Q: Among your earliest papers is a study of forbidden line emission from young stellar objects.

A: By the early—mid 1980s, optical forbidden lines such as [O I], [S II] and [N II] had become powerful tracers of the collimated outflows associated with young stars, linking optical jets and HH objects to the much larger bipolar molecular outflows mapped in millimetre-wave CO emission. The low optical depths of these forbidden lines made them particularly valuable kinematic diagnostics, since their profiles were unaffected by the self-absorption that can complicate the interpretation of strong per-mit-ted lines such as H$\alpha$. Imaging and spectroscopy revealed chains of knots and measurable proper motions, de-mon-strat-ing that HH structures traced rapidly moving ejecta and providing estimates of velocities, dynamical ages and flow geometry.

My early work focused on the geometry of active star-forming regions and their HH/GGD nebulosities. Deep optical imaging of regions such as NGC 7129 helped es-tab-lish which GGD objects were shock-excited HH structures, revealed previously unknown emission and HH objects, and related them to the molecular-outflow environment. The broader aim was to understand how apparently sep-a-rate HH condensations, optical jets and large-scale molecular outflows fitted into a common flow geometry.

By the late 1980s, high-resolution forbidden-line spec-tros-co-py had revealed high-velocity emission in many TTSs, in-clud-ing sources with no visible extended jet. Solf, Mundt and collaborators had shown that, where jets were re-solved, this component could be traced back toward the star. This raised an important question: did apparently “jet-less” TTSs actually harbour unresolved microjets? In the early 1990s, while working at McMaster University in Canada, I had access to the experimental High-Resolution Camera at CFHT. Its subarcsecond imaging capabilities allowed us to reveal the CW Tau microjet, supporting the identification of high-velocity forbidden-line emission with compact jets even when no prominent large-scale optical jet was apparent. Later, HST observations established that microjets are common among classical TTSs. The emerging picture was that jets are much more ubiquitous than large-scale optical jets would suggest.

These observations naturally led to the question of how jets and forbidden-line emission were connected to disk ac-cre-tion. Together with Ralph Pudritz, we explored this in the framework of magnetohydrodynamic disk winds. We proposed that a hydromagnetic wind launched from a Kep-ler-ian accretion disk could be magnetically recollimated, producing a focal region where an MHD shock forms. The shocked gas reaches densities and velocities appropriate for the observed forbidden-line emission—up to about 250 km s$^{-1}$ in the fiducial model—providing a physical connection between disk accretion, magnetically driven mass loss and the observed jet.

Q: Using high-density gas ultraviolet (UV) tracers you have studied jet collimation in RY~Tau and RU~Lup. What were the results?

A: By 1992, I had returned to Spain and was working at the European Observatory of the International Ultraviolet Explorer at the Villafranca Satellite Tracking Sta-tion, now ESA’s European Space Astronomy Centre. This gave me the opportunity to extend optically thin emission-line diagnostics into the UV, probing the hotter and denser plasma close to the base of the outflow while avoiding the severe self-absorption affecting many per-mit-ted UV resonance lines. Particularly powerful were the semiforbidden C III] 1908 \AA\ and Si III] 1892 \AA\ transitions. Their ratio is sensitive to very high electron den-si-ties, while their much weaker forbidden components pro-vide sensitivity at lower densities. Together, these tran-si-tions discriminate densities from approximately $10^3$ to $10^{13}$ cm$^{-3}$, particularly valuable in an unresolved T Tauri environment where low-density jet material and extremely dense plasma may coexist.

The UV observations revealed very broad C III] and Si III] emission associated with the outflow, tracing plasma at $T_e \sim 10^5$ K and densities of $10^9$—$10^{11}$ cm$^{-3}$ very close to the star. The large line widths suggested that the emission could not arise solely in the narrow, collimated optical jet, and were consistent with a bow-shaped shock or similarly complex velocity field at its base. More importantly, the unexpectedly high densities placed a stringent constraint on jet-launching models: they had to explain not only acceleration and collimation, but also this hot, extremely dense component close to the star.

These results opened two related theoretical questions. With Constantino Ferro-Font\‘an and later on Brigitta von Rekowski, I investigated whether the hot, dense UV emitting plasma could be produced by magnetically driven outflows from the inner disk and, ultimately, the star—disk shear layer. We confronted dynamical models directly with observables such as the centroid and width of the Si III] profiles. The calculations showed how strongly the predicted profiles de-pend on launching geometry, magnetic configuration and viewing angle, and explored whether non-stationary star—disk winds could reproduce the observed connection between accretion and ejection.

In parallel, the high densities and temperatures raised another possibility: how much of the UV semiforbidden emission might instead be produced by accretion shocks? Distinguishing between accretion, stellar/magnetospheric plasma and the base of the outflow therefore became central to interpreting these lines. Rather than simple “jet tracers”, C III] and Si III] had become probes of the physical interface where accretion, magnetic activity and mass ejection meet.

Q: One of the instruments you have used on HST is STIS, in particular to study the very active star RW~Aur~A.

A: Indeed. RW Aur was a particularly intriguing case, where the C III] and Si III] profiles revealed extremely dense, hot plasma very close to the star. From the UV spectroscopy, we inferred an ionized belt-like structure confined to only a few stellar radii,
far below the angular resolution attainable at the time. Two decades later, near-infrared interferometry with VLTI/GRAVITY has begun to probe these same spatial scales in T Tauri systems. In CI Tau, for instance, the Br$\gamma$-emitting region has been spatially resolved to a characteristic scale of only $\sim 5\,R_\star$, comparable to the magnetospheric truncation radius.
Thus, structures that could once only be inferred spectroscopically from their density, temperature and velocity signatures are now becoming accessible to direct interferometric constraints at the star—disk interface.

Q: You have studied accretion shocks in TTSs using semiforbidden UV line ratios. What was the technique and what did you learn?

A: In 1990, Simon et al. reported rotationally modulated variations of the UV flux from BP Tau observed with IUE. At the time, hot and cool spots were being extensively studied through optical monitoring, notably by Bouvier and collaborators. In the optical, however, hot spots could be interpreted as relatively modest temperature enhancements, also discussed in connection with the complex surface magnetism of young stars. The IUE observations suggested something different: the UV modulation required a substantially hotter component than was apparent from the optical continuum.

I remember reading the Simon et al. paper shortly after it appeared, while visiting STScI, and finding the result particularly exciting. Jim Pringle happened to ask a group of us in the library what we considered the most interesting recent development in the field, and my immediate answer was the BP Tau result. The UV seemed to be revealing a much more energetic phenomenon than the optical hot-spot picture suggested. The timing was remarkable: less than a year later, K\“onigl formulated the magnetospheric accretion framework for classical TTSs, in which the stellar magnetic field truncates the inner disk and channels material toward the surface at close to free-fall velocity, producing a strong accretion shock.

We subsequently monitored BP Tau and DI Cep with IUE and found similar behaviour. Using spectral tracers spanning a broad range of ionization, from C II to C IV, we asked whether the UV spectrum could constrain the physical conditions of this hot plasma. This eventually led to our collaboration with Sergei Lamzin, whom I invited to the Universidad Complutense de Madrid to confront his detailed accretion-shock calculations with the observed UV spectra of TTSs.

We concentrated on the semiforbidden C III], Si III] and O III] lines. The calculations showed that these lines are produced predominantly in the pre-shock gas, photoionized by high-energy radiation from the hot post-shock plasma. Being optically thin and relatively close in wavelength, their ratios provide unusually clean diagnostics of the accretion flow. Comparison with the IUE observations constrained densities and infall velocities consistent with magnetospheric accretion columns and material falling from the inner disk onto the star. The UV spectrum had therefore become more than evidence for a hot spot—it provided a quantitative probe of the gas immediately upstream of the accretion shock.

Q: AK Sco has been a particular focus of your work. What has attracted you to this star?

A: AK Sco attracted my attention for a rather different reason. It is an extraordinary natural laboratory: two nearly identical pre-main-sequence stars on an eccentric orbit, with a separation shrinking from about 30 stellar radii to only 11 at periastron. The orbit therefore provides a clock for the accretion process. Our numerical simulations showed that the binary acts rather like a gravitational piston: material is drawn from the circumbinary disk into the central cavity, forming streams that interact as the stars approach periastron, lose angular momentum and trigger enhanced accretion.

HST observations provided a striking signature of these streams. Near periastron, the fluorescent H$_2$ emission from the disk decreased for several hours while the stellar UV emission did not, suggesting that infalling gas temporarily blocked the stellar Ly$\alpha$ radiation, casting a shadow over part of the molecular disk. Another unexpected signature was a coherent UV oscillation, with a period of about 790 s lasting nearly two hours around periastron, which we interpreted as a response to the sudden release of accretion energy.

The UV monitoring also showed that periodic forcing does not necessarily produce a periodic plasma response. Different ionization stages have different light curves: high-ionization tracers such as N V, Si IV and C IV can brighten strongly, while C II and O I behave differently, and the Si III]/C III] ratio reveals substantial density changes. Successive periastron passages are not identical either. AK Sco thus taught us that the binary orbit provides the clock, but the accretion flow retains its own dynamics and memory.

Q: You have made several compilations of UV spectra obtained by the IUE satellite.

A: Another aspect of my work has been to make UV observations more accessible to the community. During the final years of IUE, I compiled, with Merche Franqueira, the ULDA Access Guide to TTSs, bringing together observations, spectra and basic information for the TTSs observed by the mission. Later, with Angel Robles, we produced the first INES Access Guide, devoted to Herbig—Haro objects. INES — the IUE Newly Extracted Spectra archive — incorporated improved extraction procedures developed at ESA, particularly valuable for emission-line spectra, providing a more homogeneous and higher-quality data set.

My involvement in UV facilities continued with Spektr-UF/World Space Observatory-Ultraviolet, the 1.7-m UV space telescope led by Russia, for which I served as Spanish Principal Investigator, and more recently through my involvement in the scientific development of NASA’s project Habitable Worlds Observatory. In parallel, we have compiled a catalogue connecting UV sources observed by IUE, HST, GALEX, Swift and XMM-Newton, soon to be released through the Joint Center for Ultraviolet Astronomy at the Universidad Complutense de Madrid.

There is a certain continuity in all of this: from helping astronomers navigate the archive of a single UV observatory to connecting the legacy of several generations of missions and helping prepare the observatories of the future. For me, preserving what we have learned from IUE and HST while ensuring powerful UV capabilities for HWO is an important part of keeping UV astronomy scientifically alive.

Q: In a recent paper you have uncovered a new group of TTSs in the well-studied Taurus-Auriga clouds. How did you do that?

A: Another long-term project has been the search for the more dispersed population of young stars around the Taurus-Auriga molecular complex. The idea was simple: accreting TTSs should reveal themselves through both a UV excess, associated with the accretion shock, and an infrared excess from circumstellar material. Using GALEX and 2MASS, we explored combinations of UV and near-infrared colours. The FUV$-$NUV versus $J-K$ diagram proved particularly effective: from more than 160,000 UV sources, we selected 63 new T Tauri candidates while recovering all 31 previously known TTSs in the surveyed area, giving us confidence that the method was sensitive to the young stellar population.

Gaia subsequently transformed the experiment. Its accurate parallaxes and proper motions allowed us to determine which candidates shared the location and kinematics of the Taurus-Auriga population. Using well-established TTSs as a reference, we found that the association separates naturally into two kinematic populations at average distances of about 130 and 160 pc. An independent analysis by Nayak et al. (2023) found a remarkably similar bimodal distance distribution, reinforcing the picture of Taurus-Auriga as a structured three-dimensional and kinematic population rather than a single association at one distance.

Applying these Gaia constraints dramatically reduced the original sample: only about 16\% of our 63 candidates could be considered reliable members of the association.

Q: You have worked on magnetic fields over a remarkable range of scales, from molecular clouds to T Tauri stars and their immediate environments. Looking back, do you see a common thread connecting these different aspects of your research?

A: Yes, although these were initially parallel lines of research rather than a single programme. Alongside my work on T Tauri stars, accretion and outflows, I was investigating the role of magnetic fields on much larger scales: how the Parker–Jeans instability could contribute to the formation of molecular complexes in Taurus, how MHD waves propagate through and fragment molecular material, and what observational signatures these processes might leave—from structures extending out of the plane of the nearly edge-on galaxy NGC 4013 to polarization studies of the Galactic magnetic field toward high-latitude clouds and high-velocity gas. My direct involvement in this line of research decreased after 1998, but the interest remained and resurfaced later on with the study of dust charging under various relevant UV radiation fields and its impact on MHD waves propagation. In retrospect, what connects these studies with my work on young stars is the physics of magnetized matter across scales: how magnetic fields redistribute energy and momentum and influence the organization of the interstellar medium, the fragmentation of molecular gas, and, eventually, the accretion and outflow processes that accompany the birth of a star.