AtmoDist: Self-supervised Representation Learning for Atmospheric Dynamics

• URL: http://arxiv.org/abs/2202.01897v1
• Date: Wed, 2 Feb 2022 14:49:48 GMT
• Title: AtmoDist: Self-supervised Representation Learning for Atmospheric Dynamics
• Authors: Sebastian Hoffmann and Christian Lessig
• Abstract summary: We introduce a self-supervised learning task that defines a categorical loss for a wide variety of unlabeled atmospheric datasets. We train a neural network on the simple yet intricate task of predicting the temporal distance between atmospheric fields. We demonstrate this by introducing a data-driven distance metric for atmospheric states based on representations learned from ERA5 reanalysis.
• Score: 4.873362301533825
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Representation learning has proven to be a powerful methodology in a wide variety of machine learning applications. For atmospheric dynamics, however, it has so far not been considered, arguably due to the lack of large-scale, labeled datasets that could be used for training. In this work, we show that the difficulty is benign and introduce a self-supervised learning task that defines a categorical loss for a wide variety of unlabeled atmospheric datasets. Specifically, we train a neural network on the simple yet intricate task of predicting the temporal distance between atmospheric fields, e.g. the components of the wind field, from distinct but nearby times. Despite this simplicity, a neural network will provide good predictions only when it develops an internal representation that captures intrinsic aspects of atmospheric dynamics. We demonstrate this by introducing a data-driven distance metric for atmospheric states based on representations learned from ERA5 reanalysis. When employ as a loss function for downscaling, this Atmodist distance leads to downscaled fields that match the true statistics more closely than the previous state-of-the-art based on an l2-loss and whose local behavior is more realistic. Since it is derived from observational data, AtmoDist also provides a novel perspective on atmospheric predictability.

Related papers

• Generative Data Assimilation of Sparse Weather Station Observations at Kilometer Scales [5.453657018459705]
  We demonstrate the viability of score-based data assimilation in the context of realistically complex km-scale weather. By incorporating observations from 40 weather stations, 10% lower RMSEs on left-out stations are attained. It is a ripe time to explore extensions that combine increasingly ambitious regional state generators with an increasing set of in situ, ground-based, and satellite remote sensing data streams.
  arXiv  Detail & Related papers  (2024-06-19T10:28:11Z)
• Generalizing Weather Forecast to Fine-grained Temporal Scales via Physics-AI Hybrid Modeling [55.13352174687475]
  This paper proposes a physics-AI hybrid model (i.e., WeatherGFT) which Generalizes weather forecasts to Finer-grained Temporal scales. Specifically, we employ a carefully designed PDE kernel to simulate physical evolution on a small time scale. We introduce a lead time-aware training framework to promote the generalization of the model at different lead times.
  arXiv  Detail & Related papers  (2024-05-22T16:21:02Z)
• AtmoRep: A stochastic model of atmosphere dynamics using large scale representation learning [1.677718351174347]
  We propose AtmoRep, a task-independent computer model of atmospheric dynamics. AtmoRep can provide skillful results for a wide range of applications. Our work establishes that large-scale neural networks can provide skillful, task-independent models of atmospheric dynamics.
  arXiv  Detail & Related papers  (2023-08-25T10:02:26Z)
• Inverse Dynamics Pretraining Learns Good Representations for Multitask Imitation [66.86987509942607]
  We evaluate how such a paradigm should be done in imitation learning. We consider a setting where the pretraining corpus consists of multitask demonstrations. We argue that inverse dynamics modeling is well-suited to this setting.
  arXiv  Detail & Related papers  (2023-05-26T14:40:46Z)
• Temporal Subsampling Diminishes Small Spatial Scales in Recurrent Neural Network Emulators of Geophysical Turbulence [0.0]
  We investigate how an often overlooked processing step affects the quality of an emulator's predictions. We implement ML architectures from a class of methods called reservoir computing: (1) a form of spatial Vector Autoregression (N VAR), and (2) an Echo State Network (ESN) In all cases, subsampling the training data consistently leads to an increased bias at small scales that resembles numerical diffusion.
  arXiv  Detail & Related papers  (2023-04-28T21:34:53Z)
• Uncovering the Missing Pattern: Unified Framework Towards Trajectory Imputation and Prediction [60.60223171143206]
  Trajectory prediction is a crucial undertaking in understanding entity movement or human behavior from observed sequences. Current methods often assume that the observed sequences are complete while ignoring the potential for missing values. This paper presents a unified framework, the Graph-based Conditional Variational Recurrent Neural Network (GC-VRNN), which can perform trajectory imputation and prediction simultaneously.
  arXiv  Detail & Related papers  (2023-03-28T14:27:27Z)
• ClimaX: A foundation model for weather and climate [51.208269971019504]
  ClimaX is a deep learning model for weather and climate science. It can be pre-trained with a self-supervised learning objective on climate datasets. It can be fine-tuned to address a breadth of climate and weather tasks.
  arXiv  Detail & Related papers  (2023-01-24T23:19:01Z)
• Palm up: Playing in the Latent Manifold for Unsupervised Pretraining [31.92145741769497]
  We propose an algorithm that exhibits an exploratory behavior whilst it utilizes large diverse datasets. Our key idea is to leverage deep generative models that are pretrained on static datasets and introduce a dynamic model in the latent space. We then employ an unsupervised reinforcement learning algorithm to explore in this environment and perform unsupervised representation learning on the collected data.
  arXiv  Detail & Related papers  (2022-10-19T22:26:12Z)
• Pre-training via Denoising for Molecular Property Prediction [53.409242538744444]
  We describe a pre-training technique that utilizes large datasets of 3D molecular structures at equilibrium. Inspired by recent advances in noise regularization, our pre-training objective is based on denoising.
  arXiv  Detail & Related papers  (2022-05-31T22:28:34Z)
• Towards Representation Learning for Atmospheric Dynamics [6.274453963224799]
  We present a novel, self-supervised representation learning approach specifically designed for atmospheric dynamics. Our approach, called AtmoDist, trains a neural network on a simple, auxiliary task. We demonstrate this by using AtmoDist to define a metric for GAN-based super resolution of vorticity and divergence.
  arXiv  Detail & Related papers  (2021-09-19T07:43:30Z)
• Laplacian Denoising Autoencoder [114.21219514831343]
  We propose to learn data representations with a novel type of denoising autoencoder. The noisy input data is generated by corrupting latent clean data in the gradient domain. Experiments on several visual benchmarks demonstrate that better representations can be learned with the proposed approach.
  arXiv  Detail & Related papers  (2020-03-30T16:52:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.