Scholarlink

In various global zones—equatorial, polar, and auroral—ionospheric scintillation proves a formidable hurdle to Global Navigation Satellite System (GNSS) performance. Over the past couple of decades, GNSS Radio Occultation (RO) has emerged as a vital resource, furnishing top-tier atmospheric data for seamless integration into Numerical Weather Prediction (NWP) models and the advancement of meteorological research. Nevertheless, scintillations disrupt measurements in satellite-to-satellite GNSS-RO geometry, influenced by solar flares, seasons, geomagnetic activity, geographical positions, and local time. These disturbances introduce positioning errors, deteriorating GNSS performance and underscoring the importance of accurate detection. Traditional algorithms often lack sensitivity, particularly in identifying strong scintillation due to its sporadic nature, leading to dataset imbalances. To address this, we’ve harnessed a myriad of machine learning (ML) algorithms to detect varying degrees of ionospheric scintillation, aiming to rectify dataset imbalances and bolster accuracy. Our study delves into Total Electron Content (TEC) and Ionospheric Phase Scintillation classification, predicting TEC using regressors like XGBoost Regressor, Autoregressive model (AR), and Exponential Smoothing model (ES), and classifying ionospheric phase scintillation using classifiers such as XGBoost, Naïve Bayes, and Light Gradient Boosting Machine (Light GBM). Through comprehensive comparative analysis, we evaluate regressors and classifiers using standard metrics like Mean Squared Error, Root Mean Squared Error, R2 Score, Accuracy, Recall, Precision, F1-Score, and Confusion Matrix.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Leveraging State-of-the Art Computational Models to Forecast TEC and Categorize Ionospheric Scintillations Using ML Methodologies

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Kothapalli Sri Brinda,
  • B. Modini,
  • V. B. S. Srilatha Indira Dutt

摘要

In various global zones—equatorial, polar, and auroral—ionospheric scintillation proves a formidable hurdle to Global Navigation Satellite System (GNSS) performance. Over the past couple of decades, GNSS Radio Occultation (RO) has emerged as a vital resource, furnishing top-tier atmospheric data for seamless integration into Numerical Weather Prediction (NWP) models and the advancement of meteorological research. Nevertheless, scintillations disrupt measurements in satellite-to-satellite GNSS-RO geometry, influenced by solar flares, seasons, geomagnetic activity, geographical positions, and local time. These disturbances introduce positioning errors, deteriorating GNSS performance and underscoring the importance of accurate detection. Traditional algorithms often lack sensitivity, particularly in identifying strong scintillation due to its sporadic nature, leading to dataset imbalances. To address this, we’ve harnessed a myriad of machine learning (ML) algorithms to detect varying degrees of ionospheric scintillation, aiming to rectify dataset imbalances and bolster accuracy. Our study delves into Total Electron Content (TEC) and Ionospheric Phase Scintillation classification, predicting TEC using regressors like XGBoost Regressor, Autoregressive model (AR), and Exponential Smoothing model (ES), and classifying ionospheric phase scintillation using classifiers such as XGBoost, Naïve Bayes, and Light Gradient Boosting Machine (Light GBM). Through comprehensive comparative analysis, we evaluate regressors and classifiers using standard metrics like Mean Squared Error, Root Mean Squared Error, R2 Score, Accuracy, Recall, Precision, F1-Score, and Confusion Matrix.