Abstract

The GOES-R Advanced Baseline Imager (ABI) instrument represents a considerable challenge for engineers to monitor its health, safety, and radiometric performance. As there are 7856 detectors across 16 spectral channels, the existing approach for monitoring the radiometric performance of the GOES N-P Imagers, each with only 16 detectors, is no-longer possible. The ABI radiometric monitoring has to be automated in order to handle such a huge volume of data. To help with this task, a machine learning system has been developed and implemented that automates the trending and monitoring of ABI radiometric calibration datasets. This machine-learning framework is used to capture the time dependent trends of these datasets through data training; the result of which consists of a time dependent function and noise level for each parameter. Because the time dependent trend for a dataset is highly sensitive to changes above its noise level, outliers that deviate from the existing pattern are detected by comparing the values of a dataset against predictions of its time dependent trend. Anomalies in the machine-learning framework are defined as unexpected data pattern changes, which are quantitatively characterized by dimensionless metrics. These metrics are used in a clustering analysis to separate anomalous datasets from the normal ones. In addition, the machine learning outputs enable the assessment of relative detector data quality by comparing the noise level of a detector with the average behavior for a given channel. A detector with a higher noise level has significant impact on the radiometric accuracy. Initial results from our radiometric monitoring show that many of the anomalies characterized in the machine learning approach cannot be detected through the existing static approach. The machine learning approach brings fundamental advances in data trending, anomaly detection, and analysis, and it leads to a more dynamic, proactive, and autonomous monitoring of ABI radiometric performance.

Share

COinS
 
Jun 19th, 4:25 PM

Monitoring GOES-R ABI Radiometric Performances with a Machine Learning System

The GOES-R Advanced Baseline Imager (ABI) instrument represents a considerable challenge for engineers to monitor its health, safety, and radiometric performance. As there are 7856 detectors across 16 spectral channels, the existing approach for monitoring the radiometric performance of the GOES N-P Imagers, each with only 16 detectors, is no-longer possible. The ABI radiometric monitoring has to be automated in order to handle such a huge volume of data. To help with this task, a machine learning system has been developed and implemented that automates the trending and monitoring of ABI radiometric calibration datasets. This machine-learning framework is used to capture the time dependent trends of these datasets through data training; the result of which consists of a time dependent function and noise level for each parameter. Because the time dependent trend for a dataset is highly sensitive to changes above its noise level, outliers that deviate from the existing pattern are detected by comparing the values of a dataset against predictions of its time dependent trend. Anomalies in the machine-learning framework are defined as unexpected data pattern changes, which are quantitatively characterized by dimensionless metrics. These metrics are used in a clustering analysis to separate anomalous datasets from the normal ones. In addition, the machine learning outputs enable the assessment of relative detector data quality by comparing the noise level of a detector with the average behavior for a given channel. A detector with a higher noise level has significant impact on the radiometric accuracy. Initial results from our radiometric monitoring show that many of the anomalies characterized in the machine learning approach cannot be detected through the existing static approach. The machine learning approach brings fundamental advances in data trending, anomaly detection, and analysis, and it leads to a more dynamic, proactive, and autonomous monitoring of ABI radiometric performance.