July 26, 2024
Report

Data Assessment and Assimilation for Atmospheric Radiation Measurement Data Using Dynamic Bayesian Networks

Abstract

A data-driven machine learning approach to scientific sensor data assessment may automatically learn directly from the structural and temporal patterns in abundant observations. The project has developed data assessment and assimilation models to screen datasets using dynamic Bayesian network (DBN) and deep learning (DL) methods, which rely on correlations between variables, across time, and between spatial locations, to detect poor or invalid data with a high degree of confidence.

Published: July 26, 2024

Citation

Chin G., J.M. Brandi-Lozano, W.I. Gustafson, K. Porterfield, and L. Riihimaki. 2019. Data Assessment and Assimilation for Atmospheric Radiation Measurement Data Using Dynamic Bayesian Networks Richland, WA: Pacific Northwest National Laboratory.