SearcharxivSearch

arXiv subjects

Ryan J. Roberts

Publications and source records attributed to Ryan J. Roberts.

3 recordsLinked to original sources

Galaxy discs regulate the growth of supermassive black holes

We examine the relationship between the mass of present-day central supermassive black holes (SMBHs, $M_{\rm BH}$), and the stellar mass ($M_{\star}$) and halo mass ($M_{200}$) of their host galaxies in the EAGLE simulation, and find that scatter about these relations correlates with both halo structure and galaxy morphology. EAGLE reproduces the observed $M_{\rm BH}$-$M_{\star}$ relation, including (qualitatively) its dependence on morphology: at fixed $M_{\star}$, disc-dominated galaxies host less massive SMBHs than ellipticals. We show that $M_{\rm BH}$ correlates with $M_{200}$, as expected if SMBHs are regulated by processes acting on the scale of the host dark matter halo, but exhibits a tighter correlation with the halo binding energy ($E_{\rm bind}$), signalling that this quantity, which encodes information about both halo mass and halo structure, is more fundamental to $M_{\rm BH}$. As with $M_{\rm BH}$-$M_{\star}$, scatter about the $M_{\rm BH}$-$E_{\rm bind}$ relation is strongly correlated with morphology. Gas in the central few parsecs of galaxies with present-day discs retains strong rotational support as the galaxy grows, inhibiting inward transport and precluding periods of rapid SMBH growth by gas accretion. In galaxies destined to be present-day ellipticals, however, this rotational support is disrupted, enabling gas to be funnelled onto the central SMBH, triggering rapid growth. Evolution of the mass fraction of stars formed ex-situ indicates that this disruption is caused by galaxy-galaxy interactions and mergers. Our findings corroborate the conclusion of recent studies, based on controlled simulations of an ~$L^{\star}$ galaxy, that prolonged secular galaxy evolution inhibits central SMBH growth.

astro-ph.GA

Dargana: fine-tuning EarthPT for dynamic tree canopy mapping from space

We present Dargana, a fine-tuned variant of the EarthPT time-series foundation model that achieves specialisation using <3% of its pre-training data volume and 5% of its pre-training compute. Dargana is fine-tuned to generate regularly updated classification of tree canopy cover at 10m resolution, distinguishing conifer and broadleaved tree types. Using Cornwall, UK, as a test case, the model achieves a pixel-level ROC-AUC of 0.98 and a PR-AUC of 0.83 on unseen satellite imagery. Dargana can identify fine structures like hedgerows and coppice below the training sample limit, and can track temporal changes to canopy cover such as new woodland establishment. Our results demonstrate how pre-trained Large Observation Models like EarthPT can be specialised for granular, dynamic land cover monitoring from space, providing a valuable, scalable tool for natural capital management and conservation.

physics.geo-ph

AstroPT: Scaling Large Observation Models for Astronomy

This work presents AstroPT, an autoregressive pretrained transformer developed with astronomical use-cases in mind. The AstroPT models presented here have been pretrained on 8.6 million $512 \times 512$ pixel $grz$-band galaxy postage stamp observations from the DESI Legacy Survey DR8. We train a selection of foundation models of increasing size from 1 million to 2.1 billion parameters, and find that AstroPT follows a similar saturating log-log scaling law to textual models. We also find that the models' performances on downstream tasks as measured by linear probing improves with model size up to the model parameter saturation point. We believe that collaborative community development paves the best route towards realising an open source `Large Observation Model' -- a model trained on data taken from the observational sciences at the scale seen in natural language processing. To this end, we release the source code, weights, and dataset for AstroPT under the MIT license, and invite potential collaborators to join us in collectively building and researching these models.

astro-ph.IM