My Rewarding Life in Science

Abstract

This memoire covers my life history starting with my family’s background and their immigration to the US. It continues with my childhood, my early education, and my introduction to science. It then covers my professional research career including a variety of institutions and areas of physics ending ultimately in solar physics.

1 Skumanich Family Background

My father Petro Skumanich was born to a farming family in Pčolina (the town of Bees) about 1894 in the Zemplen district of the Austro-Hungarian empire (now Slovakia). My mother Mariya Scrip was born about 1895 in Čukalovtse (Chukalovtse), the next village higher up Northward from Pčolina. These towns and villages were on the South side of Carpathian Mountains, whose spine or crest divided Polish-Rusyn Galicia from Austro-Hungary. The Skumanich and Scrip families owned their farms, as did many others.

Hungarian was the primary language of the Zemplen administration. My father could speak and read Hungarian. His mother’s tongue at home was Rusyn, an Eastern Slav dialect. He learned to read and write at the local (and only) Uniate (Greek Catholic or Byzantine) Church school. He also knew how to read (and presumably speak) Polish, Ukrainian, and Slovak. He could read and understand Church Slavonic. Much of this must have come from his studies to be a priest. This ambition was never realized due to an unfortunate accident, in which he was responsible for the death of a village youth. This crime caused the priesthood to be closed to him.

My mother had a 3rd-grade education. She could slowly read and write Rusyn. She often had me write her postal letters in Rusyn and later in English when I learned how to read and write English. Nonetheless, she was quite competent in the New World.

With the outbreak of WWI, my father was conscripted into the Austro-Hungarian army to fight on the Eastern front. Early in the war, he was exposed to poison gas due to a “friendly fire” incident. He was captured by the Russians soon after and spent the remainder of the war in a Ukrainian commune, a farming cooperative, escaping only after rising through the administrative ranks.

After the war, my father decided to join his older brother, who had immigrated before the war to Cleveland, OH. My father wanted to buildup a “nest egg” and return to Pčolina to buy up land to add to his inheritance.

His brother was working for a railroad company and managed to get my father a good job in the company as a switch-yard locomotive driver. Unfortunately, somewhat later, he had an alcohol-related accident, which disabled the locomotive. To avoid prosecution, he fled to a mining town, Blackman’s Patch in Wilkes-Barre Township, neighboring the city of Wilkes-Barre, Pennsylvania. There were countrymen from my father’s village living in the Patch drawn to work in the Anthracite mines.

My father passed the requirements for obtaining Mining Papers (engineering certificate). That allowed him to direct the work of laborers and to decide which coal veins to attempt to open, where to set dynamite charges to break up the veins and to direct their removal. He was also able to make tools that he required to facilitate his work.

Ultimately, he decided to stay in the US and had my mother join him in late December 1928. It must have been a joyful reunion. I was born at home in the Patch just 9 months later on October 5, 1929. I was my mother’s second pregnancy, this time in the US, and the first to survive.

My father’s life as a miner was a difficult one. He would come home black with coal dust and wash in a portable galvanized circular floor tub. My mother would heat water on the kitchen stove and pour it over him and then wash his back. It was not an easy bath. Given all the other attendant mining ill effects, I vowed that mining was not for me.

2 Childhood and Education

The language of discourse in my parents’ home was always Rusyn, a western dialect of Russian, also identified as Ruthenian or Carpatho-Russian. My literacy training in Rusyn was at home. I started primary school with a smattering of street English, which was quickly replaced. My Russian literacy was developed in a contemporaneous Russian Church school along with Russian and Church Slavonic vocabularies. The latter was used for liturgical purposes. On rainy Sundays, my parents celebrated the Liturgy at home with my father as the priest and cantor with my mother, brother, and me as the choir. My earliest ambition was to become a priest. This ambition was ultimately quashed when my father resigned from the Church after a bitter row with a new ambitious priest.

In my childhood years, I was drawn to the comics. My mother had a boarder that subscribed to a Sunday paper. My younger brother and I would surreptitiously remove the comics, carefully read them, and return them to the “undisturbed” paper. The comic that had a lasting effect on my love of Art and History was Hal Foster’s Prince Valliant. An adventure strip, set within the Middle Ages around knighthood, with a line art that pleased the eye with its verisimilitude and richness of color. My new ambition was to learn to draw as well as Foster. I bought books that offered self-taught courses in drawing.

I had the pleasure of an art accomplishment in a primary school incident. I drew a free-hand copy of the façade of my school building and showed it to my class teacher. She liked it and had it printed in the school bulletin. My art was on display!

3 Introduction to Science

The singular event, which launched me into the study of astronomy and ultimately physics, was a brilliant red (oxygen) nighttime aurora in August of 1939. My mother called me outdoors to view a bright sky that scintillated with blood-red colors. I was awestruck. She claimed it was an omen foretelling a bloody war. Indeed, Germany invaded Poland a month later, September 1939, of course a coincidence.

The next night I went out to see if the event would be visible again. No, the sky was dark, but I noticed that there were points of light that I had never noticed before. I asked about them of my grade school teacher, and she directed me to the school librarian, who labeled them as “stars.” She began my study of the stars by bringing me various books from the public library.

My primary school years were involved in learning the constellations and their associated classical myths, the names of the visible stars, and other night-sky phenomena. I attempted to make a telescope with Edmunds Scientific lenses and tubes but never finished the project due to a lack of tools.

At one Christmas, I was given a chemistry set, from which I learned of the chemical elements and chemical reactions. One of them was making a mixture of powdered zinc and sulphur, which when heated, would “explode” with a flash and much fumes, a fabulous experiment.

Early in my secondary school experience, I was enthralled by Edwin Hubble’s “The Realm of the Nebulae.” In the equation relating absolute magnitudes to luminosities, I found the mystifying word “log.” When I asked the school math teacher to explain what was meant, he brushed me off with the answer that I would learn about “logs” in the next year’s course. I was disappointed but realized that I had to learn mathematics to understand the language of science.

Later I joined The American Meteor Society as a student observer. This entailed reporting meteor number counts and entering paths onto celestial maps of the night sky. I even convinced a group of my high school peers, while we were doing contract farm labor in the Finger Lakes (NY) area, to help me with observations of the 1947 August Perseid Shower. We brought out our mattresses and arranged them in a circle with our heads at the center. Someone kept count, the others would call out an event, and I would plot the path. Alas my “crew”, exhausted by the day’s labor, fell asleep before the peak of the shower. I labored on alone.

4 Higher Education

I had held a variety of jobs during my final years in secondary school, initially for family finances and finally for college tuition. By the summer of 1947 my father had succumbed to anthrosilicosis (black lung) from coal mine dust. Thus my mother was thrust upon public welfare.

The summary nature of my secondary school physics courses inspired me to pursue additional more detailed studies. For economic reasons, I started at the local Wilkes College (initially Bucknell Jr. College) in Wilkes-Barre and entered into a two-year Physics Certificate program in the fall of 1947. I supported my school expenses by working in the college bookstore as well as by commercial jobs.

I met three other physics matriculants from the suburbs, William (Bill) Holak, Angelo Campanella, and Fred Bellas. We became life-long friends. Holak’s ethnic and religious background was essentially identical to mine, so our bond became familial.

Near the end of the first semester of our sophomore year (last year at Wilkes), Holak, Campanella, and I learned that Penn State would not insure our transfer there for the fall 1949 semester. However, we could and did transfer in the second half of our sophomore year at Wilkes.

After settling down at Penn State, I learned that I had to pay a matriculation and dorm fee, money that I did not have. I was on the verge of being dropped when Campanella’s father, a grocery store merchant, paid my fees. I have been forever grateful for his kind help.

During my junior year (1949 – 50) at Penn State, I was awarded a one-year State Senatorial scholarship. Subsequently, in 1950, I started to work at menial jobs. Later I was awarded a summer research assistantship in the department to work with one of the professors to reduce molecular gas phase data to allow the determination of the Van der Waals parameters. In my senior year, I was hired by the technical firm HRB,¹ where one of my duties was the numerical inversion of fairly large matrices associated with infrared images with the then newly available desktop Wang electromechanical calculator.

Although there were no astronomy courses at Penn State, there was an Astronomy Club, Alpha Nu, in the Physics department sponsored and led by Prof. Yeagley. One of our activities was to make ourselves our own 4-inch reflecting telescope. To this end, we ground our own parabolic mirror and then bought our own supporting tube. Unfortunately for me, I left in the middle of the project and seldom had the time or an available mechanical shop to finish my project.

A second life-changing event was a chance encounter in my final semester (1951) with one of my physics professors, Richard Stoner (Princeton PhD in Physics), who, when I told him I was planning to do graduate study in Solid State physics at Case (Western Reserve) University, insisted that I apply to Princeton instead. He helped me with the forms. The Physics department student roster was full, so my late application was forwarded to the Astronomy department, where I was accepted (one of only two admitted annually) and awarded a Higgins Fellowship. Now I had the opportunity of studying both fields, physics and astrophysics.

I entered Princeton in the fall of 1951 with a BS in Physics and left in the fall of 1954 with a PhD in “astronomia vestiganda.” The scientific and social milieu at the Department and its associated Observatory during this period is well described by Osterbrock (2000).

The department had a two-year cycle of the standard courses, which allowed much time for physics courses. Thus I rounded out my physics knowledge with a variety of courses such as quantum mechanics by R. H. Dicke (well known for his work in Cosmology), thermodynamics by H. Snyder (gravitational collapse singularity with Oppenheimer), and mathematical physics by P. G. Bergmann.

The departmental rule was to assign graduate assistants to work with each academic member. In my first semester (1951), I was assigned to Lyman Spitzer to help determine the internal motions within interstellar clouds. It appeared in the Astrophysical Journal and was my first scientific paper (Spitzer and Skumanich, 1952).

My next assignment, spring 1952, was with Uco van Wijk to help rectify the light curve of the eclipsing binary GL Carinae so as to determine the orbital properties of the star. The method of “rectification of light-curves” was published in 1946 by H. N. Russell, formerly head of the Department. The analysis was continued by J. Rogerson, my comatriculant. The results appeared in Van Wijk, Rogerson, and Skumanich (1955).

In the summer 1952, I was sent to work with Prof. A. N. Vyssotsky at the University of Virginia, Leander McCormick Observatory in Charlottesville, VA. The purpose was to determine if there was a difference in the dispersion of the “proper” motions of two spectroscopically distinct Main Sequence dwarf G and K field stars. The two groups were identified by a difference in their spectral G-band or molecular CH composition. This band is an indirect measure of the carbon ash, derived from the nuclear “burn” region, which increases with age. A kinematic age effect was indeed detected, with the weak G-band stars having a smaller dispersion than the strong G-band stars (Vyssotsky and Skumanich, 1953).

Note that Spitzer and Schwarzschild (1951) and Osterbrock (1952) had shown that stellar collisions with interstellar clouds would “heat up” stars, i.e., increase the dispersion of their “proper” velocities.

The Leander McCormick Observatory was a large structure similar to Princeton’s with offices, classrooms, and visitor living quarters. An attached building housed a 26-inch refractor for use in parallax and proper motion studies. It was sited on a hill overlooking the city. I lived in the visitor quarters.

I found Vyssotsky to be a very knowledgeable and warm mentor, whose Russian cultural background enriched my Slavic sentiments. He had a rich collection of talents. One of his talents, which he helped me to develop, was to find a particular page in a book without a search, i.e., at first pass.

I learned from Vyssotsky that he had been a staff astronomer at Pulkovo Observatory (St. Petersburg) when the Russian revolution broke out. He became a captain in the White Army and later, when the Reds won, escaped via Crimea to, ultimately, the University of Virginia. He later learned that the subsequent observatory administration erased his name from all the Observatory records. Years later, when I was at a conference on magnetic fields and Stokes polarization at Pulkovo, I mentioned this history to the local staff, and when I asked about his name being reinstated, I was greeted with a culpable silence. I fondly remember one of my last occasions to meet with him, at the 1958 AAS meeting at University of Florida (Figure 1).

Figure 1

Attendees of University of Florida 1958 AAS Meeting. Among them: F. K. Edmondson (AAS Secretary, front row, 6th from left); A. N. Vyssotsky (Leander McCormick, front row, 7th from left); A. Skumanich (LASL,, 2nd row, head tilted slightly right, just behind Vyssotsky); A. N. Cox (LASL, 4th row, just to the left of the woman above Skumanich and Vyssotsky); George Field (Harvard, right back row, tallest figure); J. Bev Oke (CalTech, in the next column, left of Field but two heads down).

My final assignment in the fall of 1952 was to Donald (Don) E. Osterbrock, at the time a visiting Post-Doc. I was to calculate, using a tabletop machine, a parametric sequence of interior models of red dwarf stars (Osterbrock, 1953). Thus I learned to construct stellar interiors from Osterbrock. His course in stellar atmospheres introduced me to the field of radiative transfer.

It is interesting to note that at this time (Spring 1953) the Princeton electronic computer (ENIAC at the Institute for Advanced Study) was online and was being used by Schwarzschild (Osterbrock, 2000) to solve the static hydrodynamic equations (stellar structure) with hydrogen fusion, whereas at the same time, Foster Evans (Los Alamos Scientific Laboratory) was at Princeton solving the time-dependent hydrodynamic equations with nuclear reactions (basically supernova-like equations) in his study of the feasibility of the “Super” (thermonuclear H-bomb; Evans, 1996).

My third life-changing event was in the spring of 1953 when I, by pure chance, met Bill’s first cousin Mary J. Berdy. It was while Bill and I were standing on a street corner waiting for a bus to downtown Wilkes-Barre. She stopped for the red light, saw him, and offered him a ride. We spent the evening with her. I was completely charmed. She became my wife in the fall of 1955.

The University required proficiency in two languages for the Doctoral program. My German exam was given by Schwarzschild, who complimented me on my native-like use of the language. My Russian exam was given by John Turkevich, Eugene Higgins Professor in Chemistry, which I passed easily.

Turkevich proposed that I and another graduate student, Alexander Tulinsky, translate the Russian Журнал Зкспериментальной и Теоретической Физики (Journal of Experimental and Theoretical Physics) cover to cover, which he would submit to NSF to initiate a translation program of the said journal. He was awarded the contract to this effect with us as translators. This started the Russian translation program at NSF. This translation program was transferred to the American Institute of Physics (AIP) and ultimately to Consultants Bureau as a publisher. I remained a translator through these changes till the late 1950s.

I started on my thesis in the Summer of 1953 with a three-month visit to the Mt Wilson Observatory in Pasadena, CA, under the tutelage and guest investigator privileges of Martin Schwarzschild, who became my thesis advisor. The trip protocol was part of the Department process for thesis candidate to obtain their observational material or observational experience. I traveled out there on the Atchison Topeka and Santa Fe Railroad, a romantic introduction to the American West.

In my case, Martin had suggested I look into the bright–dark symmetry or lack thereof in granulation to investigate the nature of the convection. The extant white-light images archived at the Observatory and the appropriate analysis devices were available to me.

My analysis of the bright–dark wings of the lateral intensity distribution indicated no significant areal asymmetry. I found a null result as far as the nature of the convective turbulent heat flow (Skumanich, 1955a).

During my Pasadena stay, I became friends with William (Bill) Livingston and Dale Vrabec, grad students at CalTech. Given the imminent Summer 1953 AAS meeting in Boulder, CO, Vrabec, an experienced pilot, decided to fly out to Boulder with a loaner Cessna and invited Bill and me to join him. It was a remarkable experience (Figures 2 and 3).

Figure 2

Bill Livingston (left), Skumanich (right), and Dale Vrabec (photographer) have arrived at the Prescott, AZ airport.

Figure 3

Skumanich (Princeton Univ., far left) with four young colleagues on the Top of the World (Loveland Pass) to attend the 1953 AAS-Meeting in Boulder, CO. From the left: (unidentified), Art Cox (Indiana Univ.), Dale Vrabec (CalTech), and Allan Sandage (Mt. Wilson).

On one of several Pasadena outings, I raised the question with Martin as to what might fix the dominant granulation size. I asked whether this was known. He knew of no such studies, so he suggested I look into the problem as part of my thesis.

Thus I turned to the study of convective flow in a density-stratified atmosphere that is thermally unstable. The previous study was Rayleigh’s (1916), who considered a homogeneous atmosphere. My results, for an adiabatic flow, indicated that the smaller the perturbation compared to the thickness of the unstable zone, the more unstable the mode (Skumanich, 1955b). I found that for larger perturbations, Rayleigh’s solutions were confirmed. My analysis benefited from discussions with Arnulf Schlüter (Dr. Rer. Nat., Rheinische Friedrich-Wilhelms University, Bonn, 1947), a Department Visiting Fellow, and with Richard Härm, Schwarzschild’s research assistant, in regard to the numerical integration method.

My thesis stimulated a decade of further studies by others with one or another of my assumptions removed (see, e.g., [3], Table 1). These studies also failed to produce a horizontal convective mode size that might explain the observed solar granulation.

My thesis examination committee included the full Department staff, Spitzer, Schwarzschild, van Wijk, and the venerable J. Q. Stewart of Russell, Dugan, and Stewart fame (Russell, Dugan, and Stewart, 1926). The procedure was going well until it was Stewart’s turn. He started by describing an experiment that consisted of dropping a steel ball onto a plate sheet of glass and ranking by size and counting the number of shards in any size group. I was nonplussed when Stewart asked me whether I would discuss why the frequency-rank distribution was given by a power law. I had no answer and reacted angrily to what I thought was a nonthesis question. I was embarrassed by my intemperate reaction and felt I had ruined my exam. It seems my committee overlooked this small brouhaha and passed me after all.

Looking back after so many years, I still feel it was not a suitable thesis question, but nonetheless it posed an interesting problem. By now this question must surely be resolved. It could have relevance in determining the nature of the debris from the collision of two asteroids. My diploma was formally issued October 22, 1954.

In the early winter of 1954, I began to think of employment prospects for my postgraduate future. My first effort resulted in an interview at the Army Ordnance Lab in Durham, NC. I took a train down to Durham to interview for a position as a project scientist to review and select research proposals from industry and universities. I would be in charge of overseeing whatever research I approved. What I remember most of this trip was the red clay of this southern state. I declined the position.

I also decided to visit the Winter 1954 meeting of the American Physical Society (of which I was a member) to be held in New York City. At the meeting, I came across a professional booth staffed by Los Alamos Scientific Laboratory (LASL) personnel presenting the science to be found at the Lab. They were also advertising a position that required knowledge of spectroscopy and radiative transfer. I could certainly fill the position, so I applied then and there. This ultimately led to filling out various forms to be submitted to the relevant individuals. After some time, I learned that I was a possible candidate, and I was invited to come to Los Alamos for a personal interview. I passed the interview, so that an extensive background investigative process, including a security clearance investigation, began. I was offered a position in the weapons test division.

In the meantime, I also sent my CV to Donald (Don) H. Menzel, whom I had met when he visited the Department, for a post-doc position at Harvard. My thought was to continue my convective flow analysis. Apparently, Schwarzschild had recommended me highly to Menzel, who proposed to Walter Orr Roberts, director of the High Altitude Observatory in Boulder, CO, that I be considered for a position as staff scientist. I was offered the appointment to pursue solar physics at the Observatory.

By August 1954 I had two job offers to consider. I decided to accept the Los Alamos position because it offered an interaction with very knowledgeable academic physicists from a wide range of fields, a post-doc experience. In addition, it paid more, which allowed me to financially help my mother.

5 Los Alamos Scientific Laboratory (LASL)

For my interview, I had flown to Albuquerque and from there to Los Alamos with Carr Airlines, a small carrier, that had a contract with the University of California to serve that specific (classified) route. I remember how appalled I was to see from the air a burnt out and desiccated region. But after some time at LASL, I fell in love with the landscape. I finally understood why St. Anthony choose to live in a desert.

Los Alamos was a small, closed city and mostly confined to two mesa tops. The buildings, except for Fuller Lodge, a large log-style two-story inn, were government style, wood sided, one or two story structures. There was a water tower and a machine gun posted guard tower at the fenced single entrance with a high wire fence around the entire site. It was possible to access the town only with an official ID card, mine also carried my Q-clearance (top secret) designation. I found the population very friendly and well cultured.

When I arrived at Los Alamos in the Fall of 1954, I was assigned a dormitory room in one of the temporary multirooms, two-story barracks on the North side of the main Canyon Rd. The town center was between Canyon on the North and Trinity Drive on the South. The lab was on the South side overlooking a deep canyon. I walked to work to Building C, a wooden structure with an IBM punch card computer, which sat off Trinity Drive.

In 1955, now married, Mary and I were assigned to one of many one-bedroom Gold Street apartments. When our son was born in 1956, we were assigned a two-bedroom house overlooking a neighboring canyon at 4321 Fairway Drive. With our second child, we were assigned to a three-bedroom portion of a duplex house. With our third child, we did not move.

I was hired by Herman Hoerlin, renowned for his early work on cosmic rays, as head of a newly formed section, J-10, in the Weapons Test Division or J-division of LASL. The section was created to initiate physics experiments treating the nuclear detonation as a high-energy source. Ralph Speece, an electronics engineer, and I were the first members of the section.

I soon discovered that there were two other Astrophysicists at LASL. Ralph Williamson (PhD 1943, U. of Chicago) in the Weapons Design division or T-division and Arthur N. Cox (PhD 1954, U. of Indiana) in the Weapons Effects section, J-15. This section was involved in studying the damage effects from a detonation. Such things as the heat and shock load on near and distant objects as well as nuclear radiation effects were its purview.

During my career in J-10, I participated in four weapons test operations. Two at the atomic bomb test site in Nevada, namely Teapot in 1955 and Plumbbob in 1957, and one test, Redwing in 1956, at the thermonuclear test site on Bikini Atoll in the Marshall Islands in the Pacific, and the last one a single thermonuclear detonation, labeled Teak, at a high altitude above Johnson Island as part of the Pacific Hardtack Operation in 1958.

My first experiment, with Ralph, was at the 1955 Teapot series of tests at Camp Mercury, Nevada. We planned to determine the temporal history of air fluorescence, the so-called Teller light, induced by the initial burst of gamma rays and later X-rays from atomic bomb devices that were being tested. The issue was to see what diagnostics might be learned about the nuclear reaction and to determine the emission state of the irradiated air. The observations were performed from an underground bunker with a surface carriage overhead that held the optics needed to access any line of sight. The bunker was accessed by a ramp with a blast-proof door. A quartz spectrograph with photoelectric detectors placed at appropriate spectral locations was used.

I was excited to see the results of my first experimental effort, so after the first “shot”, I drove out to my bunker from the observing station overlooking the Yucca Flats where the various tests were sited. There was no data! A quick check showed that the power supply to the detectors was not turned on. So, lesson #1, always use a check list! Director Hoerlin was not amused. The subsequent results were published in a LASL technical report ([1], Table 1).

Table 1 List of LASL technical reports (both classified and unclassified).

The full list of my LASL Technical Reports is provided in Table 1 and gives information to indicate the various experimental activities with which I was involved. Some of these reports are not classified and are available online.

One of the most unusual and complex experiments was a test of the Planck function distribution at approximately $\simeq 10^{7}\text{ K}$, in both optical and X-ray regions, filling the hohlraum using radiant energy from a thermonuclear device at Operation Redwing at Bikini Atoll in the Pacific (1956).

The design phase was led by Hoerlin (see [2], Table 1) and included, in addition to myself, Bill Ogle (J-Division Head), Harold S. Stewart (NRL-Optics Division, a LASL subcontractor), and Ralph Williamson (T-Division). Later, in the field execution, two new physicists hired by J-10, Don Westervelt and Roy Blumberg along with Dennison Bancroft (PhD 1939, Harvard), a J-10 scientific consultant, contributed their talents.

One of my responsibilities was the design of the hohlraum into which the bomb energy is dumped. The box had to be made of a high inertial mass material to slow down dynamic response to the rapid heating, surfaced with a material that would vaporize due to the strong surface heating and exert outward pressure on the walls. Finally, the box had to be thick enough so that the propagating Marshak front (a radiation “wave”, i.e., radiation field density “shock”) would not break through before the end of the measurement. The answer was a lead box with a gold plated inner surface!

Once the box had equilibrated, the black body X-rays and visual radiation exiting the hohlraum aperture traveled along a 1-mile-long vacuum pipe to the detector station. Bancroft designed and supervised the construction of the vacuum pipe system.

The experiment was successful, and we found that the Planck function was valid, within our experimental errors, at very high temperatures unattainable in laboratory conditions.

The subsequent year (1957) at Operation Plumbbob² in Nevada I was responsible for another unusual experiment. I was tasked to obtain an image of the neutron environment around the secondary of a mock-up thermonuclear device, using a series of lead pinhole baffles, the so-called PINEX experiment (see [5], Table 1).

This was executed with the device on a 500-ft tall tower, where I laid a lead brick blanket with a “pinhole” aperture at various levels of the tower to obtain an image of the neutron density at the source. The detector to record the image was at the bottom of the tower. The experiment was successful. The tower and lead “camera” were ultimately vaporized, and to this day, I still rue the need to have so much lead introduced into the atmosphere.

My last field experiment dealt with a single thermonuclear detonation (labeled Teak) at a high altitude over Johnson Island as part of the 1958 Pacific Operation Hardtack. The bomb was rocketed aloft from Johnson Island. The Teak experiment consisted of the study of the atmospheric physics effects of a megaton-range high-altitude energy release.

The experiment required a variety of theoretical calculations beginning with the X-ray dose-distance relation due to prompt radiation (see [6], Table 1). This provided the local excitation radiation field and surface level dosages. The latter indicated that to avoid eye damage to Marshall Island natives, the test had to be moved from the Bikini test site to the isolated Johnson Island site.³

My task was to predict the brightness and growth of the X-ray Fireball and to continue the study of high-energy excitation of air fluorescence (see [8], Table 1). The radiative evolution of the Fireball with time was calculated with the assistance of the J-10 support scientist Françoise Ulam, wife of famed Manhattan-Project mathematician Stanislaw Ulam.

A study of possible atmospheric emission mechanisms was designed (see [7], Table 1). This was expedited by consultation with the LASL visiting scientist Gerhard Dieke, the well-known molecular spectroscopist at Johns Hopkins. His expertise in molecular spectroscopy was of invaluable use to us because most of the luminescence was due to emission by molecular oxygen (O₂) and nitrogen (N₂ and the singly ionized nitrogen molecule N₂⁺) The different stopping depths of the prompt emission and X-radiation allowed a height resolution of the emitting molecules and hence of the atmospheric properties.

The test was successful and yielded an extensive data set. The preliminary analysis of the air fluorescence data was published by Westervelt, Bennett, and Skumanich (1960).⁴

With a nuclear detonation in a nearly empty space, the emergent radiation field is no longer blanketed by air, and we have the possibility of radiation pressures effects at the bomb surface. I analyzed this effect and reported on its feasibility at the 1959 LASL Symposium on Scientific Applications of Nuclear Explosives (see [9], Table 1). Edward Teller, who was advocating peaceful uses of nuclear explosives at the time, told me he found my analysis sound and of considerable interest.

A more extensive study of this effect is presented in a paper that deals with other possible high-altitude explosion effects, viz., the interaction of radiation accelerated particles, the fireball shock interaction with the Earth’s magnetic field, and generation of MHD waves, artificial aurora, bomb plasma interactions, etc. (Argo et al., 1959).

In addition to the programmatic activities at LASL, there were also others of an academic character such as frequent lectures, colloquia, and minicourses. For example, Burt Wendroff gave several talks on a new numerical method for solving the hydrodynamic equations, Connie Longmire gave a minicourse on plasma physics, Nick Metropolis presented a new approach to Monte Carlo methods, whereas Bengt Carlson covered numerical solutions of the Boltzmann equation for particles including photons (radiative transfer equation). Thus my physics understanding was significantly extended.

Mary and I found our Los Alamos life to be culturally enriching. Indeed, when Mary first joined me at Los Alamos, she was pleased to discover the presence of the Los Alamos Choral Society. She immediately enrolled us. One of the first pieces we studied and sang was Mozart’s Missa Brevis in F Major (K. 192). The cultural milieu in Santa Fe was rich with visiting quartets and the resident Santa Fe Opera Company.

6 University of Rochester

With a comprehensive test ban treaty negotiated by President Eisenhower in 1958, there were no 1959 or 1960 atmospheric nuclear tests. It seemed the opportune time for a year’s leave of absence. Harold Stewart, who had left NRL to become the Director of the Institute of Optics at the University of Rochester, urged me to come to the Institute as a Visiting Scientist for the 1960–61 academic year. In addition, based on Stewart’s recommendation, Robert Marshak, the head of the Physics and Astronomy department, offered me a 1-year Assistant Professorship appointment.

In the fall of 1960, I gave my first academic undergraduate course in thermodynamics with a new book by H. B. Callen. I had decided that it was the best book available at the time. It presented the concept of entropy from an entirely different point of view than that available in other standard texts. The book was a success with my students. I found them to be quite bright and extremely competent and diligent. It was a pleasure to work with them.

Coming from a nuclear weapons program, it should be no surprise that my research continued to be involved with the nuclear weapons effects program at the Institute. A. Battacharjie, a coworker at the Institute, and I calculated, using a Monte Carlo method, the ground radiation dose to be expected with a nuclear explosion at different heights above a layer of clouds (Skumanich and Bhattacharjie, 1961).

Academically and culturally, Rochester was a welcome interlude. It was a pleasure to be in the depths of the academic environment. The faculty was quite interactive, so I mixed with historians, linguists, mathematicians, and physicists. The library was a delight, and as a faculty member, I had open access to the inner sanctum of the humanities, sciences, and rare books.

My wife had worked in Rochester as a registered nurse at an earlier time which renewed previous associations. In addition, she had been a member of the Choral Society at the University’s Eastman School of Music and was able to rejoin. Furthermore, the Dryden Theater of the Eastman Museum presented a wealth of national and international free movies. We enjoyed the rich cultural environment. Finally, from an emotional need, we were only hours away from our families, friends and the Rusyn culture in Pennsylvania.

My rewarding experience at the University of Rochester led me to decide to leave Los Alamos and weapons work and reenter the academic environment. But here I was in a bit of a quandary. My first opportunity was a staff position without term at the High Altitude Observatory (HAO) of the University of Colorado, Boulder (CU Boulder), offered to me by Walt Roberts while at Los Alamos, where he gave a colloquium. He told me that HAO was to undergo incorporation into a new larger institution to be dedicated to the atmospheric sciences in general. This was a standing appointment.

I also explored the possibility to stay at the University of Rochester. Marshak extended my appointment for another year. However, I feared that in such a short time and with my experience only in weapons physics, I would not be able to develop an astrophysical research program that would ultimately lead to a permanent appointment. Perhaps, I should have argued for a three-year appointment, but I was reluctant to do so. I was worried that even three years might not lead to a viable research problem. So I decided to accept the position at HAO.

We (two parents and three young children) left Rochester in the late summer of 1961, stopping to spend some time in Pennsylvania before leaving with sad hearts for the new opportunity in Colorado.

7 High Altitude Observatory – Early Years

My family and I arrived at Boulder in the first week of October 1961 after an early fall snowstorm. We were temporarily assigned a house originally intended for an Australian visitor whose arrival was delayed. I was duly appointed as a staff scientist at HAO and as a Lecturer in the University of Colorado Department of Astrogeophysics on September 16, 1961. Very shortly afterward, in January 1962, HAO became the initial division within the National Center for Atmospheric Research (NCAR) managed by the University Corporation for Atmospheric Research (UCAR), and I became a staff scientist of NCAR.

Walt Roberts, the then director of HAO, became the first Director of NCAR and President of UCAR, whereas John Firor was hired to the Directorship of HAO, and R. Grant Athay became Head of the Astrogeophysics department. John was an understanding person and accepted my initial lack of direction for my research. Coming from a weapons laboratory, I had no “irons in the fire” to pursue. Even at the University of Rochester, my research was related to atomic weapon effects.

7.1 Chromospheric Activity Decay with Age

In reviewing the literature looking for research ideas, I came across a short note by Vyssotsky and Dyer (1957) on “Population Differences Among M Dwarfs.” Using the presence or absence of chromospheric Ca II H-K emission to assign class membership rather than a chemical difference, as was done by Vyssotsky and Skumanich (1953), Vyssotsky and Dyer (1957) found similar peculiar velocity dispersion differences. Main sequence dM stars with H-K in emission were in the kinematically cooler, smaller peculiar dispersions, population class, and hence younger.

A previous evidence for age dependence of H-K emission dwarfs was that of Delhaye (1953). Using local solar neighborhood dM and dMe stars listed in the extant literature, he found that dMe stars had a significantly smaller scale height (lower mean velocities) perpendicular to the galactic plane than dM stars. He surmised that “ces étoiles formeraient donc un sous-système très plat et pourraient être très jeunes.” (these stars […] might be very young.) This result was treated with circumspection, as the sample number was quite small.

This prompted me to initiate a study of the peculiar velocities of the field dG Stars (solar like) using Ca II H-K emission (presence or absence) as the classifier, as was done by Vyssotsky and Dyer (1957). The results indicated that the Ca II H-K emission group was younger than the nonemission group (Skumanich, 1965).

Concomitant with the search for stellar chromospheric emission data for field stars was the search for photometric data that would yield stellar luminosity, color (a proxy for surface temperature), and degree of chemical metallicity. I found a homogeneous photometric data set in an unpublished Yerkes Observatory stellar catalogue by Stromgren and Perry.

In the fall of 1963, Olin Wilson gave a colloquium talk at HAO on his investigation into Ca II H-K emission intensities (eye estimates) in Galactic Clusters and field stars. The object was his search for a probable correlation between chromospheric activity and age in Main-Sequence Stars. Cluster ages were derivable from their deviation from the zero age distribution in the Hertzsprung–Russell diagram (luminosity-color plane), but the lack of an accurate homogeneous data set for the field stars precluded his doing so.

During the colloquium, I stated that I could date his stars using the Stromgren and Perry catalogue. We could construct an accurate and coherent H-R diagram and thus date the emission field stars by their location. He suggested to collaborate, and we found that indeed the strongest emission field stars inhabited the zero-age H-R region (Wilson and Skumanich, 1964).

Subsequently, Wilson, with a newly designed photometric detection system, began publishing the “apparent” equivalent widths of the Ca II emission lines (widths that included the equivalent width due to the underlying photospheric absorption line) for the Hyades cluster as well as that of field stars.

I had found, in Olin’s field star catalogue, the Ca II emission data for the Ursa Major cluster and its associated stream stars. Luckily, an age for this cluster was also available.

In the case of the Sun, whose age was well determined, I had Olin’s measure of the solar Ca II H-K emission “apparent” equivalent width (in angularly integrated light, i.e., reflected Moon light).

After subtracting the underlying photospheric line’s equivalent width and removing the color dependence of the normalizing continuum (Skumanich and Eddy, 1981), I had three data points, Hyades, Ursa Major, and the Sun in an emission luminosity vs. age diagram.

I assumed that the temporal evolution of the chromospheric emission was in an asymptotic state and hence initial-value free, i.e., governed by a power law. I learned of this type of behavior from my late 1950s experience at Los Alamos. I found that an inverse square root of age power law fitted my three points.

In 1970, Robert (Bob) Kraft was in Boulder on a sabbatical, and in one of my conversations with him, I discussed my emission-age power relation finding. Kraft suggested that his observations with Greenstein of the Ca II emission equivalent widths in late-type Pleiades and Hyades would allow us to add the Pleiades to the graph. His preliminary calculation indicated that the Pleiades extended the square-root relation. He also directed me to his rotation spin down data, which I found also followed a square root relation.

Finally, I discovered Peter Conti’s (1968) suggestion of a “relation between rotational braking and Lithium content,” so I added his Lithium data to my graph. Lithium followed a square root drop to the Hyades age but showed a significant overdepletion for the Sun.

I wrote up my results and sent the paper to the Astrophysical Journal (ApJ) Letters editor, Don Osterbrock. Don called me and urged me to withdraw my proposed letter and resubmit it as a paper to the ApJ main journal, and if so, he would deliver my manuscript by walking it across the hall. With Osterbrock’s imprimatur, the paper was published directly (Skumanich, 1972).⁵ We note that the data reduction procedures used to arrive at calibrated data were published much later (Skumanich and Eddy, 1981).

At the 1969 AAS meeting, Edward (Ed) T. Frazier presented scatter plots of simultaneous photometric observations of the solar network in a $2.4''\times 2.4''$ square region at the center of the solar disk. He plotted the relative contrast (i.e., mean-normalized brightness deviation from the mean over the map) of the Ca II K core (1.1 Å bandwidth) vs the same quantity for the Fe I 525.0 line. In turn, the relative contrast of Fe I 525.0 line was plotted against the vertical magnetic field (inferred from Fe I 523.3). From my seat at the meeting, my “eyeball” fit to the scatter plots implied a linear relation between Ca II K contrast and the magnetic field.

This sparked my interest in deriving a direct relation between Ca II K contrast and the magnetic field. I knew that I could calibrate the Ca II K relative contrast to derive the emission equivalent width index and examine its dependence on the magnetic field.

I proposed to Ed that I do so, and he agreed to share his data. We found, after a detailed analysis, that, indeed, the Ca II K emission index was linearly dependent on the magnetic field (Skumanich, Smythe, and Frazier, 1975).

This conclusion coupled with the Skumanich square-root law (Skumanich, 1972) led to interesting implications regarding magnetic braking and the spin-down of stars. This paper was completed during my 1973-1974 stay at the Laboratoire de Physique Stellaire et Planétaire (LPSP) of the Centre National de la Recherche Scientifique (CNRS).

7.2 Radiative Transfer and Resonance Spectral Lines

The formation of resonance spectral lines, i.e., the prediction of their shape and strength in the emergent radiant spectrum of the Sun, depends on the equation of transfer (the Boltzmann evolution equation), which allows us to calculate the radiation field (the radiative Boltzmann function) through the Sun’s outer layers given the emission and absorption coefficients for the spectral line in question. These coefficients depend on the number of atoms that occupy the internal energy states that define the energy of the line. They are determined by population rate equations that contain the detailed processes, collisional and radiative, that populate or depopulate the energy state.

In the case of a steady state, detailed balance holds, and the population rate must equal the depopulation rate so that with conservation of the total atom density, we have a single equation for the population ratio of the two energy states involved.

The radiative process in the population equations is, in terms of kinetic theory, given by the binary product of the local photon density and exciting particle and the interaction (absorption) cross-section. The photon density is defined as the angular integral of the Boltzmann photon function or its equivalent, the specific intensity. The latter usage leads to the term “mean specific intensity.” I find it preferable to use the directly understandable photon density designation.

Following the Harvard astrophysical approach, the photon density is represented by an integral operator, the so-called “Lambda” operator, which maps the source function into the monochromatic photon density. This leads to a nonlinear system of equations, whose solution yields the ratio of occupation numbers (i.e., emission to absorption coefficients), the so-called source function.

I received a letter from Grant Athay, dated 19 Aug 1965, where he writes: “I stand in the peculiar and unenviable position of claiming that all existing theories of line formation, including all solutions for the two-level atom, are devoid of physical meaning. I now have one convert to my cause, viz. [J. C.] Pecker. If you are willing […] meet me in Munich. We (you and I) have some exciting work to do.” This was a remarkable assertion and stirred my interest.

I soon met Grant in Munich, and he went through his argument with me. I began studying his set of equations. Ultimately, I found errors in his analysis, which, when corrected, negated his conclusion. The world was not to be changed.

At one point in his argument, Grant made use of the gradient of the radiant flux-density that occurs in the first angular moment of the transfer equation, i.e., an energy conservation equation.

On reflection, it occurred to me that we could replace the photon density in the population equation with its form as defined by the energy equation rather than by its primitive definition. This introduces the gradient of the flux-density into the population equation as an integro-differential operator (see Kourganoff 1963) acting on the source function. Because its kernel has a shorter range than that for the Lambda operator, we have better convergence properties.

At Los Alamos, I found that inversions based on the energy equation were more stable than using the primitive integral representation of the photon density. So I proposed to Grant that we use my formulation given above to obtain better source functions (i.e., population) solutions. The analysis and derivation of the numerical algorithm is given by Athay and Skumanich (1967)⁶ and was programmed for the Cray by William B. (Buck) Frye, a HAO research assistant.

With a robust and accurate flux-gradient inversion algorithm, Athay and I undertook to find a thermal (kinetic) model that reproduces the observed Ca II emission features. We were successful, and the results appeared in a series of papers (Athay and Skumanich, 1968a,b,c,d,e, 1969). How accurately these models represent reality depends on how good the two-energy state model of the singly ionized calcium captures reality.

Our results were presented to the 1967 Bilderberg Conference (see Figure 4) and helped to define the nature of the low chromosphere. At this conference, I met Roger Bonnet, director of Laboratoire de Physique Stellaire et Planétaire (LPSP), as well as Yvette Cuny, on the scientific staff of Service d’Astrophysique de l’Observatoire de Meudon. Roger presented his space observations of the solar UV spectrum. Yvette reported on her successful determination of nonthermodynamic equilibrium structure of a hydrogen model of the solar atmosphere. Both became my lifelong colleagues and friends.

Figure 4

Participants at the Bilderberg conference, April 1967. Front row, left to right: Jean-Claude Pecker, John T. Jefferies, Carla E. Boot (support staff), Marijke Burger, Robert W. Noyes, Edith A. Müller, Roger M. Bonnet, Simone Dumont, David L. Lambert, Yvette Cuny. Second row: Jacques E. Blamont, Jacob Houtgast, R. Grant Athay, Christiane Guillaume, Nicolas Grevesse, Osamu Namba, Cornelis de Jager. Third row: J. Paul Mutschlecner, Owen Gingerich, George Withbroe, Marcel G. J. Minnaert, Tom de Groot, Hartmut Holweger, Jacques Sauval, Hans Vesters (support staff), Pierre Souffrin, Hans Hubenet, Michel Hersé, Robert J. Rutten, Pierre J. Léna, Andrew Skumanich, Philippe Delache, Jaap B. Vogel (support staff).

The theoretical investigation of radiative transfer became a dominant subject at HAO – and has remained as such to these days – and led to the formation of a dedicated group of scientists, which included, in addition to myself, Grant Athay, Lew House, and Dimitri Mihalas (Figure 5).

Figure 5

Members of the HAO Radiative Transfer Applications Section (c. 1973). From the left, Grant Athay, Andrew Skumanich, Dimitri Mihalas (standing), Lew House (sitting).

7.3 Orbiting Solar Observatory OSO-8 (1975 – 1978)

During my 1973 visiting scientist position at LPSP, I was invited by Bonnet to participate as a coinvestigator in the French UV spectroscopic experiment on the NASA orbiting solar satellite OSO-8 launched in June 1975. It carried in its pointed section an LPSP multichannel spectrometer designed for the highest angular and spectral resolution achieved by spacecraft at the time. The chromospheric resonance lines of Ca II, Mg II, and H I as well as the Transition Region lines of O VI and Si III were to be observed.

The observing team was composed of R. M. Bonnet, P. Lemaire, J. C. Vial, G. Artzner, P. Gouttebroze, A. Jouchoux, J. W. Leibacher, A. Skumanich, and A. Vidal-Madjar. All are from LPSP except for Leibacher (Lockheed Palo Alto Research Laboratory, Palo Alto, CA) and A. Skumanich (High Altitude Observatory, National Center for Atmospheric Research and University of Colorado, Boulder, CO).

The pointed section of the satellite operated in real time from an observing center housed at the Laboratory for Atmospheric and Space Physics (LASP) of the University of Colorado. LASP had the other UV spectroscopic experiment in the pointed section. The Boulder center had direct real time access to the dedicated services of the OSO-8 Control Center at the Goddard Space Flight Center.

The control of the LPSP pointing system rotated among the team members. The need for continuous observation of the Sun was an exhausting experience. Indeed, in July 1975, upon my return from a scientific visit to the Observatoire de Nice, I was immediately plunged into OSO-8 operations, where I was submerged for the entire year. To make matters worse, through an oversight, I had committed myself to teach a first-year graduate course in Modern Astrophysics, listed by the Departments of Physics and Astrophysics, Aerospace Engineering, and Astrogeophysics, for the spring of 1976. The upshot was that I was working seven days a week during the spring. However, I had the honor of winning the prize bottle of Bourbon for having the most frequent successes in acquiring targets on the first try.

In November 1975, a joint Sunspot Observing program was implemented by Skumanich (chromosphere, LPSP-CNRS), Bruce Lites (transition region, CU Boulder) along with Lew House and Tom Baur (HAO-NCAR), who were to provide magnetic fields derived from concurrent ground-based Stokes polarization observations. My pointing talents were “spot-on,” and we harvested a wealth of data for a large inactive spot with a homogeneous umbra.

The OSO-8 spot observations of H I, Ca II, and (previously unobserved) Mg II yielded a unique upper chromosphere and transition region thermal model similar to that of the quiet-Sun, which implied that the intense magnetic field played only a passive role for the chromospheric heating mechanism (Lites and Skumanich, 1982). The thermodynamic model derived from the UV data was unique and led to many citations (see, for example, Kneer, 2010).

The updating of the thermal parameters in this case and future others was made significantly simpler than that in general use at the time by the application of a sensitivity analysis procedure to the multiline equations (Skumanich and Lites, 1986). This allowed us to determine the most important interlocking radiative transition for the line in question that needed to be changed.

My interest was also captured by the serendipitous observation of simple two ribbon class-C flare (Jouchoux et al., 1977). After collecting radio, X-ray, and hydrogen H$\alpha$ data, I presented an initial analysis at the OSO-8 workshop (Skumanich et al., 1977).

A review of the performance and preliminary OSO-8 results is presented by Bonnet et al. (1978). This was a soft-publication option to get data out to the community as soon as possible.

8 High Altitude Observatory – Later Years

8.1 Magnetic Regions and Stokes Analysis

In early 1981, Athay, as acting head of the Stokes–Magnetic Fields Program, asked me to take on the scientific responsibility for the Disk Fields Analysis Section. The responsible program scientist had left, and the extant magnetic field inversion algorithm, called AHH, failed in the inversion more often than not. My first challenge would be to update or replace the AHH Stokes-inversion algorithm. The subsequent task would be to analyze part of the archived Stokes data. I thought this to be a challenge and decided to drop my OSO-8 involvement. This plunged me into the world of Stokes polarimetry.

My entry into the field was facilitated by HAO visiting scientists at the time, David E. Rees (Department of Applied Mathematics of the University of Sydney, Australia) and Egidio Landi Degl’Innocenti (University of Florence, Italy). They became colleagues and friends.

The incorporation of polarization or photon spin to describe the radiative state introduces a three-dimensional quantity (the two amplitudes of the two circular spin states and their relative phase difference), which with the scalar specific intensity leads to a 4-vector radiative Boltzmann “function” (the Stokes 4-vector), which satisfies a vector transfer equation with the scalar absorption coefficient replaced by a generalized “absorption” matrix.

I note that the early 1980s versions of the vector transfer equation and the “absorption” matrix were plagued by various sign errors and other associated handedness issues for the circular states. A complex quantum theory formulation by Landi Degl’Innocenti (1983) resolved these issues and provided a physically consistent “absorption” matrix, called the propagation matrix.

Later, Jefferies, Lites, and Skumanich (1989, hereafter JLS), derived the propagation matrix and vector transfer equation using classical EM theory, which allowed us to understand the physical origin of the various terms in the quantum-mechanical treatment. Their results yielded an expression for the propagation matrix, which proved to be in agreement (after introducing quantum oscillator strengths) with the results of Landi Degl’Innocenti (1983).

The physics that JLS used in their construction of the propagation matrix entails the fact that the interaction of the EM wave with the magnetic field is mediated by the complex index of refraction for each polarization component. This index appears only in the spatial term of the phase function. The imaginary part leads to extinction, whereas the real part represents the magneto-optical effects.

Thus we only need to increment the phase at the current spatial location $z$ along the ray and use it as the phase at the new point $z+dz$. A factoring of the exponential with this new phase yields the product of the exponential of the original phase with the exponential of the differential phase. This second exponential function is linearized, which yields a multiplicative factor of the vector amplitude at $z$ to obtain the vector amplitude at $z+dz$. The linear part of the multiplier contains the index of refraction (see JLS, Eq. (3)). The difference of the amplitudes ultimately leads to the propagation matrix.

8.2 A New Stokes Inversion Algorithm

The vector transfer equation is easily integrated (see JLS, Eq. (39)) to obtain analytic Stokes profiles (primitive profiles) for a simplified thermal structure, represented by a linear source function vs. optical depth, permeated by a uniform magnetic field, with a propagation matrix with constant atomic line parameters. The two source function parameters, the vector magnetic field, and the absorption shape parameters are then determined by a least-square fit of the analytic profiles to the observed Stokes profiles.

The HAO authors of the problematic AHH inversion scheme (Auer, Heasley, and House, 1977) made a number of simplifications that reduced the number of free parameters to be fitted. They transformed the line intensity to an absorption line depth, normalized by the depth at line center, which eliminated the two source function parameters but at the expense of introducing a greater degree of nonlinearity. This restricted their fit to the three Stokes variables, normalized line depth, linear and circular polarizations, as functions of wavelength.

Thus a new algorithm was required that used the primitive equations with all four Stokes variables and introduced more physics into the propagation matrix, namely a line profile with damping and magneto-optic birefringence. The specific intensity would be fitted to determine the source function parameters and any parasitic light or zero point.

My development of a new HAO algorithm was joined by Rees and Landi Degl’Innocenti. Athay also asked Bruce Lites, who was working at the Sacramento Peak Observatory (SPO), in Sunspot, NM, and was a member of the HAO-SPO Stokes Consortium, to join in the development.

Rees was instrumental in my being awarded a semester-long Visiting Scientist position in the Department of Applied Mathematics, University of Sydney in 1983. There I met Graham Murphy, who was a graduate student being supervised by Rees. His interest was in Stokes polarimetry. We soon developed a friendship as well as a mentor–student relationship. For Mary and me, the stay in Sydney was enjoyable. We had access to the famous Sydney Opera house as well as great Fish & Chips bars.

The new HAO algorithm (hereafter, SL84) was tested by Skumanich and Lites (1985) on 1975 Stokes-II observations of a large OSO-8 spot (see Section 7.3) previously inverted with AHH. It was found that the added line opacity parameters yielded more plausible line strength values than those derived by AHH.

Skumanich, Rees, and Lites (1985) further tested the SL84 inversion on synthetic Stokes profiles for Fe I $\lambda 5303$ calculated with an extant physical model of a sunspot with an assigned vector magnetic field. The inversion field parameters successfully represented the vector magnetic field to within a few percent, except that the derived line opacity parameters differed considerably and were a poor diagnostic of the thermodynamic state. However, the derived source function slope was representative of that in the spot model.

Much later, Skumanich and Lites (1987b) published an updated version of the SL84 algorithm along with an extended comparison of the OSO-8 spot inversions, with a successive introduction of each new SL84 parameter with the AHH inversion. It was found, for example, that the AHH magnetic field was 30% below the SL84 value derived with the full SL84 parameter set.

Thus we found that the SL84 inversion method was robust and that it provided more reliable and accurate estimates of sunspot vector magnetic fields without significant loss of economy than any other extant methods.

8.3 Stokes-I and Stokes-II Science

Our first effort (early 1986) was to analyze the diagnostic usefulness of the neutral Mg I $\lambda 4571$ line using Stokes-I 1978 data. This was initiated as a collaborative effort with Rees and Murphy, who were in Australia at the time, where we explored the magnetic and thermal information content of the $\lambda 4571$ line. This line is formed in the relatively unexplored umbral temperature minimum region of sunspots. The line is meta-stable, is collisionally dominated, and can be represented by a linear source function.

In Lites et al. (1987), we found that the temperature minima of umbrae were more extended in height than previously believed. Hence the spot’s chromosphere was optically thinner than extant models indicated. In addition, although the Stokes profile of $\lambda 4571$ is a sensitive diagnostic of the minimum temperature, it is also very sensitive to the assumed value of the absorption line wing-damping constant, which is not a well-known atomic parameter for this line. For this reason, the $\lambda 4571$ line was considered of limited utility as a diagnostic of magnetic fields.

In the fall of 1986, I was reappointed a Visiting Scientist in the Department of Applied Mathematics (University of Sydney), whereas Bruce was awarded a Visiting Scientist position. There we initiated a new collaborative study with Rees and Murphy to invert the Stokes-I sunspot umbral data of the strong Mg I b lines at $\lambda 5172.7$ Å and $\lambda 5183.6$ Å. The analysis was finished at HAO with Rees as a Visiting Scientist and Murphy as a Newkirk Graduate Research Assistant.

The Mg I b lines are sufficiently strong that their source function cannot be approximated by a linear function of optical depth. An exponential form had to be added to the linear term to approximate the source function predicted by non-LTE computations for an extant model of a sunspot umbral atmosphere. The exponential was used because it was integrable and adds a new analytic component to the existing SL84 least squares inversion scheme.

The new inversion methodology was tested on synthetic Stokes profiles derived, as before, for a realistic sunspot thermal model. We found that the new scheme was most effective in recovering the magnetic field used in the synthetic spot model Stokes profiles if we restrict the fit to the Doppler cores of such chromospheric lines and fit both the intensity and polarization profiles. In addition, it was necessary to fit both members of the multiplet with their different Zeeman splitting patterns to obtain the optimal comparison with the synthetic field.

Applying the revised inversion scheme to the 1978 Stokes-I data (Lites et al., 1988), we found reasonable values for the magnetic field but only if an additional “macroturbulent” profile smearing was introduced, as well as a correction of the observed intensity profiles for stray unpolarized light. Due to the Stokes-I polarimeter limited spatial resolution, the results were considered descriptive rather than definitive.

The value of the Mg I b multiplet is that it is the most favorable of pairs of chromospheric lines for quantitative analysis of chromospheric magnetic fields.

Note that the analysis methodologies introduced in the magnesium papers (Lites et al., 1987, 1988) have been recently adopted and extended to magneto-hydrodynamic simulations by Dorantes-Monteagudo et al. (2022).

The ultimate analysis of the archival Stokes data was the inversion of the 1980 Stokes-II observations of four large sunspots. The data contained only the $\lambda 630.26\text{ nm}$ line of the Fe I multiplet.

Lites and Skumanich (1990) found that the magnetic field occupied significant fractions of the area with both light and dark penumbral filaments and that the intrinsic field has a similar threshold value both in the sunspot penumbra and in the surrounding plage areas.

The variation of the poloidal field strength and inclination with distance from the center of the symmetric sunspots exhibited little nonpotential character and was well represented by the potential field of a buried dipole except in the case of spots with twisted fields.

This work demonstrated that it was possible to invert the Stokes profiles of a single solar absorption line and to derive the corresponding vector magnetic fields with considerable confidence in the results. This optimistic situation was welcome from the standpoint of advancement of our understanding of MHD processes acting in the solar atmosphere, and as such it further validated efforts to improve the resolution of the Stokes observations.

9 ASP Science

9.1 Advanced Stokes Polarimeter

The successful vector magnetic field results from the Stokes-II data inspired the drive for a new polarimeter, to be called the advanced Stokes polarimeter or ASP. For this purpose, an ASP program was established as a joint project between the National Solar Observatory (NSO) and the High Altitude Observatory (HAO). It became a part of a larger thrust for Stokes polarimetry being advanced on instrumental, analysis, and theoretical fronts by a new Stokes Consortium, namely, the HAO, NSO, University of Hawaii, and the Astrophysical Observatory of Arcetri, Florence, Italy. Lites was appointed an Instrument Scientist to lead the HAO project. I continued my theoretical role.

The ASP would have updated components and would use the SPO Vacuum Tower Telescope (VTT) and SPO spectrograph to obtain higher spatial, spectral, and temporal resolution than was possible with the Stokes-II instrument. The use of CCD cameras would allow us to obtain simultaneous Stokes parameters of the solar image along the spectrograph slit rather than at a single point as in Stokes-II. To build up a Stokes map, we would move the solar image across the slit by slewing the telescope.

In addition to the polarimeter itself, a Telescope Model would need to be devised to compensate for the spurious polarization introduced by the VTT. Such a Telescope Mueller matrix, whose inverse is needed, proved to be fairly complex. The calibration model is described by Skumanich et al. (1997).⁷ The final polarimeter instrument is detailed by Elmore et al. (1992). The ASP system was commissioned in 1992 and became available to the scientific public.

The HAO effort to develop a complicated and expensive ASP program was not without its detractors. At an AAS meeting, Harold Zirin (Big Bear Observatory) approached me (standing at my ASP poster display) to criticize HAO for wasting federal money on a project that was not really needed. He said his work with a filter-based magnetograph in conjunction with H$\alpha$ images produced magnetic data whose functional use was superior to that from ASP.

Lites, Martinez Pillet, and Skumanich (1994) compared the inversion results from the ASP data with those derived from a filter-degraded version of the same data. Significant differences were found particularly more so in areas where “filling factors”, which are not accounted for in filter magnetographs, are important. Lites (1996) gives a more direct response to Zirin’s criticism.

Initially, Bruce and I were the only members of the ASP Analysis Group. Valentin Martínez Pillet, a HAO visitor on leave from the Instituto de Astrofísica de Canarias (La Laguna, Tenerife, Spain), joined us for an extended period of time. K. D. Leka, a 1994 – 1997 Postdoctoral Fellow in the NCAR Advanced Study Program, also joined the group.

We had help from Paul Seagraves, the HAO Senior Programmer, on software development. Other visitors or users often joined us as collaborators, or, conversely, we became their collaborators.

Bruce, as an Instrument Scientist, was responsible for the public use of the ASP facility and had help from David Elmore (HAO Senior Engineer) on facility glitches.

The ASP became the most subscribed instrument at the Dunn Solar Telescope. The instrument concept was put on the Japanese-US Hinode Satellite with Bruce Lites as the PI. The validation of precision spectropolarimetry with the ASP provided essential scientific arguments for the justification to build a 4-meter solar telescope (DKIST).

9.2 Magnetic Field Expulsion

The most singular event observed with the ASP was in 1992 during the passage across the disk of a decaying active region, NOAA 7201. After the disappearance of the trailing spot, in the same area the emergence and rise of a nearly closed magnetic system was observed (Lites et al., 1995). The intersection of the system with the photosphere yielded a topographic sequence from which the 3D structure of the magnetic field was found.

The initial flux emergence region shows a rather simple geometry, but it subsequently develops a small $\delta$-sunspot configuration⁸ with a highly sheared vector field along the polarity inversion line running through it. The magnetic system persists in the corona well after the dark $\delta$-sunspot has disappeared from the photosphere.

Observations of associated H$\alpha$ prominences (Lites et al., 1995) and accompanying X-ray emission by Yohkoh (Tsuneta et al., 1992) of the event indicated a similarity to a coronal mass ejection (CME).

B. C. Low (HAO) found that the magnetic field could be described by an analytic three-dimensional magnetostatic model representing a closed, spheroidal magnetic system, in which the Lorentz force arising from cross-field currents is balanced by the gravitational and pressure forces.

At this point in the history of this event, B. C. had a flash of genius and realized that we could develop a time-dependent evolution by invoking the concept of self-similarity. He explored this approach with the HAO post-doc Sarah Gibson, and the consequence was a new model of CMEs. The use of this model to analyze individual CMEs transformed the nature of space weather research.

9.3 Spots and Pores

Table 2 lists the observational highlights of the ASP related to spots and pores.

Table 2 ASP observational highlights.

The ASP data demonstrated that penumbra have fine scale structures and associated dynamical behaviors quite different from the extant understanding. The penumbras were found to contain narrow radial “spines” of more intense magnetic field more vertically oriented than their surrounding field. This second field, more horizontal, flux tube component with its associated strong downward material motion, arches downward into the photosphere at the edge of the spot.

In the case of pores, it was found that their size was related to their magnetic flux content and that the field inclination at the pore boundary increases as total flux increases. Both of these observations support the recent pore models proposed by Hurlburt and Rucklidge (2000).

The unprecedented ASP results caught the attention of an international group of mathematicians, Nigel O. Weiss (University of Cambridge), John H. Thomas (University of Rochester), Nicholas H. Brummell (University of Colorado), and Steven M. Tobias (University of Leeds). They argued (Weiss et al., 2004) that the field lines that plunge below the solar surface near the edge of the spot are pumped downward by small-scale granular convection outside the sunspot. They also found that such magnetic pumping could explain the abrupt appearance of a penumbra for pores above a certain flux level.

Note that Weiss et al. (2004) gave an excellent synthesis of the relevant ASP observations in addition to the contributions listed in Table 2.

9.4 Observing the Previously Unobserved

Two new magnetic manifestation were discovered in ASP observations, namely Horizontal “Internetwork” Fields (Lites et al., 1996) and Plage Azimuth Centers (Martínez Pillet, Lites, and Skumanich, 1997).

Due to the high linear polarization sensitivity of the ASP, we were able to detect weak transverse fields. This allowed us to discover that quiet regions near the center of the solar disk were found to contain transient small-scale (typically 1″ – 2″ or smaller), predominantly horizontal magnetic flux features that often occur between regions of opposite polarity (but weak) Stokes circular polarization profiles. These features occur in isolation of the well-known, nearly vertical, flux concentrations usually seen in the photospheric “network.” Hence these small-scale horizontal “internetwork” fields were labeled as HIFs.

We view the HIFs as the emergence of loops of flux, carried upward either by granular convection or magnetic buoyancy. Even though these entities show weak field strengths, they also seem to be fairly common and may contribute significant flux to the upper atmosphere.

The ASP has, for the first time, allowed us to observe the vector magnetic field of active region plages⁹ with an angular resolution of approximately 1 arcsecond. It was found that maps of the azimuth of the regions vector field show magnetic structures that are small-scale, several arcseconds, and that have the full 360-degree azimuthal distribution corresponding to diverging (or converging) field lines from a magnetic center similar to spots.

Some such structures are associated with darkening in the continuum and are examples of normal pores. In many cases, these structures do not show darkening in the continuum. We designate such features “azimuth centers” (ACs). Such structures could only be revealed by the type of precise measurements of the linear polarization obtainable with the ASP.

The AC structure appears to represent an intermediate state between elemental flux tubes and pores (or small sunspots). Within ACs, the flux tubes presumably still retain their identity, but the magnetic flux tubes have been concentrated together, perhaps by a converging flow that imposes a magnetic center. The magnetic flux in an AC is not large enough to strongly inhibit heat transfer to the upper layers, so the structure has normal photospheric brightness. Their role in the magnetic evolution of solar magnetism remains to be explored.

10 Solar-Stellar Chromospheres

10.1 Solar Activity

With the advent of the International Ultraviolet Explorer (IUE), circa 1980, a large data set of UV irradiances (radiant flux-density at the observer) for Solar-like stars became available. Furthermore, episodic measurement of certain solar UV irradiances, which play an important role in terrestrial atmospheric studies, were also available.

Hence there was a need for a general model to calculate any such irradiances for the Sun where surface sources could be uniquely identified and whose radiance (specific intensity) could be determined.

The first attempt to statistically identify such sources and determine their absolute chromospheric Ca II K emission radiance appeared in Skumanich, Smythe, and Frazier (1975).¹⁰ From a study of a magnetically quiet region at the center of the “disk”, it was found that two sources, the supergranulation “cell”, which is essentially free of elemental magnetic flux tubes, and “network”, a loose collection of elemental flux tubes at the downflow lanes of the supergranulation, could be uniquely identified and their absolute radiance determined.

To obtain any particular activity state in the solar activity cycle, we must add the contribution of plages as a third component to the minimum state. This requires knowledge of the radiance, size, and positional property of each plage. Fortunately for the former, the calibrated synoptic observations by Dick White (HAO) and Bill Livingston (KPNO) were available. For the latter, the WDC/NOAA Ca II K plage records sufficed. These were provided by Judith Lean, a post-doc at the Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado, Boulder, because of her interest in UV terrestrial irradiances. All three joined me in this effort.

It was found that a successful agreement between the constructed K irradiance and the measured irradiance was possible only if the (elemental) flux dispersed from the breakup of plages was added as a third “active network” component. The full description of the new K irradiance model is presented by Skumanich et al. (1984).

A well-received review of the long-term nature of solar variability circa 1981 can be found in Skumanich and Eddy (1981).

10.2 A Unique Ca II K Emission Dwarf M Star (dMe)

In 1982, Arthur Young, Professor of Astronomy at San Diego State University, came to HAO on a sabbatical leave. After a number of discussions with him regarding his solar-stellar connections research, he invited me to join him in the analysis of his recent observations of both Ca II K and H$\alpha$ emissions in dwarf M (dM) field stars. Our initial effort indicated anomalous properties of a number of individual stars (Young, Skumanich, and Harlan, 1984). Discovered here was Gliese 890, classified as dM2.5e (“e” for the presence of K emission) and estimated to be a fast rotator.

To fully investigate GLS 890, we initiated a series of observations using the 2.1-m telescope, Coude spectrograph, and CCD detector at the Kitt Peak National Observatory. This was a pleasurably experience for me as my last observing experience was with Olin Wilson back in 1964.

The analysis of our various observations (Young et al., 1984, 1986, 1990) indicated a single star with the shortest rotation period known, which had a photometric wave due to sunspots and a nonuniform distribution of H$\alpha$ emission plages on its surface. The plage and spot activity appeared to be at high latitudes compared to the occurrence of active regions on the Sun. Whether the extremely rapid angular velocity (60 times solar) confines such activity to the polar latitudes, or merely allows it to migrate to such latitudes, could not be established through our observations, which were confined to just the high latitude phase of the activity cycle. Perhaps we are observing the activity of a different type of dynamo than the solar one.

10.3 dMe Stars in General

My last venture in stellar activity with Young was to investigate the energetics of the most active population of field dMe stars. Extant evidence from observations conducted at the McCormick Observatory (Vyssotsky and Dyer, 1957) indicated that of the general population of dM stars ($N=305$) in the solar neighborhood 21% were Ca II K emission stars, or dMe stars. Of those, 20% also had the H$\alpha$ line in emission. The kinematic evidence indicated that this third group was younger than the dMe stars.

Our investigation of this third state (Young et al., 1989) used H$\alpha$ equivalent-width data from three observatories (Stauffer and Hartmann, 1986; Bopp, 1987; Young, Sadjadi, and Harlan, 1987). After subtracting background, the resulting excess equivalent widths were converted to H$\alpha$ luminosities. It was found that rotation plays no role in the degree of H$\alpha$ activity, i.e., the stars are not in the spin-down phase, so they must be quite young. The H$\alpha$ luminosities appear to be saturated at a fixed fraction, 20% of the coronal X-ray luminosity, which may be their ultimate energy source or indicative of the same heating mechanism in both spectral domains. Our results were the first to describe the nature of the saturated chromospheric state.

11 The Retirement Years

11.1 Early Period (Circa 2001)

I became interested in the effects of rotational distortion on the internal structure of solar-like stars during conversations with Steve Jackson (PhD, University of Chicago), who was at HAO at the time. I learned that he created the first code to successfully solve the coupled stellar interior equations and the Poisson equation for the gravitational potential because of the nonspherical shape of the density profile. The numerical methodology was based on a Princeton (Ostriker’s) scheme (labeled SCF) of iterating between individual sequential solutions for the interior and potential equations. The code worked very well for rapidly rotating early spectral type (massive stars).

Unfortunately, the code could not be used to study solar-like stars since it failed to converge for stars with masses less than 9 M_⊙. The methodology had to be reformulated.

Steve complained to me that his attempt to revise the SCF procedure was stymied by a lack of computer resources to test his progress. When I discussed the desirability of having a revised version of Steve’s code at HAO with Keith McGregor (Head of the Stellar Interiors Section), he agreed and arranged for a visitor’s appointment for Steve.

To our surprise, Steve arrived at Boulder with a punch-card version of his code and associated subroutines (a vestige of his work being done at Princeton – I recall my own experience there). A card-reader was found at one of the older federal labs in Boulder, and Steve transferred his extant code to a SUN workstation. He began testing various reformulations and finally arrived at a new successful SCF version that converged for any stellar mass.

We explored the nature of these solutions across the two-dimensional parameter space (rotational amplitude and gradient) and compared the new SCF solutions with those from other extant methods. The result was a robust computational algorithm (Jackson, MacGregor, and Skumanich, 2005) with updated auxiliary physical processes, such as the equation of state, opacity subroutines, energy generation reactions, convective energy transport, chemical composition, and so forth. The new code was able to treat distortions not attainable by other methods.

The first precise interferometric measurement of the shape of $\alpha$ Eri (Achernar, a B star) provided our first opportunity to model the physical structure of a rotationally flattened star. Employing masses for main-sequence stars of mid- to early-B spectral type, we could reproduce Achernar’s inferred equatorial and polar dimensions. This was achieved through a combination of rotational flattening/distension and a suitable inclination of the rotation axis (Jackson, MacGregor, and Skumanich, 2004).

Despite these successes, the models were discrepant in other respects, being (on average) cooler and more rapidly rotating than indicated by the observations.

Finally, we calculated a number of models for chemically homogeneous, differentially rotating, main-sequence stars with masses in the range 1 – 2 $\mbox{M}_{\odot}$ (MacGregor et al., 2007). For a rapidly rotating Sun, we found a reduced radiative luminosity. Relative to nonrotating stars of the same mass, all of our rapidly rotating models exhibited reduced luminosities and effective temperatures, and they displayed a flattened photospheric shape (i.e., decreased polar radii).

Thus solar-like stars not only brighten because of chemical evolution (nuclear transmutation), but also because of rotational spin-down. For a fixed ratio of axial to surface equatorial rotation rates, increased rotation typically deepens convective envelopes and shrinks convective cores. It may also lead to a convective core ($M = 1~\mbox{M}_{\odot}$) or envelope ($M = 2~\mbox{M}_{ \odot}$). These features are absent in a nonrotating star of the same mass.

Theoretical studies by D. R. Reese (Department of Applied Mathematics, University of Sheffield, UK) suggested that we might use asteroseismology, the study of stellar pulsations, to probe the internal structure of rapidly rotating stars. This allows the observed pulsation spectra to constrain/test theoretical models. We provided several SCF models for Reese to study with his pulsation code (Reese et al., 2009b). A successful identification of pulsation modes for a rapidly rotating $M = 25~\mbox{M}_{\odot}$ model was found (Reese et al., 2009a). This advance has opened new avenues for asteroseismology.

Another highlight of my later years at HAO was the celebration of the 75th Anniversary of the observatory, where I had the good fortune of meeting again with old friends and colleagues (Figure 6).

Figure 6

Attendees to 75th Anniversary Celebration of HAO (2015). Left to right, L. House (HAO), D. Gough (U. Cambridge, UK), and A. Skumanich (HAO).

11.2 Late Period (2017) – Solar Physics Handbook

In December of 2016, my wife Mary and I had the occasion to enjoy a reunion with our French friend and colleague Jean-Claude Vial (see Section 7.3) in San Francisco, during our customary winter stay in California. He mentioned a handbook that he and another colleague, Oddbjørn Engvold, were proposing to Elsevier. Because of my work spanning both solar and stellar astrophysics, he invited me to join them. The rationale for the project is quoted here from a January 2017 email to me from Jean-Claude: “The aim of this book is to present an up-to-date view of the entire field of solar physics for illustration of the significance of the Sun as a guide star in stellar astrophysics. The nature and physical processes, which were first revealed and studied in the Sun, shed clarifying lights on observed characteristics of various type stars. Some methods first used in solar physics were later applied in stellar physics. […] In the event that you would find this book project meaningful and worthwhile, would you consider joining us as co-editor?”

I responded on February 12, that the idea of such a book as he described seemed interesting and would be a resource to students and professionals alike and that I had decided to be a part of the effort. We prepared a proposal and sent it to the editor at Elsevier. It was refereed, and after some changes, accepted and ultimately published (Engvold, Vial, and Skumanich, 2019). The chapters and their authors are listed here:

The Sun as a Guide to Stellar Physics

Edited by Oddbjørn Engvold, Jean-Claude Vial, and Andrew Skumanich

• 1. Discoveries and Concepts: The Sun’s Role in Astrophysics

  Jack B. Zirker and Oddbjørn Engvold

• 2. Stellar and Solar Chromospheres and Attendant Phenomena

  Tom R. Ayres

• 3. The Sun’s Atmosphere

  Alexander I. Shapiro, Hardi Peter, and Sami K. Solanki

• 4. Helioseismic Inferences on the Internal Structure and Dynamics of the Sun

  Sarbani Basu and William J. Chaplin

Atmospheric Structure, Nonequilibrium Thermodynamics, and Magnetism

• 5.1. Spectroscopy and Atomic Physics

  Philip G. Judge

• 5.2. Models of Solar and Stellar Atmospheres

  Petr Heinzel

• 5.3. Spectropolarimetry and Magnetic Structures

  Kiyoshi Ichimoto

• 6. Coronal Magnetism as a Universal Phenomenon

  B. C. Low

• 7. Magnetohydrodynamics and Solar Dynamo Action

  E. R. Priest

• 8. Solar and Stellar Variability

  Marianne Faurobert

• 9. High-energy Solar Physics

  H. S. Hudson and A. L. MacKinnon

• 10. Space Weather at Earth and in Our Solar System

  Noé Lugaz

• 11. The Solar-Stellar Connection

  Gibor Basri

Instrumentation

• 12.1. Observations of the Sun From Space

  Alan Title

• 12.2. High-Resolution Ground-based Observations of the Sun

  Oddbjørn Engvold and Jack B. Zirker

• 13. Solar Data and Simulations

  Neil Hurlburt

• 14. Challenges and Prospects for the Future

  Jean-Claude Vial and Andrew Skumanich

11.3 The Last Hurrah (2019)

The last paper of my career (Skumanich, 2019) was published 10 years after my prime age (Figure 7). It arose from editing Tom Ayres’s contribution to the Handbook (see Section 11.2). Ayres, in discussing rotational spin-down, displayed an explicit interpolation formula for the torque derived from numerical solutions of a magnetized solar wind for different values of the wind parameters, magnetic field, and mass-loss rate (Matt et al., 2012).

Figure 7

A. Skumanich, 2009 (80 years).

The torque formula contained the product of two known monomial (power law) functions of each of the wind parameters. To obtain a solution for the temporal variation of the angular momentum, we must know how they vary with the rotation rate.

I contacted Sean Matt for his advice regarding coronal wind theory and rotational spin-down. I also mentioned that in the absence of a theory predicting the dependence of these parameters on the rotation rate, we could derive their dependence by imposing two observationally derived rules governing the spin-down. He thought my idea would make for an interesting paper, so he encouraged me to develop the argument. He provided insightful comments and valuable suggestions at various stages of my analysis.

The torque parameters are assumed to depend on the rotation rate by monomials with different exponential indices. In the case of the magnetic field, I had recourse to the relation implied in my popular 1972 paper (Skumanich, 1972). I found that the chromospheric calcium luminosity (which depends on the global magnetic field) was proportional to the rotation rate. By representing the chromospheric emission by an unknown power law of the field strength, i.e., with an unknown exponential index, we obtain an explicit form for the field factor in the torque.

For the mass-loss factor, I assumed that it varied with a monomial function of the field with an unknown index. This led the torque to have a cumulative index for its monomial dependence on rotation. By imposing the Skumanich law requirement we obtain a linear relationship between the two indices. If we have an observational value for either index, then we know the value of the other.

The index pair was calculated for twenty-five known activity correlations. The wide distribution of points along the solution line in the plot of one index against the other indicated a lack of consistency among the activity correlations. This is probably due to observational data errors or data reduction errors (e.g., incorrect baseline or zero-point corrections) or nonglobal sources.

The box-average of the distribution implies that the global surface magnetic field of a solar-like star scales linearly with rotation, whereas the mass-loss rate scales with the square of the global field strength, i.e., with the magnetic energy density in the system. The first conclusion reflects on the dynamo field and its emergence, whereas the latter has implications for the energetics involved in driving the wind.

12 Academic Engagement Through My Career

12.1 Faculty Courses

When I joined HAO, I was also appointed as an adjoint lecturer and finally professor in the CU Boulder graduate program of Astrogeophysics. Later I was appointed as professor in the Physics department.

At the behest of Mahinder Uberoi, Chair of the Aerospace Engineering Department at CU Boulder, who wanted his students to have some background in astrophysics, I devised a first year Modern Astrophysics course, Physics 580, open to students in Astrogeophysics, Physics, and Aerospace Engineering.

I was invited to lecture at the August 1976 Erice summer school (Sicily), where I discussed the role of escape probabilities in estimating “back of the envelope” source functions, often to check the convergence of more elaborate numerical calculations. The nature of radiation operators and their matrix representation was also presented.

Among the students auditing my lectures, there was Ester Antonucci, who was very perceptive and interactive with regard to the materials I presented. She was at the school to see if solar physics was a possible career path. Perhaps my lectures helped her decide that it was. We soon became friends and colleagues. I was pleased to recently learn that she was chosen by a Solar Physics panel as a memoirist and published her Memoirs last year (Antonucci, 2022).

12.2 Doctoral Candidates

I have benefited both scientifically and personally from working with graduate students as either their PhD advisor or a mentor. I was involved with the following students.

1) Loren Wallace Avery entered (1963) the Department of Astrogeophysics at CU Boulder as a PhD student and selected Lewis House as his thesis advisor. House suggested that Avery consider the use of the Monte Carlo method to determine the internal radiative field for a resonance line (two-level atom) in open cylindrical geometries for the self-excited case with and without external illumination (see Section 7.2 for a discussion of the formation of resonance lines).

With my knowledge of Monte Carlo methods, I was invited by House to participate in Avery’s PhD candidacy as a mentor and cochair on his thesis committee.

With our tutelage, Avery developed a Monte Carlo program to determine the source function for a level-atom for cylindrical geometries. The resulting thesis, titled “A Monte Carlo Calculation of Radiative Transfer in Cylinders with Application to Solar Spicules,” was accepted, and his degree was awarded in 1969.

Avery’s work on the scaling of the source function for a variety of sizes appears in Avery, House, and Skumanich (1969). His results with the Monte Carlo method confirmed my results derived by the standard deterministic method (Skumanich, 1967).

2) Gary J. Saliba was a Doctoral candidate at the Department of Applied Mathematics (University of Sydney, Australia) when I met him during my first appointment to the department in 1983.

His thesis title was “Non-LTE Scattering Resonance Polarization in Solar Spectral Lines.” The thesis was accepted, and his degree was awarded in 1986. I was involved only as a mentor sharing my understanding of the resonance scattering process and discussing various aspects his analysis.

3) Harrison P. Jones entered as a PhD student in the Department of Astrogeophysics and started to work with me on generalizing the Athay and Skumanich (1967) flux-divergence formalism for the statistical equilibrium of a two-level atomic model to a noniterative scheme applicable to multidimensional media based on the method of characteristics (Jones and Skumanich, 1968). The conversion of the resultant flux-divergence operators to their algebraic forms appears in Jones and Skumanich (1973).¹¹

Jones incorporated these results in his thesis “Line Formation in Multi-Dimen-sional Media,” which was accepted, and his degree was awarded in 1970. The thesis was published as NCAR Cooperative Thesis No. 21 by University of Colorado and the High Altitude Observatory, NCAR.

4) Philippe Lemaire (Associate Scientist, LPSP) came to HAO in the fall of 1970 to study the formation of resonance lines with me. He had designed and executed a balloon experiment to measure the solar Mg II h and k resonance lines at high spectral resolution. We derived one of the first models of the Mg II chromosphere (Lemaire and Skumanich, 1973). This work was ultimately included in his PhD thesis “Recherches sur l’émission de la chromosphère solaire dans les raies du magnésium ionisé,” which was accepted in June 1971. I was an invited member on his thesis committee in Paris.

5) Michel Hersé, a graduate student at the Université Pierre et Marie Curie (Paris VI), was a scientist assistant to Jacques E. Blamont at Service d’Aéronomie du Centre National de la Recherche Scientifique (CNRS). I had met Michel in 1973 during my sabbatical stay at LPSP. He was working on high resolution pictures of solar bright ‘grains’ in the ultraviolet’ as observed from a ballon. At this time I was working on such bright ‘grains’ as observed in the Ca II K emission line (Skumanich, Smythe, and Frazier, 1975). I was soon invited by Blamont, Président of Michel’s thesis panel, to participate as an advisor. Michel received his Docteur ès Sciences degree in 1976 with the thesis “Structure Fine du Soleil dans l’Ultra-Violet”. We have remained friends and colleagues since.

6) Benedict (Ben) Domenico became a graduate student in the Department of Physics and Astrophysics and started to work with me to convert the statistical population equation with flux-divergence operators for a two-level atomic model (Athay and Skumanich, 1967) to a multilevel model that allows for subordinate lines (nonground state transitions). The now vector population equations were solved by applying an inversion procedure based on quasi-linearization (Newton–Raphson; Skumanich, 1968).

This system was successfully tested and used to analyze the formation of Hydrogen Lyman $\alpha$ and Lyman $\beta$ resonance lines as well as Balmer H$\alpha$ for the solar atmosphere (Skumanich and Domenico, 1971).

Domenico adapted the new vector population algorithm in his thesis “On the Application of the Generalized Newton–Raphson Method to the Singly-Ionized Calcium Line Formation Problem In Model Stellar Atmospheres,” where he investigated the formation of the two resonance Ca II H and K lines, the Ca II infrared triplet (IRT), and the ground state of Ca III. In addition, he studied the implications of Olin Wilson’s observations of Ca II emission stars on the derived thermal models. The thesis was accepted, and his degree was awarded in 1972. The thesis was published as NCAR Cooperative Thesis No. 25 by University of Colorado and the High Altitude Observatory, NCAR.

7) Graham A. Murphy, another doctoral candidate at the Department of Applied Mathematics (University of Sydney, Australia), became a research assistant to the “Gang of Three” (Rees, Lites, Skumanich). One of his tasks was to develop a method to integrate the Stokes vector equation of transfer that would yield the emergent Stokes vector for a given model atmosphere (Lites et al., 1987, 1988). This became a part of his thesis “The Synthesis and Inversion of Stokes Spectral Profiles”. It was published as Cooperative Thesis No. 124 by the University of Sydney and the High Altitude Observatory, NCAR, in 1990.

8) Arturo López Ariste, a graduate student at l’Université Paris 7, Denis Diderot, was a research assistant to Meir Semel, his thesis advisor, at DASOP, Observatoire de Paris, Section de Meudon. I had met Arturo in 1995 when I joined Semel at DASOP. I was soon invited to join Ariste’s thesis committee as an advisor. He received his PhD degree in October 1999 with a thesis entitled “La Spectropolarimetrie en Astrophysique: Application au diagnostic des champs magnétiques solaires et stellaires”. After his degree, he came directly to HAO in January 2000 as a post-doc and then in 2001 was appointed Scientist I on the regular staff. Eventually, he took an appointment at CNRS, France. We continued our relationship as friends and colleagues.

13 Additional Collaborations Through My Career

13.1 Durney, Bernard (HAO)

The Durney and Skumanich (1968) paper was the first to study nonradial adiabatic oscillations of slowly rotating polytropes. Using a spherical harmonic expansion method, they identified structures that are now called gravitational gyroscopic waves or gyres. This work has been referred to as “pioneering” in recent publications (see, e.g., Townsend and Teitler, 2013).

13.2 Poland, Arthur (HAO)

To provide thermodynamic models for use in the analysis of the hydrogen Balmer H$\alpha$ observations in absorption or emission (and alternatively, lines from minority species), from spicules and prominences, Poland et al. (1971) calculated the ionization and excitation equilibrium for hydrogen in a variety of slab model atmospheres irradiated from both sides by photospheric, chromospheric, and coronal radiation fields. Such vertically standing structures were used as a 2D representation of their form. This was possible because the major photon escape route, which fixed the internal radiation field, is fixed by the slab thickness. The flux-derivative version of the statistical equilibrium equations proved to be efficient. The high citation number for this publication indicates that the models were very useful.

13.3 Liu, Sou-Yang (HAO)

Sou-Yang arrived at HAO in 1973 for a post-doc appointment with a high-resolution observational data set dealing with the fast time evolution of the Ca II K line. The observations refer to internetwork 1-arcsecond bright grains in a quiet region at the center of the disk.

The data indicated a wavelength localized brightness enhancement in the wings of the line that moved closer to line center (hence upward) and ended with a blue shifted emission in the Doppler core of the line (Liu, 1974).

Liu wanted to learn how to apply the HAO radiative resonance line inversion algorithms, so I instructed him and joined him to determine the thermal and dynamic model atmospheres needed to fit the spectral line data for several specific temporal line profiles (Liu and Skumanich, 1974). I proposed that we consider a rest thermal state for the upper photosphere and lower chromosphere that is then affected by a thermal and dynamical perturbation with height and width changing with time. The resulting inversion yielded a downward flow containing an upward moving wave. These results implied that the grains were an acoustic phenomenon.

This was demonstrated ultimately with a hydrodynamical simulation of the grains by Carlsson and Stein (1997), who showed that upward propagating acoustic waves convert nonlinearly into shocks in a downflowing stream and produced cell grains.

13.4 Jones, Harrison P. (NASA-GSFC)

Harry, who was now at the NASA Goddard-Southwest Solar Station, and I continued our association and considered the general effects of a finite geometrical structure on the internal radiative field density (Jones and Skumanich, 1980). We showed that the volume-averaged escape probabilities, which are closely related to radiation loss rates, scale with the mean chord length (thickness $D$ for a slab, radius $R$ for a cylinder), and thus the radiative losses are approximately the same in the two geometries. A limitation of this result is that it is only strictly valid when the absorption coefficient per unit length is constant with geometric position.

This geometric scaling law approach was confirmed by the 2D NLTE computations of Mihalas, Auer, and Mihalas (1978) with a two-level atom and later applied to prominences for the case of the H I, Ca II, and Mg II lines by Vial (1982).

13.5 Lean, Judith (CIRES)

While at CIRES,¹² Judith, with whom I had collaborated on the development of the solar Ca II K irradiance model (see Section 10.1), continued our collaboration, which included a variety of others from the atmospheric irradiance “cartel”. In addition to estimating the UV and Lyman $\alpha$ irradiances for terrestrial atmosphere analysis (Lean et al., 1982; Lean and Skumanich, 1983), our studies included the nature of the Maunder minimum (Lean, Skumanich, and White, 1992; Lean, White, and Skumanich, 1995), as well as the issue of a noncycling state (White et al., 1992). With the exception of Lean, White, and Skumanich (1995), all these works are fairly well cited into 2022.

13.6 Semel, Meir (DASOP)

In the fall of 1995, I took a leave of absence to join Meir Semel, whom I met during my early sojourns to France, at the Département d’Astronomie Solaire de l’Observatoire de Paris (DASOP) in Meudon to study the nature of the radial current densities (in the heliocentric coordinate frame, HCF) associated with magnetic field regions on the Sun as observed with the ASP polarimeter. These currents yield important boundary conditions for the outer solar atmosphere.

We note that the following methodology (Semel and Skumanich, 1998) requires us to assign the observer’s frame magnetic field to both a parity free line of sight (LOS) component and a parity-dependent (perhaps arbitrarily assigned) plane of the sky (POS) component. These are then mapped to the HCF and projected onto the solar tangent plan. Both fields yield a radial current via Ampère’s law. These currents are then treated separately and differently.

In the case of the radial POS-current, an algebraic method was used to modify the current into a sum of two quadratic forms with parity-dependent coefficients but whose absolute value was independent of any particular azimuth disambiguation (parity assignment) procedure. This yielded an invariant or “free” current, which was used to replace the parity-dependent form in other expressions.

The use of this construction yielded a smoother radial current than that derived from applying Ampère’s law to the POS field derived by using the disambiguation from HAO-AZAM.¹³ It was found that similar current discontinuities occur in both maps, confirming the correctness of the new construction, but whose location was found to be occasionally disambiguation dependent.

The final construction adds the parity-free radial current derived from the LOS field to our invariant current in such a way so as to yield two independent and disambiguated currents, one of which represents the smallest possible current allowed by the data. The selection of the minimum current disambiguates any particular magnetic map at any position on the solar disk.

We applied this minimum current disambiguation (Skumanich and Semel, 1996) to two HAO magnetic maps with different positions on the disk of the active region NOAA 7201 (see Section 9.2). The site of interest contained an emerging delta spot.

A comparison of our minimum current density map with that derived with Ampère’s law applied to the AZAM maps showed that either disambiguation method yielded quite similar results. Our construction yielded strong linearly extended features that appear on all maps near the neutral line in the plage and delta spot region. They are not of uniform amplitude along their length, i.e., they are often fragmented.

Note that long after its introduction, the value of our parity-free absolute-radial-current density method was well demonstrated by its use by Georgoulis¹⁴ as part of the initial current in an iterative disambiguation scheme. It significantly improved the ultimate accuracy of his disambiguation of synthetic data.

Considering the complex transformations involved in the paper, the unusual publication history of Semel and Skumanich (1998) was not surprising. It was submitted to A&A on July 11, 1996, accepted July 24, 1997, and published in 1998.

13.7 Leka, K. D. (NCAR-ASP)

Leka came to HAO in 1995 via her appointment as a Fellow in the Advanced Studies Program at NCAR. She expressed an interest in the ASP science program, so I became her scientific guide and tutor in the various subjects associated with Stokes polarimetry. Her first Stokes data reduction adventure was with the Stokes observations of the so-called quiet (“no activity”) region at the center of the solar disk. This led to the discovery of “HIFs”, transient small-scale horizontal “internetwork” fields as presented in Lites et al. (1996) (see discussion in Section 9.4).

I suggested to K. D. that we consider the temporal evolutions of fine structures and magnetic fields around pores. The consequence was Leka and Skumanich (1998), one of the first quantitative studies that set the basis for the development of the field (refer to discussion in Section 9.3).

Our final collaborative work was to derive a suitable method to study the nature of an active region’s helicity coefficient $\alpha$AR, which is defined by the ratio of the photospheric radial current (density) by the photospheric radial magnetic field. The result was the paper by Leka and Skumanich (1999). We tested three methods to calculate $\alpha$AR and discussed their limitations and examined the influence of data noise on their results. The discrepancies, agreements, and overall robustness of the different methods are also discussed. The paper was well received in the coronal activity field and has been frequently referenced up to this day.

13.8 López Ariste, Arturo (HAO)

In thinking about the new Stokes inversion method based on applying principal component (PC) analysis to a synthetic and observed Stokes data base, I realized that we could obtain an unbiased estimate of the physical content of each Stokes parameter profile by applying the principal component or orthogonal decomposition directly to the spectral covariance matrix (using wavelength pairs) for the Stokes intensity and net polarization spectral profiles, at any position. This matrix is then diagonalized by singular value decomposition to obtain the eigenfunctions (orthogonal components) and eigenvalues.

The PCA method expands the Stokes intensity and net polarization spectral profiles, at any position, in terms of these eigenfunctions. The expansion is ordered by the magnitude of the relevant eigenvalue from largest to smallest. The ordering represents a perturbation expansion.

Arturo volunteered to use his PC codes to help me with my project, which appeared in Skumanich and López Ariste (2002), one of the last papers of my career.

We found that the ordering allowed us to examine the physical content of the first few orders of the set of 40,000 profiles for each Stokes parameter for a solar active region (see Skumanich and Lites, 1987a, OSO-8 spot).

In particular, the analysis showed that the eigenvalues found by the PCA construction can be related to certain physical quantities, e.g., to the line-of-sight velocity and the magnetic Zeeman splitting for the unpolarized Stokes $I$ profiles, to the transverse field components for the linearly polarized Stokes $Q$, $U$, and the longitudinal magnetic flux for the $V$ profiles.

Note that this quick-look method easily found that the bright quiet-Sun points (grains) have an upflow signature, whereas the dark regions have a downflow one, in good agreement with that derived by Liu and Skumanich (1974).

The most recent reference to this quick-look PCA-spectral method was a publication by Lehmann and Donati (2022). They adopted the method for the determination of stellar magnetic fields from a time series of observations of the stellar circular or Stokes $V$ polarization profiles.

They concluded that the method allows an easy first glance at key parameters of the stellar magnetic field topology before more advanced techniques like Zeeman-Doppler-Imaging (ZDI) are used to reconstruct a full map of the surface magnetic field. It can be conveniently applied to a large number of stellar targets.

14 The 50th Year Anniversary

The influence of my 1972 paper (Skumanich, 1972, hereafter Sku72), which studied the age dependence of surface rotation state (magnetic wind braking) and chromospheric luminosity (Ca II emission, signature of magnetic activity) has persisted since its publication.

It has spawned many studies relating rotation periods to magnetic activity for other mass stars, e.g., by Noyes, Weiss, and Vaughan (1984) and many other studies. A new field of gyrochronology was inspired, which uses surface rotation rates for the first time as an age diagnostic (Lachaume et al., 1999). The advent of the IUE satellite data (c. 1980) allowed one to search for Sku72-like relations in the UV and X-ray domains. Likewise with the Kepler satellite (c. 2016), the studies of young and intermediate seismic ages confirm the Sku72 results for solar analogs in general except for stars older than 2 Gy.

The current influence is well illustrated by the range of papers presented at a 2022 meeting to celebrate the 50th anniversary of the Sku72 publication date (https://skumanich.wdrc.org). The celebration was initiated by Travis Metcalfe and colleagues and hosted as an in-person conference organized around the major science themes of the Sku72 paper. The announcement notes that “the paper is one of the most impactful publications in the history of HAO, with more than 1500 citations to date.”

The Festschrift abstracts for all of the talks and posters presented at the meeting and the presentations themselves have been archived for posterity.¹⁵ My own presentation “The discovery path of the inverse square root of age relations for solar-type stars”, with explanatory slides, can also be downloaded there.¹⁶

Holly Gilbert, the Director of HAO, opened the meeting with a welcome and wishes for a fruitful celebration. After a few words commenting on my long association with the observatory, she invited me to the front to be presented with a framed poster by the American Astronomical Society congratulating me “On the fiftieth anniversary of the publication […] in the Astrophysical Journal. This famous and highly cited paper has had a profound influence on the development of stellar astrophysics in the intervening years.” I was surprised and quite pleased with the AAS recognition.

At the celebration, I was excited to see and met some of the young people contributing to the advancement of solar-stellar astrophysics. After 20 or so years of retirement, I was also overcome with emotion on meeting many dear former colleagues and friends again. “Those were the days my friends.”

15 Personal Reflections

My scientific career path has been an eclectic one. With Martin Schwarzschild’s tutelage, I developed an interest in mathematical analysis and model building, with Alexander Nikolaevich Vyssotsky, an interest in the relationships or correlations between observations. This was, to some extent, the fulfillment of my childhood curiosity and need to understand. Published work of, and interaction with, peers and collaborators often opened other possible paths to explore. You will find such examples in this memoire. I had no specific research plan when I was ready to enter the work force. My direction was driven by the available opportunities.

One feature of my early scientific career was my involvement at LASL to do physics experiments with high-energy density sources as provided by nuclear weapons. To have witnessed the detonations of atomic (fission) and thermonuclear (fusion) sources was an awesome experience. The fission explosions turned night (predawn) into blue daytime that lasted a fair number of heart beats, whereas the fusion explosion “sunrises” seemed to last fearfully longer. Once I left LASL, I would have an infrequent dream that while walking along a street, there would be a flash of light above me, and recognizing the BOMB, I would drop to the curb, cover my head, and after the shock passed, would rise and start brushing dust off my clothes … as I imagined the Japanese fishermen did during the first H-bomb event.

On the occasion of a stellar Ca II observing session at Mt Wilson with Olin Wilson, I had a night time opportunity to climb (furtively) to the top of the 100-inch dome and behold the luminous carpet of Los Angeles below me. In my fancy the curved dome became Earth, and I was Atlas (in reverse). It was an unforgettable image. Heights were never a problem for me. At Las Alamos, I worked on a 500-ft tall open tower in the case of the neutron PINEX experiment.

My interest in Art and History, which developed in my youthful years, was augmented by visits to museums and historical sights in different US cities, where I attended professional conferences. Conferences abroad proved to be even more enriching, given the birth of the Renaissance in Europe. Thus my 1-year sabbatical at LPSP (1973 – 1974) in the environs of Paris and my semester long stays at the University of Sydney (1984, 1986) and University of Florence (1988, 1993) were fulfilling. During our year-long stay in Versailles, my family and I were able to do the fabled “Grand Tour,” which was the apex of a humanistic education.

I have had extraordinary luck and unexpected changes – good breaks – in my life line at various stages of my life. My greatest satisfaction was to be able to work with outstanding colleagues, post-docs, grad students, and scientific staff assistants. My education was possible through the unexpected assistance from my “village”, personal labor, librarians, and teachers who cared, college and university endowments, and, ultimately, through government expense. I am grateful for all of the above and very thankful to have lived in a society and at a time when the pursuit of knowledge was considered a worthy direction of community resources.