dorsal/arxiv
View SchemaNetworks as Renormalized Models for Emergent Behavior in Physical Systems
| Authors | Maya Paczuski |
|---|---|
| Categories | |
| ArXiv ID | physics/0502028 |
| URL | https://arxiv.org/abs/physics/0502028 |
| DOI | 10.1142/9789812701558_0042 |
Abstract
Networks are paradigms for describing complex biological, social and technological systems. Here I argue that networks provide a coherent framework to construct coarse-grained models for many different physical systems. To elucidate these ideas, I discuss two long-standing problems. The first concerns the structure and dynamics of magnetic fields in the solar corona, as exemplified by sunspots that startled Galileo almost 400 years ago. We discovered that the magnetic structure of the corona embodies a scale free network, with spots at all scales. A network model representing the three-dimensional geometry of magnetic fields, where links rewire and nodes merge when they collide in space, gives quantitative agreement with available data, and suggests new measurements. Seismicity is addressed in terms of relations between events without imposing space-time windows. A metric estimates the correlation between any two earthquakes. Linking strongly correlated pairs, and ignoring pairs with weak correlation organizes the spatio-temporal process into a sparse, directed, weighted network. New scaling laws for seismicity are found. For instance, the aftershock decay rate decreases as 1/t in time up to a correlation time, t[omori]. An estimate from the data gives t[omori] to be about one year for small magnitude 3 earthquakes, about 1400 years for the Landers event, and roughly 26,000 years for the earthquake causing the 2004 Asian tsunami. Our results confirm Kagan's conjecture that aftershocks can rumble on for centuries.
{
"annotation_id": "fed4eeee-4174-47da-a302-757ae7045604",
"date_created": "2026-03-02T18:00:57.097000Z",
"date_modified": "2026-03-02T18:00:57.097000Z",
"file_hash": "427e9fa34bfa9379ef210183239701d0a300bede6f73ae86b9fd6d8e95784863",
"private": false,
"record": {
"abstract": "Networks are paradigms for describing complex biological, social and\ntechnological systems. Here I argue that networks provide a coherent framework\nto construct coarse-grained models for many different physical systems. To\nelucidate these ideas, I discuss two long-standing problems. The first concerns\nthe structure and dynamics of magnetic fields in the solar corona, as\nexemplified by sunspots that startled Galileo almost 400 years ago. We\ndiscovered that the magnetic structure of the corona embodies a scale free\nnetwork, with spots at all scales. A network model representing the\nthree-dimensional geometry of magnetic fields, where links rewire and nodes\nmerge when they collide in space, gives quantitative agreement with available\ndata, and suggests new measurements. Seismicity is addressed in terms of\nrelations between events without imposing space-time windows. A metric\nestimates the correlation between any two earthquakes. Linking strongly\ncorrelated pairs, and ignoring pairs with weak correlation organizes the\nspatio-temporal process into a sparse, directed, weighted network. New scaling\nlaws for seismicity are found. For instance, the aftershock decay rate\ndecreases as 1/t in time up to a correlation time, t[omori]. An estimate from\nthe data gives t[omori] to be about one year for small magnitude 3 earthquakes,\nabout 1400 years for the Landers event, and roughly 26,000 years for the\nearthquake causing the 2004 Asian tsunami. Our results confirm Kagan\u0027s\nconjecture that aftershocks can rumble on for centuries.",
"arxiv_id": "physics/0502028",
"authors": [
"Maya Paczuski"
],
"categories": [
"physics.geo-ph",
"astro-ph",
"cond-mat.stat-mech",
"physics.plasm-ph"
],
"doi": "10.1142/9789812701558_0042",
"title": "Networks as Renormalized Models for Emergent Behavior in Physical Systems",
"url": "https://arxiv.org/abs/physics/0502028"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "ab580003-2187-4355-9e82-fa3479d95187",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}