A generative deep neural network as an alternative to co-kriging
Year: 2024
Type: article
Abstract: In geosciences, kriging is leading spatial interpolation, and co-kriging is the most commonly used method for accomplishing spatial interpolation of a target variable by incorporating information from a secondary variable. However, the application of cokriging, is not without limitation as spatial stationarity assumptions and data sparsity can significantly impact its performance. Moreover, it can be resource-intensive when used for interpolation with a substantial volume of data, especially in the case of 3D interpolation. In this paper, we introduce a new method for spatial interpolation that considers two variables using a generative deep neural network. This approach utilizes a convolutional neural network with an encoder-decoder architecture, featuring a single encoder and two decoders to handle the two variables. Additionally, we introduce a loss function that facilitates the control over the relationships between the two variables. Our approach eliminates the need for prior training, making labeled data unnecessary.In order to assess the performance of our method, we use two real-world datasets. The first one is a 2D dataset of total soil organic carbon combined with the Normalized Difference Vegetation Index. The second one is a 3D dataset that combines concentrationsof Hydrocarbon and Fluoride obtained from hyperspectral analysis of soil cores. The experimental results demonstrate that theproposed method outperforms ordinary kriging and co-kriging, showing a significant improvement when both variables are used. We also demonstrate how the inclusion of the auxiliary variable serves as a means to mitigate the overfitting of the model.
Institutions: Université Le Havre Normandie, Laboratoire d'Informatique, du Traitement de l'Information et des Systèmes, Université de Rouen Normandie, Institut National des Sciences Appliquées Rouen Normandie, Centre National de la Recherche Scientifique, Environnements et Paléoenvironnements Océaniques et Continentaux, Institut Polytechnique de Bordeaux
Language: English
Subfield: Environmental Engineering
Field: Environmental Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,920
OpenAlex ID: W4403016476