Studies in chromospheric and transition region events and their relationship with the corona using IRIS and AIA

Student thesis: Doctoral ThesisDoctor of Philosophy


This thesis is concerned with studies and analyses for Far-Ultraviolet (FUV), NearUltraviolet (NUV) and Extreme Ultraviolet (EUV) observations of chromospheric and transition region (TR) dynamic phenomena. The solar chromosphere is host to a large number of dynamic phenomena and processes on both large and small spatiotemporal scales. The corona is arguably even more enigmatic, extending millions of kilometers into space and exceeding temperature of 1 million kelvin. Not only are the behaviour of these layers complex but the relationship between them is also poorly understood. A multitude of theoretical models attempt to explain the coronal heating problem but are not unambiguously proved through observations. While micro-flares and other small-scale, ubiquitous phenomena have been proposed, a deeper analysis of these small-scale events is required to provide meaningful constraints on at least some of these models. The following paragraphs provide a brief summary of the collection of work presented in this thesis: Chapter 1 provides an introduction to general theories and solar mechanisms relevant to the work presented in subsequent chapters. Extensive accounts of observation, image processing and data extraction methods are provided in chapter 2, particularly those details that could not be included in the studies described in subsequent chapters. Chapter 3 presents high-resolution observations and some spectroscopic properties of chromospheric jets emerging from a bright, penumbral foot-point. Distancetime plots of images from the Interface Region Imaging Spectrograph (IRIS) and the Atmospheric Imaging Assembly (AIA) are constructed using a path-tracing method. From these, plane-of-sky emergence speeds of 23-130 kms 1 are determined from two discrete data sets. These plane-of-sky speeds are considerably higher than the expected chromospheric sound speed. Maps of velocity, line-broadening and intensity are also constructed from the spectroscopic data. The jets exhibit signatures in observations at several wavelengths in cool and hot channels of IRIS and AIA, suggesting a multi-thermal nature with instances of coronal temperatures. Spectral analysis of these jets reveal Doppler line-of-sight speeds of 10-22 kms 1 , facilitating true velocity estimates. Full-Width Half Maximum and intensity spikes also overlap these areas of strong blue-shift. Each jet lasts an average of 15 minutes and occur 5-7 times over a period of 2 hours. Several physical mechanisms for this phenomenon are discussed, particularly Alfv’en wave shock trains based on the presence of line-broadening throughout the jet region and the suggestion that the line-of-sight Doppler movements represent wave amplitudes as opposed to bulk plasma motion. The close match between this study’s results and those of other jet and spicule observations are also discussed. It is imperative that the contribution of microflares to the heating of the corona is understood. What is their maximum contribution? This may be constrained by detecting small-scale brightenings. Detection of these brightenings in quiet regions is useful as these are the most commonly occurring regions on the solar surface, and are also less susceptible to large, rapid changes that may disrupt the detection process. While this study cannot guarantee any distinct conclusions, it can provide limitations on leading theories that attempt to solve the coronal heating problem, whereby accurate statistical relationships may be used as a form of constraint by which plausible theories must comply. Chapter 4 concerns these detections and limitations by describing and testing a new filtering method on synthetic data and and demonstrate its application to IRIS time-series imagery in the EUV as a proof of-concept. This method is based on spatio-temporal band-pass filtering, adaptive thresholding and centroid tracking, and records an event’s spatial position, duration, total brightness and maximum brightness. Average area, maximum and total brightness, and position are also recorded as functions of time throughout the event’s lifetime. A synthetic data cube is generated using a real IRIS slit-jaw time series as the background. A Poisson distribution is added to each pixel based on their mean intensity as an analog for randomly varying background intensity. 1280 Gaussian profiles are superimposed on to the synthetic cube as artificial brightening events with random distributions of spatial and time positions, amplitude, area, duration, speed, etc. The band-pass filtering in space and time is achieved via convolutions with appropriate kernels, while the varying threshold is based on the Poisson noise model for each spatial pixel. The method yields an event detection rate of 81% and a voxel detection rate of 49% in tests on the synthetic cube. The limits of this method are then tested by applying it to a very dense, real-world-analog data cube with 10; 000 artificial brightenings. The properties of these brightenings are based on power-law distributions. The power-law distributions and event properties of these are accurately determined while demonstrating the difficulty of extracting power law properties from real data due to instrument and method sensitivities. Applying this detection method to real IRIS QS slit-jaw imaging data spanning 19 minutes over a 54:4000 55:2300 FOV yields 2997 detections. Additionally, an extra measure of complexity of each detection is designated by their potential to fragment or merge over time. Therefore the number of distinct regions constituting the brightening event is recorded over time, and a brightening’s maximum number of regions over time is recorded. 1340 of these detections either remain un-fragmented or fragment to two distinct regions at least once during their lifetime, equating to an event density of 3:96 10 4 arcsec 2s1. Chapter 5 presents a multi-wavelength analysis based on the results of chapter 4 by applying the same method to an active region (AR) IRIS time series across multiple channels. Threshold values chosen for 1400 ˚A in the previous study remain unchanged while new threshold values are applied to both 1330 ˚A and 2796 ˚A. Using a stringent and statistical event-match criterion, it was determined that 3248 events occur across each wavelength channel, 2400 of which remain unfragmented or fragment only once during their lifetime, yielding an event density of 1:3 10 4 arcsec 2s 1 within a 9000 10000 FOV over the course of 34.5 minutes. The characteristics of these detected events are compared between channels and against the results of chapter 4. A power-law fitting method is also applied to each characteristic from each wavelength which shows that the events’ average area, maximum brightness and total brightness obey power laws above some lower bound, while their average speed and duration do not. The Appendix provides greater detail of the methods described in chapters 2-5, particularly any obstacles, limitations or aspects not sufficiently explained in the studies of chapters 3-5 due to their published format.
Date of Award2020
Original languageEnglish
Awarding Institution
  • Aberystwyth University
SupervisorHuw Morgan (Supervisor) & Maire Gorman (Supervisor)

Cite this