5. SIMULATED CG LIGHTNING

The majority (> 90%) of lightning discharges in the simulations were intracloud flashes. Currently, a flash is classified as a CG when a leader reaches within 2km of the ground. (This assumption will be tested in the future.) The polarity of the leader determines the polarity of the CG. Once a flash becomes a CG, the upward-propagating leaders are allowed to continue as for an IC flash, but the downward leaders are halted.

 

Figure: 1   Intracloud flash from a supercell simulation. The positive leaders are shown in gray, the negative leaders in black. Distances are in km, and $\Delta x = \Delta y = \Delta z = 500$m. The initiation point indicated by the arrow in each view.

All lightning in the simulations initiated between regions of opposite charge (Fig.  3 b,d), where the electric field is strongest, as has been inferred previously for IC flashes and negative cloud-to-ground flashes. However, positive cloud-to-ground flashes also were initiated between opposite charges in our simulations, a relationship not suggested previously. Positive cloud-to-ground lightning in the simulations occurred only when the lowest significant charge region near the initiation point was negative (i.e. roughly a positive dipole structure about the initiation point).

Positive cloud-to-ground lightning (e.g. Fig.  2 ) occurred predominantly in the supercell storm simulations that used the Gardiner noninductive charging parameterization. The airmass storm simulation with the Gardiner parameterization and strong inductive charging, on the other hand, produced almost exclusively negative CG flashes. As mentioned above, the only environmental difference between the airmass and HP supercell storms was the magnitude of the shear in the 0-5km layer.

 

Figure 2:   As in Fig.  1 but for a positive cloud-to-ground flash from a supercell simulation.

A comparison of the airmass and HP supercell storms with the Gardiner scheme illustrates the differences that led to -CG lightning in the airmass storm and to +CG flashes in the HP supercell. In the airmass case, negatively charged graupel tended to fall through the updraft into high cloud water mixing ratios, where it acquired positive charge via inductive charging. The airmass storm thus developed a strong lower positive charge region which promoted -CG lightning (Fig.  3 a-b). In the HP supercell storm (Fig.  3 c-d), however, the higher wind shear resulted in more graupel falling outside the updraft, so that a more extended volume of negative graupel developed. The horizontally extensive negative charge resulted in IC and +CG lightning with the positive charge region above (Fig.  3d). Intracloud flashes were regularly initiated between the charge layers in the forward flanks of both storms, but only in the supercell storm did some of those flashes connect to ground to become +CG flashes.

 

Figure 3:   Lightning composites (left) and contour slices (right) for simulations with Gardiner noninductive charging. (a-b) Airmass storm with -CG flashes. (c-d) HP supercell with +CG flashes. The composites are surfaces over the regions of positive (blue) and negative (red) lightning leaders during a 2.5 min period. Green surfaces indicate initiation point locations. The slices again show lightning activity regions (red fill for negative leaders, blue for positive, green for initiation locations) overlaid with charge density contours (solid for positive, dashed for negative). Contour values start at $\pm$0.25nCm^-3 with intervals of 0.5. Cloud boundary indicated by lightest gray surface (left) or thick gray line (right).

The model result that +CG flashes initiate only between oppositely charged regions (positive above negative) appears to be consistent with the observations reported by Carey and Rutledge (1998). Carey and Rutledge found that a corona point sensor indicated negative charge overhead for regions of a storm that tended to produce positive CG flashes (i.e. the lowest significant charge region was negative). Other observations (e.g. Brook et al. 1982, Fuquay 1982) also have suggested that storms producing +CG flashes have a positive dipole structure (positive charge over negative). However, the commonly mentioned hypotheses to explain the occurrence of +CG flashes all seem to assume that the lowest charge region above the +CG strike point should be positive. The oft-mentioned tilted dipole (or sheared dipole) hypothesis suggests that positive CG flashes might occur if an upper positive charge layer is shifted away from the lower negative charge and becomes ``exposed to ground.'' Likewise, the ``inverted dipole'' and ``enhanced lower positive charge'' hypotheses both assume that a lower positive charge causes +CG lightning. The present results suggest that negative charge is needed below positive charge to initiate +CG flashes. However, a flash triggered from ground on a mountain peak or tall structure would presumably access the lowest significant charge (either positive or negative), but this capability is not included in the model.