Scholarlink

<p>Climate change has accelerated the dynamics of cryospheric and oceanographic systems, leading to rising sea levels, increased ice melt, and significant shifts in polar regions. Monitoring these changes is crucial, yet traditional observational methods are often impractical in remote, harsh environments. This study presents a novel approach to monitoring oceanographic and cryospheric changes induced by climate dynamics, using a cascaded long short-term memory (LSTM) and recurrent neural network (RNN) model with ICESat and supplementary remote sensing data. The model leverages ICESat’s high-resolution laser altimetry for precise measurements of ice thickness and ice sheet elevation, combined with data from MODIS, Sentinel-1, and CryoSat-2 for comprehensive insights. Our results indicate that the cascaded LSTM-RNN model achieves high predictive accuracy, with root mean square error (RMSE) values as low as 0.08 m for ice thickness and a mean absolute error (MAE) of 0.05 m. Additionally, it attains a correlation coefficient of 0.96, significantly outperforming other models like standalone LSTM (RMSE 0.12 m, MAE 0.09 m) and CNN-LSTM hybrid (RMSE 0.10 m, MAE 0.08 m). This model captures both seasonal and long-term trends in ice dynamics, highlighting its utility for robust climate monitoring. The successful integration of ICESat and other remote sensing data demonstrates the potential for advanced machine learning techniques to enhance predictions of polar and oceanographic changes, supporting climate adaptation efforts.</p>

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

Monitoring Oceanographic and Cryosphere Changes: A Remote Sensing Approach to Climate-Induced Marine and Polar Dynamics

  • 期刊论文

  • 发表日期：

    2025年1月25日

查看全文
  • Reddi Khasim Shaik,
  • D. Santhi Jeslet,
  • Vijay Vasanth Aroulanandam,
  • B. A. Siva,
  • Ranjith Kumar Painam,
  • R Thiagarajan

摘要

Climate change has accelerated the dynamics of cryospheric and oceanographic systems, leading to rising sea levels, increased ice melt, and significant shifts in polar regions. Monitoring these changes is crucial, yet traditional observational methods are often impractical in remote, harsh environments. This study presents a novel approach to monitoring oceanographic and cryospheric changes induced by climate dynamics, using a cascaded long short-term memory (LSTM) and recurrent neural network (RNN) model with ICESat and supplementary remote sensing data. The model leverages ICESat’s high-resolution laser altimetry for precise measurements of ice thickness and ice sheet elevation, combined with data from MODIS, Sentinel-1, and CryoSat-2 for comprehensive insights. Our results indicate that the cascaded LSTM-RNN model achieves high predictive accuracy, with root mean square error (RMSE) values as low as 0.08 m for ice thickness and a mean absolute error (MAE) of 0.05 m. Additionally, it attains a correlation coefficient of 0.96, significantly outperforming other models like standalone LSTM (RMSE 0.12 m, MAE 0.09 m) and CNN-LSTM hybrid (RMSE 0.10 m, MAE 0.08 m). This model captures both seasonal and long-term trends in ice dynamics, highlighting its utility for robust climate monitoring. The successful integration of ICESat and other remote sensing data demonstrates the potential for advanced machine learning techniques to enhance predictions of polar and oceanographic changes, supporting climate adaptation efforts.