We present new results from the Interface Region Imaging Spectrograph showing the dynamic evolution of chromospheric evaporation and condensation in a flare ribbon, with the highest temporal and spatial resolution to date. IRIS observed the entire im
pulsive phase of the X-class flare SOL2014-09-10T17:45 using a 9.4 second cadence `sit-and-stare mode. As the ribbon brightened successively at new positions along the slit, a unique impulsive phase evolution was observed for many tens of individual pixels in both coronal and chromospheric lines. Each activation of a new footpoint displays the same initial coronal up-flows of up to ~300 km/s, and chromospheric downflows up to 40 km/s. Although the coronal flows can be delayed by over 1 minute with respect to those in the chromosphere, the temporal evolution of flows is strikingly similar between all pixels, and consistent with predictions from hydrodynamic flare models. Given the large sample of independent footpoints, we conclude that each flaring pixel can be considered a prototypical, `elementary flare kernel.
The temperature distribution of the emitting plasma is a crucial constraint when studying the heating of solar flare footpoints. However, determining this for impulsive phase footpoints has been difficult in the past due to insufficient spatial resol
ution to resolve the footpoints from the loop structures, and a lack of spectral and temporal coverage. We use the capabilities of Hinode/EIS to obtain the first emission measure distributions (EMDs) from impulsive phase footpoints in six flares. Observations with good spectral coverage were analysed using a regularized inversion method to recover the EMDs. We find that the EMDs all share a peak temperature of around 8 MK, with lines formed around this temperature having emission measures peaking between 10^28 and 10^29 cm^-5, indicating a substantial presence of plasma at very high temperatures within the footpoints. An EMD gradient of EM(T) ~ T is found in all events. Previous theoretical work on emission measure gradients shows this to be consistent with a scenario in which the deposited flare energy directly heats only the top layer of the flare chromosphere, while deeper layers are heated by conduction.
