This paper presents an original investigation on a time-resolved moiré deflectometry for gas temperature measurements of long sparks in air. In order to perform a rigorous comparative investigation with the widely used quantitative schlieren method, we set up an orthogonal optical measurement system. The impacts of spatial resolution and exposure time on the measurement accuracy are investigated by comparing the measured radial distribution of the gas temperature and its time evolution during the same spark discharge event. It is found that the time-resolved moiré deflectometry with an exposure time of 1.0 μs can accurately measure the gas temperature evolution during the isobaric heating and relaxation process of long sparks. The measured results of the expansion rate of cross-section π·〖R_"g" 〗^2 and average temperature Tg confirms that the energy loss is dominated by thermal conduction during isobaric heating and relaxation process after the extinction of discharge current. The moiré deflectometry can achieve a finer spatial resolution than the quantitative schlieren method, which is beneficial to the reconstruction of steep radial temperature distribution and the suppression of measured temperature fluctuation. Moreover, the moiré deflectometry can obtain a sub-microsecond exposure time by increasing the power of continuous-wave laser source, which shows its potential to capture the rapid changes in gas temperature during the fast-heating process.

Leader propagation velocity is one of the most important properties to characterize both intracloud (IC), cloud-toground (CG) and artificially triggered lightning discharges. Increasing observation results implicate that the lightning leader speed may decrease with the increase of altitude of leader propagation. The positive leader in rocket triggered lightning observed in New Mexico with an altitude of 3,200 m developed more slowly than those in low altitude regions (Edens et al., 2012). The vertical velocity of initial negative leaders in the preliminary breakdown of IC flashes shows a decreasing trend with the increase of initiation altitude (Wu et al., 2015). Recently, similar trend is also found in the propagation of positive leader in IC flashes, where the velocity of positive leader decreases with the increase of initiation altitude (Wu et al., 2019). Nevertheless, the effect of reduced air density on lightning leader velocity is difficult to recognize in numerous observations of CG lightning flashes, because the lightning leader velocity can be affected by multiple factors, such as the ambient electric field, injected current, humidity, etc. For instance, although a clear increasing trend in speed was observed for positive leaders as they approach to ground in +CG flashes (Kong et al., 2008;Saba et al., 2008;Wang & Takagi, 2011), the increase of the electric field between the downward leader and the ground may be more crucial to the acceleration of lightning leaders. Due to difficulties in decoupling of the influence of multiple factors on the leader velocity in field observations, the effect of reduced air density at high altitude on the lightning leader velocity has not been fully understood yet.During the past two decades, numerical simulation was developed to investigate the formation of leader discharge at reduced air density conditions (

In this study, we present a comprehensive investigation of positive leader discharges, with the aim of enhancing our understanding of leader decay and reactivation. Our approach involved a detailed experimental and computational analysis of the phenomena. Specifically, we employed a time-resolved quantitative Schlieren platform, which provided us with a high spatial resolution (60.0 μm/pixel) and short exposure times (0.37 μs/frame), allowing us to capture the 2D spatial-temporal evolution of gas temperature in positive leaders with a gap length of up to three meters. In addition, we employed a detailed thermal-hydrodynamic model coupled with a comprehensive kinetic scheme, consisting of 28 chemical species and 125 chemical reactions. Our simulations showed good agreement with the measured mean gas temperature and expansion rate of thermal radius. We conducted experiments under the same applied conditions to obtain both stable and decaying leaders. Our results showed that once a positive leader starts to decay, the temperature drops below 3000 K. At the same time, both the electric field and conductivity decreased significantly compared to a stable leader. In addition, before the temperature drops below 2000 K and transforms into an aborted leader, a decaying leader might be reactivated.