TEAM – The Transformer Earthquake Alerting Model

Description: TEAM, the Transformer Earthquake Alerting Model is a deep learning model for real time estimation of peak ground acceleration (TEAM), earthquake magnitude and earthquake location (TEAM-LM). This software package contains the joint implementation of both TEAM and the derivative TEAM-ML, as well as the scripts for training and evaluating these models. In addition, it contains scripts to download an early warning datasets for Japan and implementations of baseline approaches for the estimation of earthquake magnitude and peak ground acceleration. TEAM is implemented in Python. TEAM and TEAM-ML have a variety of configuration parameters that are documented in the README. These configurations need to be provided in JSON format. In addition, multiple example configuration files are provided in the subdirectories pga_configs and magloc_configs. Please note that this implementation is intended for research purpose only. Production use is discouraged.

Global identifier:

Doi(
    "10.5880/GFZ.2.4.2021.003",
)

Tags: Seismik ? Software ? Joint Implementation ? Frühwarnsystem ? Künstliche Intelligenz ? Erdbeben ? EARTH SCIENCE > SOLID EARTH > TECTONICS > EARTHQUAKES > EARTHQUAKE MAGNITUDE/INTENSITY ? python ? real time assessment ?

License: GNU General Public License 3.0

Language: Englisch/English

Organisations

• GFZ Data Services (Herausgeberschaft)
• GFZ Helmholtz-Zentrum für Geoforschung (Bereitstellung)

Persons

• Bindi, Dino (Verfasser*in)
• Bindi, Dino (Beitragende*r)
• Leser, Ulf (Verfasser*in)
• Leser, Ulf (Beitragende*r)
• Münchmeyer, Jannes (Verfasser*in)
• Münchmeyer, Jannes (Beitragende*r)
• Tilmann, Frederik (Verfasser*in)
• Tilmann, Frederik (Beitragende*r)

Issued: 2021-01-01

Status

Aktiv

Quality score

• Overall: 0.64
• Findability: 0.64

  • Title: 0.80
  • Description: 0.39
  • Identifier: true
  • Keywords: 0.67
  • Spatial: NoRegion (0.00)
  • Temporal: true
• Accessibility: 0.67

  • Landing page: Specific (1.00)
  • Direct access: false
  • Publicly accessible: true
• Interoperability: 0.25

  • Open file format: false
  • Media type: false
  • Machine-readable metadata: true
  • Machine-readable data: false
• Reusability: 1.00

  • License: ClearlySpecifiedAndFree (1.00)
  • Contact info: true
  • Publisher info: true

Accessed 1 times.