In Praise of the Unexpected

There are many advantages to having an open mind in science; not being too set on a particular direction or outcome can be a good thing. For me, a career in Space Physics was completely serendipitous and certainly unexpected. Over the years I have been mentor and advisor to many students who have set their hearts on a career in Space Physics, and it has been a pleasure to be able to encourage and help them follow that ambition, while admiring their sense of purpose and forward planning. But my own experience was quite different—things just happened. So my advice is to welcome unexpected turns, and at the scientific level, to be ready to embrace a result that surprises, or maybe even disappoints. To quote from Mike Lockwood’s Frontiers article “In Praise of Mistakes” (Lockwood, 2022, https://doi.org/10.3389/fspas.2022.852798) which inspired this article’s title, “If one can recognize a mistake, it is often a path to unexpected progress.”


Aurora, Rockets, and Electron Precipitation
The aurora, australis or borealis, was not a topic that cropped up in that life in Australia.There wasn't even an Astronomy group at Monash then.But my high-speed movies of fish in a tank eventually evolved into high-speed imaging of the dynamic and spectacular phenomenon of the aurora in the northern hemisphere.Unexpected directions are usually the result of the influence of key people.While still finishing my fish project, marriage to a physicist had brought me to the University of Southampton.Wandering the corridors of the Physics Department, unsure of what I was going to do next, I met Pamela Rothwell.On discovering that I had a physics degree, she immediately enlisted me into the Upper Atmosphere Group and set me to work, analyzing data from rockets flown from the Outer Hebrides into aurora.Pamela's instant offer of a job, and her commitment to my role as a physicist was why I stayed in physics, and in particular space physics.She re-employed me after a break for children, and then part-time for many years, on one occasion rising up to her height of barely five foot and thumping her desk, saying "You are trained as a physicist not a social worker," when I was dabbling with a second decision to move sideways.Pamela's enthusiasm and drive were an inspiration, and I learned a great deal about the importance of these qualities from her.She was an expert writer of grant proposals, which led to my continuous but patchy employment.One piece of advice that I have constantly used and passed on is to remove all words with a negative connotation, however slight.It is a small but powerful tip, and it reflected her positivity and optimism.
My unexpected entry into the field of space physics at the end of 1969 was at a time when the understanding of auroral processes above the earth's atmosphere was increasing rapidly.Results from rocket and satellite missions were providing particle distributions above auroral structures, leading to discussions about possible acceleration mechanisms for the electrons.I was set the task of searching for a particular signature in the particle data, the elusive increase in fluxes of field-aligned electrons, believed by many to be physically impossible, but which had occasionally been sighted in particle energy spectra.The required electric fields of short duration parallel to the Earth's magnetic field have been an essential ingredient in the pot of auroral acceleration theories ever since.Gerhard Haerendel (Haerendel, 2022) has expertly described the scientific journey from those early days of space data until the present, pointing out that the key ingredients for understanding auroral processes were available by the end of the 1990s, but that explanations are still missing for many of the dynamic and dazzling auroral structures, features that are now observed with greatly increased resolution in time, space, and wavelength.

Auroral Small-Scale Structure
From the earliest days Pamela Rothwell was a champion of understanding small-scale auroral structures.I was aware of her exasperation that auroral arcs were regularly referred to as being many 10s or 100s of km wide, when the optical evidence from cameras she had employed in Norway was clear that they were often as narrow as km scales.Mistaking the height of an arc for its width when interpreting all-sky camera data became a bee in the bonnet of Pamela and it was a legacy she left with me.Decades on, there is still a need for care when referring to an arc, as they have many manifestations (Borovsky, 1993).Single filamentary structures that are only 10s of meters wide are commonly observed, but they are usually embedded within a wider structure, with a broad energy distribution.Structures that are usually called arcs are predominantly east-west aligned, extending in length many 10-100s or even 1000s of km and are mostly made up of a system of roughly parallel filamentary structures, which often display counter-streaming optical signatures (e.g., video in Haerendel (2021) is a perfect example of this ubiquitous phenomenon).Theoretical understanding of the fascinating detail in the smallest scales of the aurora became a theme to my research.How these theories fit into those for the larger scales of aurora continues as a tantalizing unsolved problem, or more likely many problems.

The Auroral Spectrum
The pioneering Norwegian physicist and mathematician Carl Størmer stressed "the necessity of simultaneous photographs of the auroral spectrum and of photographic height measurements" (Størmer, 1955, p. 162).The crucial ingredient of the auroral spectrum was introduced properly to me during the time I was doing a PhD under Pamela Rothwell some years later in the 1980s.(A doctoral funding grant was about to be wasted so I was persuaded to exchange my part-time research assistant job for this studentship.)As I was beginning to write my thesis on the topic of magnetic field perturbations measured from MAGSAT, combined with auroral all-sky movies from Svalbard, Professor Fred Rees (M.H. Rees) came to stay in Southampton.Although he was busy writing his seminal book "Physics and Chemistry of the Upper Atmosphere" (Rees, 1989), he took a lively interest in my research, and insisted that it was incomplete without including data from the Alaskan meridian scanning photometers, located on Svalbard.His intervention transformed my understanding of the auroral ionosphere.At his elbow, I learned about emission brightness ratios, of the chemistry of ionization and excitation in the aurora and the power of modeling these processes.The long-lasting friendship with Fred and Marjie Rees was (and still is) a bonus, as were the annual visits to Southampton of Fred's protégés, Dirk Lummerzheim and Antonius Otto.My horizons in auroral physics were expanded with another unexpected turn of events.
By far the most exciting part of doing auroral physics is being involved in new measurements, and particularly in watching new instruments evolve through their design and build, and finally to witness their first light.I was fortunate to be at the Ramfjordmoen site for some of the earliest runs of the European Incoherent SCATter (EISCAT) radar.It was exciting to extract our first signatures of auroral arcs passing through the UHF radar beam while our cameras were running, now a common-place event.With the advent of the EISCAT Svalbard Radar came new challenges, which were a natural continuation of Pamela's early involvement in Svalbard observations, and with my PhD research.Discussions began in Southampton about building a spectrograph to install in Longyearbyen.Its main purpose was to measure the hydrogen emissions resulting from proton precipitation at higher temporal and spatial resolution than previously achieved and with excellent wavelength resolution.The whole project was a very fruitful and enjoyable collaboration with Boston University, involving telephone meetings between Southampton and Boston, which included Fred Rees, whose master plan it was, Michael Mendillo and Jeff Baumgardner, whose brilliant design was the key to its unique properties.It became a successful project, with colleagues at UCL assuming a major role in building, installing and taking care of the resulting High Throughput Imaging Echelle Spectrograph (HiTIES), with its first light coming from an opportunistic campaign to measure the solar corona during the total solar eclipse in Cornwall in 1999.One (lone) frame containing signatures of H alpha, beta, and gamma was obtained through a very small and brief gap in the clouds.Alas our plans to solve the mystery of the solar corona were thwarted.
The instrument continues as a beloved workhorse at the Kjell Henriksen Observatory (whose name will always remind me of the help and friendship given by Kjell during my PhD).The first HiTIES measurements from Svalbard were in November 2000, so it has accumulated a valuable data set longer than two solar cycles, in particular of proton precipitation signatures, with all combinations of filters including the wavelength range of either H beta or H alpha profiles.Svalbard is a unique location for such measurements, with particular challenges and opportunities for observations in the region of the magnetospheric cusp; although it is dark at noon in midwinter, measurements are affected by sunlight at heights above the shadow line (a fact that I used in my PhD research from Svalbard).The removal of this sunlit contribution from hydrogen profiles was achieved during an event with a bright and variable "cusp spot," seen in early data from the Imager for Magnetopause-to-Aurora Global Exploration mission (Robertson et al., 2006).Thus began a fascinating and productive time of combining ground-based spectrographic measurements with data from satellites, using particle spectra as input to modeling the hydrogen emission profiles, in collaboration with Marina Galand and our Boston colleagues, as well as Harald Frey (Lanchester et al., 2003).At this time Mike Lockwood was leading the Southampton group and so another opportunity came to study proton signatures within the larger magnetospheric laboratory (e.g., Throp et al., 2005), while learning from Mike's knowledge in this field (and much more).HiTIES became a treasure trove of data for postgraduate students, and so it continues.The spectral information that it gives is particularly valuable in determining rotational temperatures through synthetic modeling of molecular spectra.A statistical study was made on the sunlit N 2 + emission, comparing it with steady state theory, and finding a clear change in the rotational development of the molecular bands in different conditions (Jokiaho et al., 2009).New science is still emerging from the latest HiTIES data (D.Price, 2021; D. J. Price et al., 2019) in which the neutral temperature of the auroral thermosphere is found at spatial and temporal resolutions not before achieved.The information contained within the auroral spectrum keeps on being mined by using a judicious choice of filters and with new analysis techniques.

Electric Fields and Currents
My first direct experience of measuring ionospheric electric fields was from the EISCAT tri-static radar, in which the flow of ions at the intersection of three radar lines-of-sight allows an estimate of the field strength and direction in that volume.Collaborations are the best part of life as a researcher; during this period of my career, 10.1029/2023CN000209 4 of 6 I benefited hugely from working closely with colleagues from Oulu University and MPE Garching, who set up optical instruments at the radar site and shared their excellent data.The video referred to above (Haerendel, 2021) was made during one of these campaigns.(It was an evening of incredible aurora lying above the radar site, but because the radar was not running our field-aligned program, the Southampton party took time off in Tromsø and missed the live display.The unexpected can cut both ways!).
By combining such narrow angle camera images with radar measurements of horizontal electric fields, Haerendel et al. (1993) had confirmed that arcs drift separately from the ionospheric convection, implying that the magnetosphere and ionosphere are decoupled electrically in the region of auroral arcs.The high time resolution of the radar data, with the shortest sample time being 3 s, provided evidence for the existence of larger than expected electric fields (>400 mV/m) very close to an arc (Lanchester et al., 1996).Electric fields of 1 V/m had been measured by the Freja satellite, but the link to the optical signature was an important starting point for theoretical studies.Large plasma velocities were found to be parallel to the arc, and in the same direction and with similar speeds as bright features which moved along the arc.In other events, very narrow filaments with large fluxes and high peak energies were seen to be embedded in wider regions of lower brightness and fluxes (Lanchester et al., 1997).
The radar also gave high resolution measurements of changes in both ion and electron temperatures close to narrow auroral structures, which could be interpreted as the signature of field-aligned currents and heating in this small region.The results fed nicely into simulations carried out by Antonius Otto and colleagues (Lanchester et al., 2001;Otto & Birk, 1993;Zhu et al., 2001) which could model these small and dynamic features.New techniques in analyzing the auroral spectra from HiTIES (D. Price, 2021) are revealing gradients in temperature very close to auroral structures which are interpreted as resulting from heating from both field-aligned currents and those closing in the ionosphere.The significance of large and variable heating within these localized regions of dynamic aurora needs to be considered within the larger context of Joule heating as an input to global models, which is one of the aims of present research at Southampton.Auroral electric fields had been a feature of ionospheric research at Southampton from the beginning in the 1960s, using rockets flown into aurora (Reeve et al., 1978), and from the release of barium vapor into the auroral region from rockets (Rothwell et al., 1974).The flow of a different species of ion measured by optical methods became a renewed topic of interest in Southampton when Nickolay Ivchenko was working as a researcher in the group.It was his brainchild to use the long-lived emission of O + from the doublet at 732.0 nm to measure small-scale ion flows within the aurora, while also measuring the energy and flux in the region with different selected emissions.His design for the triple camera system, which we named ASK for Auroral Structure and Kinetics, was eventually achieved and continues in operation on Svalbard.So another unexpected privilege came my way to be the champion of an exciting and innovative instrument, watching it emerge from early dreams into unique results (Dahlgren et al., 2009).The original idea to solve the continuity equation for the O + ions is a complex problem, with ion velocity as a free parameter.The results give the dynamic changes of the ion population in the 3D volume around the magnetic zenith (Krcelic et al., 2023;Tuttle et al., 2020).By combining these results with HiTIES spectral measurements, the temperature of the ionosphere can be estimated and compared with the changes in electric fields, which is another strand in the search to understand the role of the small-scale aurora in heating the atmosphere.
Unraveling the dynamic changes in brightness and color of the aurora in space and time is fascinating but difficult, which is why there are remaining questions about the processes at play above the aurora.The size of the laboratory of the ionosphere and magnetosphere system is daunting, but new techniques, sophisticated measurements and computer simulations keep giving new insights.Another very valuable addition to the armory is Aurora Zoo, a citizen science project that evolved in the Southampton group over many years.It is now operating with excellent results.The first published paper which acknowledges Aurora Zoo contributors is about some very unusual small-scale structures of non-auroral emission, most likely resulting from ionospheric instabilities and very strong E-fields.These features, named fragmented aurora-like emissions, had been seen previously, but finding more of them using the citizen scientist approach was crucial for being able to identify their small-scale dynamics with confidence (Whiter et al., 2021).These non-auroral "gadgets" within the auroral ionosphere, visible only in certain emissions, and with no field-aligned component within their shape, were definitely unexpected.

Final Words
Although this article is not meant to be a chronologically complete record, there was a significant interval when my research diverged from the aurora.Henry Rishbeth joined the Southampton group, and my funding became 10.1029/2023CN000209 5 of 6 linked to the study of the effects of waves and sporadic E layers in the summer ionosphere, using the EISCAT radar.The work was a very fruitful collaboration with Tuomo Nygren from Oulu University, where I spent many weeks over several years as a visitor to the group, and where I learned so much, including the limited ability to say huomenta, kiitos, and anteeksi in Finnish (good morning, thankyou, and sorry).It was my good fortune to have experienced the Finnish way of life and is more evidence for the unexpected delights of scientific collaboration.The Oulu connection was eventually my route back to auroral studies, making optical observations with Kari Kaila, but I am very thankful for the opportunity I had to study a completely different subject with such expert teachers as Henry and Tuomo.
My conclusion is that with a physics degree one can tackle anything, no matter how unexpected the topic.Thanks to one teacher, my choice of physics turned out to be a journey full of surprises.There were significant people (not all mentioned here) who were responsible for making changes to my path and who added greatly to the enjoyment of the journey."Auroral Physics" was unexpected as a career description; but the aurora is a beautiful demonstration of the unexpected-at every sighting it can astonish and amaze.