We compare simulations and experiments of single positive streamer discharges in air at 100 mbar, aiming toward model validation. Experimentally, streamers are generated in a plate-plate geometry with a protruding needle. We are able to capture the complete time evolution of reproducible single-filament streamers with a ns gate-time camera. A 2D axisymmetric drift-diffusion-reaction fluid model is used to simulate streamers under conditions closely matching those of the experiments. Streamer velocities, radii and light emission profiles are compared between model and experiment. Good qualitative agreement is observed between the experimental and simulated optical emission profiles, and for the streamer velocity and radius during the entire evolution. Quantitatively, the simulated streamer velocity is about 20% to 30% lower at the same streamer length, and the simulated radius is about 1 mm (20% to 30%) smaller. The effect of various parameters on the agreement between model and experiment is studied, such as the used transport data, the background ionization level, the photoionization rate, the gas temperature, the voltage rise time and the voltage boundary conditions. An increase in gas temperature due to the 50 Hz experimental repetition frequency could probably account for some of the observed discrepancies.

Streamer discharges are the primary mode of electric breakdown of air in lightning and high voltage technology. Streamer channels branch many times, which determines the developing tree-like discharge structure. Understanding these branched structures is for example important to describe streamer coronas in lightning research. We simulate branching of positive streamers in air using a 3D fluid model where photoionization is included as a discrete and stochastic process. The probability and morphology of branching are in good agreement with dedicated experiments. This demonstrates that photoionization indeed provides the noise that triggers branching, and we show that branching is remarkably sensitive to the amount of photoionization. Our comparison is therefore one of the first sensitive tests for Zheleznyak's photoionization model, confirming its validity.

Streamer discharges often exhibit branching, which can greatly affect their behaviour and will lead to so-called streamer trees. In this work we present a methodology for investigating the structure of a streamer discharge tree by means of advanced imaging techniques.
Stereoscopic and stroboscopic techniques augment the images with depth perception and temporal information relevant to study the inherently stochastic three-dimensional and transient streamers.
A semi-automated post processing algorithm is developed to make a reconstruction of the streamer discharge tree formation.
This results in a tree of streamer segments, separated by branching events, where velocities, diameters and trajectories are used to characterize the morphology.
The workings of the algorithm is detailed using an exemplar measurement series of positive streamers in synthetic air at 233 mbar.

Streamer discharges are the primary mode of electric breakdown of air in lightning and high voltage technology. Streamer channels branch many times, which determines the developing tree-like discharge structure. We simulate branching of positive streamers in air using a 3D fluid model with stochastic photoionization. The distributions of branching angles and branching locations agree quantitatively with dedicated experiments. The simulations are sensitive to the photoionization coefficients, and they confirm the validity of the classical photoionization model.

The stochastic nature of streamers and the manual identification of features in 2D discharge images together cause great ambiguities when analysing streamer branching characteristics. Here we present the development of streamer image diagnostics by a 2D peak-finding method to obtain accurately quantified extensive statistics on streamer branching. And we present quantitative results on the growth of the streamer head number as a function of time in N 2 -O 2 mixtures at 100 and 200 mbar. Decreasing the oxygen concentration decreases the nonlocal photoionization, and hence allows for local instabilities and more branching. The oxygen concentration in N 2 -O 2 mixtures affects streamer branching not only by smoothening the electron number density in front of streamer heads but also by the creation of an inception cloud. Streamers in pure nitrogen have no noticeable inception cloud, which gives the nitrogen streamers a longer effective propagation time during a voltage pulse of 550 ns; they branch more both as a function of space and of time. However, the statistical results show that the number of streamer heads in high purity N 2 is less than in mixtures with 0.1% O 2 , and it depends on pressure.