Coordinated analysis of delayed sprites with high-speed images and remote electromagnetic fields

One of the most dramatic discoveries in solar-terrestrial physics in the past two decades is the sprite, a high altitude optical glow produced by a lightning discharge. Previous sprite studies including both theoretical modeling and remote measurements of optical emissions and associated radio emissions have revealed many important features. However, in-situ measurements, which are critical for understanding the microphysics in sprites and constraining the existing models, are almost impossible because of the sprites' small time scale (a few ms) and large spatial scale (tens of km). In this work, we infer the lightning-driven ambient electric fields by combining remote measured electromagnetic fields with numerical simulations. To accomplish this, we first extract the lightning source current from remotely measured magnetic fields with a deconvolution technique. Then we apply this current source to an existing 2-D Finite Difference Time Domain (FDTD) model to compute the electric fields at sprite altitudes. These inferred electric fields make up for the deficiency of lacking in-situ measurements. A data set collected at two observation sites in 2005 combines simultaneous measurements of sprite optical emissions and sprite-producing lightning radiated electromagnetic fields. Sprite images from a high speed camera and the measured wideband magnetic fields removed the limitations imposed by the small sprite temporal scale and allow us to precisely determine the sprite initiation time and the time delay from its parent lightning discharge. For 83 sprites analyzed, close to 50% of them are delayed for more than 10 ms after the lightning discharges and empirically defined as long-delayed sprites. Compared with short-delayed sprites, which are driven by the lightning return stroke, all these long-delayed sprites are associated with intense continuing current and large total charge moment changes. Besides that, sferic bursts and slow intensifications are frequently detected before those long-delayed sprites. These observations suggest a different initiation mechanism of long-delayed sprites. To reveal that, we inferred the lightning-driven electric fields at the sprite initiation time and altitude. Our results show that although long-delayed sprites are mainly driven by the continuing current instead of the lightning return stroke, the electric fields required to produce those long-delayed sprites are essentially the same as fields to produce short-delayed sprites. Thus the initiation mechanism of long delayed sprite is consistent with the conventional breakdown model. Our results also revealed that the slow (5{20ms) intensifications in continuing current can significantly increase high altitude electric fields and play a major role in initiating delayed sprite. Sferic bursts, which were suggested as a direct cause of long-delayed sprites in previous studies, are linked to slow intensifications but not causal. Previous studies from remote measured low frequency radio emissions indicate that substantial electric current flows inside the sprite body. This charge motion, with unknown location and amount, is related to the detailed internal microphysics of sprite development that is in turn connected to the impact sprites have on the mesosphere. In our data, the recorded high speed images show the entire development history of sprite streamers. By assuming streamers propagate along the direction of local electric fields, we estimate the amount of electric charge in sprites. Our results show that individual bright core contains significant negative space charge between 0.01 to 0.03 C. Numerical simulations also indicate that this sprite core region is at least partial or perhaps the dominant source of the positive charge in the downward positive polarity streamers. Thus the average amount of charge in each downward streamer is at least 2 - 4 10􀀀3 C. The connection between these charge regions is consistent with previous observations. The reported amount and location of the electric charge provide the initial condition and key data to constrain the existing streamer models. After initiation, sprite streamers propagate in the inhomogeneous medium from a strong field region to a weak field region. The propagation properties reflect the physics in sprite development. For the first time we measured the downward streamer propagation behaviors over the full sprite altitude extent. We found that downward streamers accelerate to a maximum velocity of 1 - 3 x 107 m/s and then immediately decelerate at an almost constant rate close to 10 10 m/s2. The deceleration processes dominant downward streamer propagation in both time and distance. Lightning driven electric fields have been inferred at streamer tip locations during their propagation. We found that most of the deceleration process occurs at a electric field less than 0.1 Ek. The results also show the dependence of sprite termination altitude on the ambient electric field. A minimum ambient electric field about 0.05 Ek is consistently observed for streamers in different sprites or at different locations in a single sprite. These streamer propagation properties as well as their connections to the ambient electric fields can be applied to further constrain the streamer models.