AHAaron Hornschild

Aaron Hornschild

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Aaron Hornschild has a total of 8 co-authors. Here are the top co-authors by number of shared publications:

  • Jan Saynisch-Wagner

    GFZ Helmholtz Centre for Geosciences, Potsdam, Germany

    2 shared publications
  • Maik Thomas

    Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences, Section 1.3: Earth System Modelling, Potsdam, Germany

    2 shared publications
  • Christopher Irrgang

    Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences, Section 1.3: Earth System Modelling, Potsdam, Germany

    1 shared publication
  • Christopher Irrgang

    Centre for Artificial Intelligence in Public Health Research, Robert Koch Institute, Berlin, Germany

    1 shared publication
  • Jan Saynisch-Wagner

    Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences, Section 1.3: Earth System Modelling, Potsdam, Germany

    1 shared publication
  • Julien Baerenzung

    Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences, Section 1.3: Earth System Modelling, Potsdam, Germany

    1 shared publication

Latest Publications

  • Open Access

    Development of an Observational Encoder for Data Assimilation in the Latent Space of a Variational Autoencoder (with Sea Surface Temperature)

    by Simone Carsey, Aaron Hornschild, Jan Saynisch-Wagner

    High-dimensional ocean datasets, e.g. of global sea surface temperature, provide crucial insight to the dynamic of physical ocean characteristics such as seasonal cycle, ENSO, and global trend, but the dimensionality often results in computational complexity. Deep learning methods, such as variational autoencoders (VAEs), offer dimension reduction techniques that retain nonlinearities while expressing the system state in a meaningful lower-dimensional latent space. We explore whether encoded spatially limited observations, such as from satellites, buoys, or ship tracks, could be assimilated in the latent space. First, we developed a VAE to create a low-dimensional representation of a global sea surface temperature anomalies dataset. Next, we built a sample environment to demonstrate data assimilation within the latent space by creating spatially incomplete observations from the global dataset by selecting specific regions and adding noise. Accordingly, we developed an observational encoder to map these observations into the latent space of the VAE. For the latent data assimilation, we created a Bayesian update (e.g. Kalman filter) and decoded assimilated observations to evaluate results. We report on the assimilation of encoded limited observations within the latent space and discuss possible applications and future development of this approach.

  • Open Access

    On the detectability of the magnetic fields induced by ocean circulation in geomagnetic satellite observations

    by Aaron Hornschild, Julien Baerenzung, Jan Saynisch-Wagner, Christopher Irrgang, Maik Thomas

    Due to their sensitivity to conductivity and oceanic transport, magnetic signals caused by the movement of the ocean are a beneficial source of information. Satellite observed tidal-induced magnetic fields have already proven to be helpful to derive Earth’s conductivity or ocean heat content. However, magnetic signals caused by ocean circulation are still unobserved in satellite magnetometer data. We present a novel method to detect these magnetic signals from ocean circulation using an observing system simulation experiment. The introduced approach relies on the assimilation of satellite magnetometer data based on a Kalman filter algorithm. The separation from other magnetic contributions is attained by predicting the temporal behavior of the ocean-induced magnetic field through presumed proxies. We evaluate the proposed method in different test case scenarios. The results demonstrate a possible detectability of the magnetic signal in large parts of the ocean. Furthermore, we point out the crucial dependence on the magnetic signal’s variability and show that our approach is robust to slight spatial and temporal deviations of the presumed proxies. Additionally, we showed that including simple prior spatial constraints could further improve the assimilation results. Our findings indicate an appropriate sensitivity of the detection method for an application outside the presented observing system simulation experiment. Therefore, we finally discussed potential issues and required advances toward the method’s application on original geomagnetic satellite observations. Graphical Abstract


  • Open Access

    The detection of the magnetic fields induced by ocean circulation - An observing system simulation experiment

    by Aaron Hornschild, Jan Saynisch-Wagner, Julien Baerenzung, Christopher Irrgang, Maik Thomas

    The movement of conductive seawater through Earth's magnetic field leads to electromagnetic induction processes in the oceans. The resulting radial magnetic fields have been successfully modeled and, in the case of tidal-induced magnetic fields, also identified in satellite magnetometer data. However, the magnetic signals caused by ocean circulation have still remained unobserved.We introduce a new method to detect these signals using an observing system simulation experiment. Our approach relies on a Kalman filter-based assimilation of satellite magnetometer data. A key aspect of identifying ocean-induced signals is separating them from other magnetic contributions. For this separation, we used both estimations of the temporal behavior and spatial constraints as prior information. The observing system simulation experiment allows us to evaluate the proposed method.  We present the results of this evaluation and report on the detectability of magnetic fields induced by ocean circulation.