ALBERT

All Library Books, journals and Electronic Records Telegrafenberg

X
feed icon rss

Your email was sent successfully. Check your inbox.

An error occurred while sending the email. Please try again.

Proceed reservation?

Export
  • 1
    facet.materialart.
    Unknown
    Layer-Wise Relevance Propagation for Echo State Networks Applied to Earth System Variability (2022)
    AIRCC Publishing Corporation
    Details
    Publication Date: 2023-01-02
    Description: Artificial neural networks (ANNs) are known to be powerful methods for many hard problems (e.g. image classification, speech recognition or time series prediction). However, these models tend to produce black-box results and are often difficult to interpret. Layer-wise relevance propagation (LRP) is a widely used technique to understand how ANN models come to their conclusion and to understand what a model has learned. Here, we focus on Echo State Networks (ESNs) as a certain type of recurrent neural networks, also known as reservoir computing. ESNs are easy to train and only require a small number of trainable parameters, but are still black-box models. We show how LRP can be applied to ESNs in order to open the black-box. We also show how ESNs can be used not only for time series prediction but also for image classification: Our ESN model serves as a detector for El Nino Southern Oscillation (ENSO) from sea surface temperature anomalies. ENSO is actually a well-known problem and has been extensively discussed before. But here we use this simple problem to demonstrate how LRP can significantly enhance the explainablility of ESNs.
    Type: Conference or Workshop Item , NonPeerReviewed
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 2
    facet.materialart.
    Unknown
    Modularizing Earth system models for interactive simulation (2022)
    Springer
    Details
    Publication Date: 2023-02-02
    Description: Interactive exploration of Earth system simulations may have great potential to improve the scientific modeling process. It will allow monitoring of the state of the simulation via dashboards presenting real-time diagnostics within a digital twin world. We present the state of the art for Earth system modeling in this context. Cross-domain data handling and fusion will make it possible to integrate model and observation data in the context of digital twins of the ocean. Domain-driven modularization of monolithic Earth system models allows one to recover interfaces for such a cross-domain fusion. Reverse engineering with static and dynamic analysis enables modularization of Earth system models. The modularization does not only help with restructuring existing Earth system models, it also makes it possible to integrate additional scientific domains into the interactive simulation environment.
    Type: Article , PeerReviewed
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 3
    facet.materialart.
    Unknown
    Fluctuations of the Atlantic North Equatorial Undercurrent and associated changes in oxygen transports (2020)
    AGU (American Geophysical Union) | Wiley
    Details
    Publication Date: 2023-02-08
    Description: Although the core velocity of the Atlantic North Equatorial Undercurrent (NEUC) is low (0.1−0.3 m s−1), it has been suggested to act as an important oxygen supply route towards the oxygen minimum zone in the eastern tropical North Atlantic. For the first time, the intraseasonal to interannual NEUC variability and its impact on oxygen are investigated based on shipboard and moored velocity observations around 5°N, 23°W. In contrast to previous studies that were mainly based on models or hydrographic data, we find hardly any seasonal cycle of NEUC transports in the central Atlantic. The NEUC transport variability is instead dominated by sporadic intraseasonal events. Only some of these events are associated with high oxygen levels suggesting an occasional eastward oxygen supply by NEUC transport events. Nevertheless, they likely contribute to the local oxygen maximum in the mean shipboard section along 23°W at the NEUC core position.
    Type: Article , PeerReviewed , info:eu-repo/semantics/article
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 4
    facet.materialart.
    Unknown
    Predictability of European winter 2019/20: Indian Ocean dipole impacts on the NAO (2020)
    Royal Meteorological Society
    Details
    Publication Date: 2023-02-08
    Description: Northern Europe and the UK experienced an exceptionally warm and wet winter in 2019/20, driven by an anomalously positive North Atlantic Oscillation (NAO). This positive NAO was well forecast by several seasonal forecast systems, suggesting that this winter the NAO was highly predictable at seasonal lead times. A very strong positive Indian Ocean dipole (IOD) event was also observed at the start of winter. Here we use composite analysis and model experiments, to show that the IOD was a key driver of the observed positive NAO. Using model experiments that perturb the Indian Ocean initial conditions, two teleconnection pathways of the IOD to the north Atlantic emerge: a tropospheric teleconnection pathway via a Rossby wave train travelling from the Indian Ocean over the Pacific and Atlantic, and a stratospheric teleconnection pathway via the Aleutian region and the stratospheric polar vortex. These pathways are similar to those for the El Niño Southern Oscillation link to the north Atlantic which are already well documented. The anomalies in the north Atlantic jet stream location and strength, and the associated precipitation anomalies over the UK and northern Europe, as simulated by the model IOD experiments, show remarkable agreement with those forecast and observed.
    Type: Article , PeerReviewed
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 5
    facet.materialart.
    Unknown
    Decomposing barotropic transport variability in a high‐resolution model of the North Atlantic Ocean (2020)
    AGU (American Geophysical Union) | Wiley
    Details
    Publication Date: 2023-02-08
    Description: A method using a linear shallow water model is presented for decomposing the temporal variability of the barotropic streamfunction in a high‐resolution ocean model. The method is based in the vertically‐averaged momentum equations and is applied to the time series of annual mean streamfunction from the model configuration VIKING20 for the northern North Atlantic. An important result is the role played by the nonlinear advection terms in VIKING20 for driving transport. The method is illustrated by examining how the Gulf Stream transport in the recirculation region responds to the winter North Atlantic Oscillation (NAO). While no statistically significant response is found in the year overlapping with the winter NAO index, there is a tendency for the Gulf Stream transport to increase as the NAO becomes more positive. This becomes significant in lead years 1 and 2 when the mean flow advection (MFA) and eddy momentum flux (EMF) contributions, associated with nonlinear momentum advection, dominate. Only after 2 years, does the potential energy (PE) term, associated with the density field, start to play a role and it is only after 5 years that the transport dependence on the NAO ceases to be significant. It is also shown that the PE contribution to the transport streamfunction has significant memory of up to 5 years in the Labrador and Irminger Seas. However, it is only around the northern rim of these seas that VIKING20 and the transport reconstruction exhibit similar memory. This is due to masking by the MFA and EMF contributions.
    Type: Article , PeerReviewed
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 6
    facet.materialart.
    Unknown
    Equatorial deep jets and their in influence on the mean equatorial circulation in an idealized ocean model forced by intraseasonal momentum flux convergence (2020)
    AGU (American Geophysical Union) | Wiley
    Details
    Publication Date: 2023-02-08
    Description: Equatorial deep jets (EDJ) are vertically stacked, downward propagating zonal currents that alternate in direction with depth. In the tropical Atlantic, they have been shown to influence both surface conditions and tracer variability. Despite their importance, the EDJ are absent in most ocean models. Here we show that EDJ can be generated in an idealized ocean model when the model is driven only by the convergence of the meridional flux of intraseasonal zonal momentum diagnosed from a companion model run driven by steady wind forcing, corroborating the recent theory that intraseasonal momentum flux convergence maintains the EDJ. Additionally, the EDJ in our model nonlinearly generate mean zonal currents at intermediate depths that show similarities in structure to the observed circulation in the deep equatorial Atlantic, indicating their importance for simulating the tropical ocean mean state.
    Type: Article , PeerReviewed
    Format: text
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 7
    facet.materialart.
    Unknown
    CICMoD - A Climate Index Collection Benchmark (2023)
    Details
    Publication Date: 2023-11-20
    Type: Conference or Workshop Item , NonPeerReviewed
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 8
    facet.materialart.
    Unknown
    A Bottom-Up Sampling Strategy for Reconstructing Geospatial Data from Ultra Sparse Inputs (2023)
    Details
    Publication Date: 2023-11-20
    Description: Working with observational data in the context of geophysics can be challenging, since we often have to deal with missing data. This requires imputation techniques in pre-processing to obtain data-mining-ready samples. Here, we present a convolutional neural network (CNN) approach from the domain of deep learning to reconstruct complete data from sparse inputs. CNN architectures are state-of-the-art for image processing. As data, we use two-dimensional fields of sea level pressure (SLP) and sea surface temperature (SST) anomalies. To have consistent data over a sufficiently long time span, we favor to work with output from control simulations of two Earth System Models (ESMs), namely the Flexible Ocean and Climate Infrastructure and the Community Earth System Model. Our networks can restore complete information from incomplete input samples with varying rates of missing data. Moreover, we present a technique to identify the most relevant grid points of our input samples. Choosing the optimal subset of grid points allows us to successfully reconstruct SLP and SST anomaly fields from ultra sparse inputs. As a proof of concept, the insights obtained from ESMs can be transferred to real world observations to improve reconstruction quality. As uncertainty measure, we compare several climate indices derived from reconstructed versus complete fields.
    Type: Conference or Workshop Item , NonPeerReviewed
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 9
    facet.materialart.
    Unknown
    Reconstruct Geospatial Data from Ultra Sparse Inputs to Predict Climate Events (2023)
    IEEE
    Details
    Publication Date: 2023-11-01
    Description: Working with observational data in the context of geophysics can be challenging, since we often have to deal with missing data. This requires imputation techniques in pre-processing to obtain data-mining-ready samples. Here, we present a convolutional neural network approach from the domain of deep learning to reconstruct complete information from sparse inputs. As data, we use various two-dimensional geospatial fields. To have consistent data over a sufficiently long time span, we favor to work with output from control simulations of two Earth System Models, namely the Flexible Ocean and Climate Infrastructure and the Community Earth System Model. Our networks can restore complete information from incomplete input samples with varying rates of missing data. Moreover, we apply a bottom-up sampling strategy to identify the most relevant grid points for each input feature. Choosing the optimal subset of grid points allows us to successfully reconstruct current fields and to predict future fields from ultra sparse inputs. As a proof of concept, we predict El Niño Southern Oscillation and rainfall in the African Sahel region from sea surface temperature and precipitation data, respectively. To quantify uncertainty, we compare corresponding climate indices derived from reconstructed versus complete fields.
    Type: Conference or Workshop Item , NonPeerReviewed
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
  • 10
    facet.materialart.
    Unknown
    Fact or Artifact? Revise Layer-wise Relevance Propagation on various ANN Architectures (2023)
    AIRCC
    Details
    Publication Date: 2024-01-09
    Description: Layer-wise relevance propagation (LRP) is a widely used and powerful technique to reveal insights into various artificial neural network (ANN) architectures. LRP is often used in the context of image classification. The aim is to understand, which parts of the input sample have highest relevance and hence most influence on the model prediction. Relevance can be traced back through the network to attribute a certain score to each input pixel. Relevance scores are then combined and displayed as heat maps and give humans an intuitive visual understanding of classification models. Opening the black box to understand the classification engine in great detail is essential for domain experts to gain trust in ANN models. However, there are pitfalls in terms of model-inherent artifacts included in the obtained relevance maps, that can easily be missed. But for a valid interpretation, these artifacts must not be ignored. Here, we apply and revise LRP on various ANN architectures trained as classifiers on geospatial and synthetic data. Depending on the network architecture, we show techniques to control model focus and give guidance to improve the quality of obtained relevance maps to separate facts from artifacts.
    Type: Article , PeerReviewed
    Format: text
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
    PAPER (OCEANREP)
Close ⊗
This website uses cookies and the analysis tool Matomo. More information can be found here...