analyzed. The brightness of the fully developed jet (frame 5) is distinctly lower than that of the type I GJs. We also found that blue starters and blue jets often occurred in the same general region before and after the type II GJs; another distinct feature that is not shared by the type I GJs.

3.3. Type III GJs (Five Events)

[10] Type III GJs are preceded by bright lightning; after that a GJ occurs near the preceding lightning and extends from the cloud top toward the ionosphere, as shown in Figure 4. The interval between preceding lightning and type III GJ varies widely from∼15 ms to 110 ms. Morphologically, type III GJs can be falsely identified as sprites. However, sprites [Sentman et al., 1994; Pasko, 2007, and references therein] extend from

∼40 km to 90 km altitudes but do not connect to the cloud top, unlike the type III GJs. Furthermore, sprites always produce clearly recognizable spectral signals in the ISUAL SP [Kuo et al., 2005; Figures 3 and 4] that characteristically are very different from those from lightning/type III GJs. Also, the average brightness of the ISUAL carrot sprites is found to be

∼3 MR [Kuo et al., 2008], which is nearly 10 times higher than that for the type III GJs. Hence the type III GJs are unambiguously distinct from the carrot sprites.

[11] After careful examining the image frames 2–4 in Figure 4a, a small luminous column is seen to protrude above the cloud emissions in each frame. During the interval, weak blue emissions are represented in SP2 (337 nm) and SP6 (224–392 nm). This suggests that the small luminous column might be a blue jet‐like event. Figure 4b shows that a long luminous column that bridges the cloud top and the iono-sphere seems to develop along the discharge channel estab-lished by the preceding luminous events, while it emits no recognizable signals in SP and AP. Therefore this clearly is a type III GJ, not a sprite, since both its generating sequence and its spectral features also differ from those for the type I and the type II GJs.

[12] The main characteristics for the three types of gigantic jets are summarized in Table 2. From the spectral data recorded by ISUAL spectrophotometer and array photometer, it is clear that the forms and the spectral properties of type I and type II GJs are very different. The spectral signals from the type III GJs are masked by the emissions of the preceding lightning, so it is extremely difficult to compare them to those of the type I and type II GJs.

[13] The spatial‐temporal evolution of the type I GJs is similar to that of the ground‐observed GJs, which are known to emit sferics that signifies they are−CI events [Su et al., 2003; Cummer et al., 2009]. Hence it is nature to expect

Figure 1. Type I gigantic jets (GJ) on 28 February 2006 0435:52.993 UTC. (a) The image sequence and the spectrophotom-eter data for this event. (b) The first image frame and the associated array photomspectrophotom-eter (AP) data. The fully developed jet (FDJ) occurred in first image, and the corresponding spectrophotometer (SP)2 and SP6 data contain double peaks and a humping continuous luminosity. The trailing jet occurred in frames 2–6. The AP blue module shows signals associated with the upward propagation FDJ and the ensuing downward return stroke‐like process. For the red module, only the signal from the FDJ was registered. The continuous cloud emissions of this event manifest themselves as the humping curves in SP2, SP3, SP6, and AP channels 10 and 11.

CHOU ET AL.: NEGATIVE AND POSITIVE GIGANTIC JETS A00E45 A00E45

3 of 13

Figure 1

CHOU ET AL.: NEGATIVE AND POSITIVE GIGANTIC JETS A00E45 A00E45

4 of 13

that the type I GJs are also consist of negative streamers.

Previous observations and theoretical studies [Wescott et al., 1996, 1998; Pasko et al., 1996; Pasko and George, 2002]

have concluded that blue jets are comprised of positive streamers extending upward from the cloud top. The type II GJ begins as a blue jet and then develops into a GJ; hence its polarity should be the same as the preceding blue jet. The polarity of the type IIIs likely is stipulated by the preceding lightning that creates the charge imbalance to initiate this type

of gigantic jets. To verify the above conjectures, magnetic field and photometric signatures of the three types of GJs will be analyzed and compared in the following sections.