See Liu et al. (2024) for further details.

Studying the physics and chemistry of nearby galaxies allows us to understand the interplay between the many processes that govern galaxy evolution across a much wider range of physical environments and extremes than in our own Galaxy, but at resolutions as close to Galactic as possible. Star formation theories based on Galactic observations cannot convey the complexity or diversity of environments in which other galaxies form and evolve (e.g. with different metallicities, total masses, positions within and size of galaxy cluster, etc.).

The closest companions to the Milky Way, the Magellanic Clouds (MCs), offer a unique perspective on low-metallicity environments with near galactic scale resolutions. From current generation facilities, we can quantify their physics and chemistry on sub-parsec scales and understand, in general terms, the relative contributions of the different components of their ISMs. But observations of [CII], for instance, suggest much of the molecular gas in the Large Magellanic Cloud (LMC) is CO-dark: we are not detecting it using standard sub-mm methods (Chevance et al., 2020). Understanding the molecular gas content of the LMC then leads us to understand how much star formation can be expected, and whether that is driven by the contents of the observed giant molecular clouds, or the ISM surrounding them. Current facilities do not have the sensitivity, FoV and resolution to explore the CO-dark gas in the MCs. With the ability to observe in the sub-mm, AtLAST will be capable of observing [CI] lines (at rest-frame frequencies of 492 and 809 GHz) across the MCs that trace this CO-dark gas.

Quantifying the dust (in terms of mass, density and temperature) as a function of the environment, energy injection and metallicity tells us about dust properties in extreme environments. There are discrepancies between theoretical expectations of dust properties at low metallicity and those seen in the MCs (Bot et al., 2010; Galliano et al., 2011). Resolving this discrepancy pushes us closer towards a complete model of cooling, heating, star formation and dust grain growth in different metallicity environments. Understanding low-metallicity environments in nearby galaxies allows us to improve our understanding of high-redshift galaxies, which are typically also lower metallicities.

Pushing out to larger distances, the physics and chemistry of the ISM in nearby galaxies tells us about the evolution of that galaxy, and by deriving both the star formation rate and gas surface density in nearby galaxies (the Kennicutt-Schmidt relation; Schmidt, 1959; Kennicutt, 1989), we understand different types of galaxies in the context of the general population of galaxies. These types of relations break down on roughly kpc scales, and understanding that deviation from the KS relation puts interesting constraints on the timescales of star formation. Through observations of chemical tracers tracing different densities and in multiple transitions, we can further explore the parameter space in the KS relation, probing to higher and higher gas densities. To do this well, in multiple chemical tracers, across statistically significant samples of galaxies, requires highly multiplexed spectroscopic instruments with high sensitivity and resolution.

Threading through these themes is the ability to measure the polarised emission of these galaxies to quantify how magnetic fields influence the evolution of galaxies. Magnetic fields regulate star formation, influence the dynamics of the ISM and are important for a number of other processes, but observing them consistently remains a significant challenge. Small samples of polarimetric observations of dust in nearby galaxies have shown a relationships between the fields in the molecular gas and star forming regions, but measurements are difficult because of either the small fields of view of sensitive interferometers like ALMA, or the relative insensitivity / low resolution of current generation single-dish facilities where observing polarisation fractions to uncertainties of a few percent requires a significant investment of telescope time.

Creating a polarisation survey of 100 nearby galaxies, reaching resolutions of 20-300 pc to an rms noise level of 16 $\mu$Jy/beam will enable detections of 1% polarisation in the dust emission. This is 100 times deeper than possible with current facilities while at the same time offering significantly higher spatial resolution (by a factor of $\sim$3 in each direction) which means more than 1000 independent polarisation vectors can be measured. While impossible with current facilities, observations like this could be done with AtLAST in 18-50 hrs per galaxy.

See Klaassen et al. (2024) for further details.

From the vantage point of our Solar System being in the plane of the Milky Way, we are able to probe the cycle of interstellar matter in a detailed way that is impossible to capture for external galaxies. We can view the assembly of clouds on the largest scales down to dense cloud core scales and protostars, then follow the subsequent dispersal of clouds by the feedback of the formed stars and the eventual enrichment from the death throes of stars at the end of their lives. We can delve into the minutiae, but to put those details into their galactic context requires large area surveys. With a large aperture, wide-field of view (sub)millimeter single dish telescope, we can explore and quantify the spatial distribution of gas and dust, the kinematics of the gas, the gas chemical composition (and its evolution) and magnetic fields from the scales of giant molecular clouds down to individual nearby, faint debris discs, analogous to our own Kuiper belt.

Large scale surveys of the gas and dust in the Galactic Plane unlock a multitude of science questions. How do cool Giant Molecular Clouds (GMCs) form within the largest scale structures in the ISM of our galaxy? How do those GMCs then continue to fragment into filaments and star forming clumps and cores? What then triggers or inhibits star formation in those clumps and cores? What is the interplay between magnetic fields and stellar feedback on star-formation? From these questions, we can see that the chemistry, dynamics, and recycling of material are all interrelated in our understanding of the ecology of our Galaxy and tie directly to the gas, dust and magnetic field properties of the ISM. The ISM is turbulent and magnetized and to describe the evolution of the ISM and star and planet formation we need to probe the density distribution of matter as well as the velocity structure (to probe the dynamics) and the magnetic fields. AtLAST will be able to study simultaneously the evolution of dust and gas from large scales to small (from $\sim$100 pc to $<$0.1 pc) and from low densities to high (1 to 10⁶ cm^-3) spanning orders of magnitude in scale and densities.

A polarised dust continuum survey of the Galactic plane will allow us to constrain, in a statistical and holistic way, the relationship between gravity and magnetic fields in the collapse and support of structures, from the scales of entire GMCs down to those of star-forming cores. This will allow us to understand the flow of material through and between states, and the nature of star forming cores throughout the plane – as we will reach from filament to individual core scales (0.1 pc) resolution on the far side of the Galaxy at the highest frequencies obtainable with AtLAST.

Spectroscopic surveys of the Galactic plane add to the picture through chemistry, turbulent velocity dispersions, temperatures and heating and cooling rates of the gas. Wide bandwidths will be essential for observing multiple species and transitions simultaneously. Here we can use a similar ‘wedding cake’ approach to the Distant Universe studies above. In a wedding cake scenario, we begin with larger area surveys (i.e the whole Galactic Plane) to sensitivities sufficient to detect common molecular and ionised gas transitions (CO, CS, N₂H⁺, HCO+, SiO, radio recombination lines, etc), and then follow up with deeper / more sensitive observations in a smaller tier towards regions deemed chemically complex (i.e. individual star forming regions) in order to derive a census of complex organic molecules (defined as carbon bearing species with at least six atoms). Only following up with highly sensitive observations in regions known to be chemically complex will save hundreds of hours of observing time.

With the resolution, sensitivity to the full range of spatial scales, and confusion limits enabled by a 50m dish antenna, we can study the most common mode of star formation: low mass stars in clustered environments which are also in the process of forming high mass ones. With current technologies, we can study the details of isolated core collapse in nearby low mass star forming regions, but with a 50m class telescope, the same resolution achieved in nearby (100-200 pc) regions will be possible in the richer and more complicated high-mass star forming regions within 2 kpc - where a full Initial Mass functions (IMF) worth of low and intermediate mass stars are also forming but cannot currently be detected or disentangled from their brighter companions.

With large format cameras, entire nearby star forming regions can captured in a single pointing, taking a single ‘snap-shot’. This makes it possible to not only trace the dust and gas in the star forming cores that then settles into protoplanetary discs and as the stars emerge and finally evolve into debris discs as the gas is dispersed from the system, but to go back to these regions in a set cadence to monitor for transient behaviour (i.e. accretion flares). With (sub-)mm observations of the early evolution of these planetary systems, the masses, morphologies, grain-size distributions and dispersal mechanisms can be tracked (Lovell et al., 2021). With repeated observations of these star-forming regions we can also expect to find large numbers of variable sources, informing us of the star formation process (see also Section 2.6). The mass assembly of protostars is known to be episodic and for the youngest, most deeply embedded sources, during the main mass assembly stage, brightness variability in the far-IR and sub-mm provide the only direct measure of both the timescale and amplitude of these accretion episodes - only AtLAST has the ability to monitor 1000s of sources with monthly cadence over decade timescales.

As the circumstellar discs evolve into debris discs, their dust masses greatly decrease, meaning that only the brightest debris discs are detected (e.g. Wyatt, 2008). Observing discs as faint as the Solar System’s Kuiper belt, requires observing the nearest stars (within around 20 pc) down to sensitivities on the order of 10s of $\mu$Jy/beam. Given their proximity, the discs can easily extend well beyond the maximum resolvable scale of interferometers like ALMA and so only a large single-dish telescope is capable of detecting them (see Figure 7 for an example). Observations in multiple bands then allow the characterisation of their dust properties (e.g. Löhne, 2020).

See Cordiner et al. (2024) for further details.

Understanding our place in the Universe requires understanding the formation and evolution of our own Solar System. (Sub-)mm spectroscopy of giant planets teaches us about the chemistry and dynamics of their atmospheres, from which we obtain information about their origins and temporal evolution (Encrenaz et al., 2005). From spectroscopy of terrestrial planets, moons and comets, we can learn about their habitability and the availability (and origins) of chemical ingredients for life in the Solar System. Deeper insights into our own planetary system are also essential in our quest to understand the properties of exoplanets, and the relationship of our solar system to other planetary systems.

Observations with current (sub-)mm facilities have produced ground-breaking results, including measuring the molecular inventories and dynamical motions of planetary and cometary atmospheres. However, current observatories suffer from limited dynamic range, spatial and temporal coverage, and sensitivity (Cordiner et al., 2024). Instantaneous map** over the entire face of a body is critical for detailed studies of the rapidly rotating and evolving atmospheres of solar system bodies (Lellouch et al., 2019; Cordiner et al., 2020).

Interferometric observations are not ideally suited to solar system studies due to the spatial filtering of the large angular scales that are a common feature of planets and comets. Furthermore, the repositioning of antennas (and consequent variation in $uv$ coverage) hinders our ability to reliably and consistently study and map temporally variable phenomena such as storms, outbursts or seasonal changes.

In order to advance our knowledge of planetary atmospheres, a large aperture, single-dish telescope is necessary. This will allow us to measure spatial and temporal variations in temperature structure and molecular abundances throughout the troposphere and stratosphere, which can be used to investigate the circulation, dynamics and composition of the atmosphere. A sensitive single-dish facility will enable short-term and seasonal variations to be studied, and facilitate the discovery of new, trace molecular species and their distribution. A 50 m dish is essential here so that Venus, Mars, Jupiter and Saturn can be resolved throughout their orbits and across all wavelengths (see Figure 8). With this, Uranus and Neptune can also be marginally resolved at the shortest wavelengths. In order to conduct sensitive searches for new spectral lines and isotopologues of interest, a spectroscopic dynamic range of $\sim 10^{5}$ is needed due to the continuum brightness of the planets — to realise this, close attention should be paid to achieving the flattest possible spectral bandpass through the careful choice of optical design, receiver components and optimal calibration strategies.

Moons of the giant planets also offer compelling targets in the search for life elsewhere in the solar system. Titan has a thick, carbon and nitrogen-rich atmosphere that can be studied in a similar way to the terrestrial planets (Nixon, 2024). Deep spectral line observations will enable the discovery of new organic molecules (Cordiner et al., 2015; Palmer et al., 2017; Thelen et al., 2020), as well as isotopologues of known molecules that improve our understanding of the long-term physico-chemical evolution of Titan’s atmosphere (Mandt et al., 2009, 2014; Nomura et al., 2023). Many of the icy moons of the outer planets are thought to have subsurface oceans, which can result in cryovolcanism (Nimmo & Pappalardo, 2016). Spectral studies of the resulting plumes provide the perfect opportunity to study the compositions of the subsurface oceans, in particular searching for organic molecules that can reveal insights into the habitability of these oceans. The Cassini mission to Saturn discovered such plumes around Enceladus (Hansen et al., 2006; Porco et al., 2006; Spencer et al., 2018). Although it had a mass spectrometer, this lacked the resolution to differentiate between certain molecules (Waite et al., 2006, 2009). Remote observations from ground-based facilities present the opportunity to overcome this. However, to do so requires a high enough angular resolution (specifically $<8^{\prime\prime}$ at 1.1 mm) to avoid interference from Saturn and its rings, and with a single-dish to avoid any interferometric artifacts. This would enable both confirmation of the Cassini result and continued monitoring for temporal changes in the plume composition.

Comets are thought to be largely unchanged since they accreted very early in the Solar System’s history (Mumma & Charnley, 2011) and thus provide unique information on the physical and chemical conditions at the time of planet formation (Bockelée-Morvan et al., 2004). The ices and organics contained within comets may well have played a key role in setting the stage for life on Earth due to their delivery via impacts. Studying comets is also useful for understanding how our Solar System relates to other planetary systems and to the chemistry of the ISM (see Section 2.3). A huge variety of different ice and organic lines are detectable in the (sub-)mm, necessitating a 32-64 GHz bandwidth to be able to observe them simultaneously and thus avoiding changing tunings during time-critical observations. Of particular interest is the deuterium to hydrogen ratio that varies across Solar System bodies and can tell us about the history of water in the Solar System, particularly when compared to observations of protoplanetary discs (Willacy & Woods, 2009; Albertsson et al., 2014; Cleeves et al., 2014). The strongest HDO line observable from ground-based (sub-)mm observatories is the 894 GHz line, demonstrating the importance of reaching these high frequencies. Cometary comae are often extended over a few arcminutes, making studies of the extended coma impossible with interferometry. In order to resolve the complex kinematical structure of the coma, a spectral resolution of $\sim 0.1\,\rm km\,s^{-1}$ is necessary.

See Wedemeyer et al. (2024) for further details.

The Sun is the only main sequence star we can spatially resolve, and as such, not only does it allow us to understand the engine that drives our solar system, but gives us insights into other stars that would be impossible to observe in such detail. At (sub-)mm wavelengths, the emission from the Sun is dominated by the magnetised plasma in the chromosphere, which we observe as (polarised) thermal continuum emission. Different layers of the chromosphere dominate at different wavelengths, hence observations across the (sub-)mm regime probe the three-dimensional structure of the Sun’s atmosphere. Observations of the chromosphere are important for a number of reasons, including understanding the ‘coronal heating problem’ – understanding why the corona is much hotter than the solar atmospheric layers below it (including the photosphere and chromosphere), solar flares, coronal mass ejections (CMEs), prominences and the solar wind. By disentangling the importance of different chromopsheric processes (such as magnetic reconnection and various forms of wave heating), we start to unpick which types of physical processes are dominant under which conditions (i.e. active or quiet Sun) and on what size scales.

Solar flares, prominences and the solar wind all play a vital role in space weather. Predicting them and understanding what drives them will help shape space weather forecasting, but doing that requires detailed long term observations of the Sun in polarised light. Because these transient phenomena are related to the magnetic field reconnecting and reconfiguring on sometimes explosive timescales, simultaneous observing of the full face of the Sun is required, with repeated observations to take in variations across the solar rotation period (around a month) and solar cycle (11 years).

Space weather directly impacts human society due to the effect it has on power grids, satellite infrastructure and human spaceflight. The more we understand space weather, the better we can mitigate these effects. Understanding space weather in our own Solar System also allows us to better understand space weather in extrasolar planetary systems and how it impacts habitability in those systems, particularly in cases where flares and CMEs are considerably stronger than those released from the Sun (e.g. Maehara et al., 2012).

Current generation facilities often either do not have the spatial resolution to resolve structures like granules (Kundu, 1959; Trottet et al., 2011; Loukitcheva et al., 2014) or, as is the case with ALMA and other interferometers, filter out much of the full disc scale (30^′) emission. For a dynamical structure like the Sun, temporal resolution is just as vital, with chromospheric evolution sometimes happening on timescales shorter than a second (Kontar et al., 2018).

See Orlowski-Scherer et al. (2024) for further details.

As can already be seen in some of the science cases presented above, another dimension has emerged as important in understanding the universe as it evolves around us: time. The study of transient and variable phenomena is still in its early stages in the sub-mm regime, with the first papers on the subject published less than 10 yrs ago, but already we are seeing evidence for variability in objects from AGN and Gamma-Ray Bursts (GRBs), through to supernovae and forming stars and the atmospheres of the Sun and planets. Transient and variability studies can help shape theoretical predictions for a wide range of astrophysical phenomena, but care must be taken to ensure the proper types of observations (e.g montoring, triggered ToOs and serendipitous discoveries), and on the right cadence, for time domain astronomy to come to prominence.

Some transient detections have come from the sub-mm (Guns et al., 2021; Lee et al., 2021; Li et al., 2023), however most are followups of detections at other wavelengths. With the right facility, software and operations model, AtLAST could instead be the source of transient detections which will require followup and monitoring by both AtLAST and other facilities. Take, for example, the JCMT transient study of nearby star forming regions (Herczeg et al., 2017; Mairs et al., 2017, 2024). These studies have shown the variability of local protostars, and the observatory is continuing to monitor the same regions to detect longer periods of variability. However, current facilities are hampered by small fields of view and sensitivity. The former limiting how many targets can be monitored, the latter limiting how precise the measurements can be as well as their overall cadence.

Variability can happen on timescales of minutes to decades. Observing strategies need to be in place to ensure that proper monitoring cadences are maintained. In addition to planned observations of variable sources, the high sensitivity provided by a large aperture telescope greatly increases the chances of serendipitous detections found in datasets intended for other purposes. To ensure that these are picked up on short timescales to enable the triggering of follow-up observations, real-time data reduction pipelines for transient detection are also required. These types of commensal observations, allow for additional science goals to be met using observations designed often for a very different purpose. Many of the upcoming CMB experiments are expected to have such data processing software in place, which will already start taking stock of the transient sky. With 100x the resolution, and significantly better sensitivity, AtLAST will be able to detect more transients that would get lost in the large beams of the CMB surveys.

Following up on transient and variable phenomena will necessitate observing modes like target of opportunity (ToO) overrides, coordinated observations with other facilities for multi-wavelength observations, monitoring campaigns and the commensal observing described above.

See Akiyama et al. (2023) for further details.

The global VLBI network has played a significant role in our understanding of black hole physics, showing the emission from lensed material very close to the black hole itself for the first time (Event Horizon Telescope Collaboration et al., 2019). The existence, nature and shape of this emission in M87 and SgrA^∗ definitively proved the existence of black holes at the centre of galaxies and as the power sources for relativistic jets; provided unique probes of the astrophysics of accretion and relativistic jet formation; and enabled the exploration of properties of general relativity.

By increasing the network, the EHT will be able to image ever fainter sources, particularly with the addition of a telescope as sensitive as AtLAST. Adding a sensitive anchor not only improves the baselines that AtLAST is a part of, but also those between other stations due to improvements in calibration. As a single dish, AtLAST has major advantages over ALMA in this regard as ALMA has to act as a phased array, which limits the sensitivity of sources it can detect. AtLAST will also be able to host an instrument with multi-frequency capabilities, enabling both simultaneous multi-frequency observations (e.g. Issaoun et al., 2023) and order of magnitude improvements in sensitivity due to the frequency phase transfer technique (Akiyama et al., 2023).

The images of M87 and SgrA^∗ are right at the resolution limits of a 230 GHz global VLBI network, yet, within those rings are an infinitely nested set of separate rings (Johnson et al., 2020), each of which reveals new insights into the nature of the black holes powering them. Shifting to higher frequencies (690 GHz) will allow future EHT observations to resolve the n=0 and n=1 rings of these black holes (at $\sim$ 6 microarcsec) and disentangle the astrophysical and relativistic effects in that emission (e.g. Tiede et al., 2022).