• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文OA文献 >Estimating top-of-atmosphere thermal infrared radiance using MERRA-2 atmospheric data
【2h】

Estimating top-of-atmosphere thermal infrared radiance using MERRA-2 atmospheric data

机译：使用Merra-2大气数据估算大气热红外光线

代理获取
本网站仅为用户提供外文OA文献查询和代理获取服务，本网站没有原文。下单后我们将采用程序或人工为您竭诚获取高质量的原文，但由于OA文献来源多样且变更频繁，仍可能出现获取不到、文献不完整或与标题不符等情况，如果获取不到我们将提供退款服务。请知悉。
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Space borne thermal infrared sensors have been extensively used for environmental research as well as cross-calibration of other thermal sensing systems. Thermal infrared data from satellites such as Landsat and Terra/MODIS have limited temporal resolution(with a repeat cycle of 1 to 2 days for Terra/MODIS, and 16 days for Landsat). Thermal instruments with finer temporal resolution on geostationary satellites have limited utility for cross-calibration due to their large view angles. Reanalysis atmospheric data is available on a global spatial grid at three hour intervals making it a potential alternative to existing satellite image data. This research explores using the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) reanalysis data product to predict top-of-atmosphere (TOA) thermal infrared radiance globally at time scales finer than available satellite data. The MERRA-2 data product provides global atmospheric data every three hours from 1980 to the present. Due to the high temporal resolution ofthe MERRA-2 data product, opportunities for novel research and applications are presented. While MERRA-2 has been used in renewable energy and hydrological studies, this work seeks to leverage the model to predict TOA thermal radiance. Two approacheshave been followed, namely physics-based approach and a supervised learning approach, using Terra/MODIS band 31 thermal infrared data as reference. The first physics-based model uses forward modeling to predict TOA thermal radiance. The second model infers the presence of clouds from the MERRA-2 atmospheric data, before applying an atmospheric radiative transfer model. The last physics-based model parameterized the previous model to minimize computation time. The second approach applied four different supervised learning algorithms to the atmospheric data. The algorithms included a linear least squares regression model, a non-linear support vector regression (SVR) model, a multi-layer perceptron (MLP), and a convolutional neural network (CNN). This research found that the multi-layer perceptron model produced the lowest error rates overall, with an RMSE of 1.22W / m2 sr um when compared to actual Terra/MODIS band 31 image data. This research further aimed to characterize the errors associated with each method so that any potential user will have the best information available should they wish to apply these methods towards their own application.
机译：空间传播的热红外传感器已广泛用于环境研究以及其他热敏传感系统的交叉校准。来自Landsat和Terra / Modis等卫星的热红外数据具有有限的时间分辨率（Terra / Modis的重复周期为1至2天，并且Landsat的16天）。由于其大视角，具有更精细的时隙卫星上具有更精细的时间分辨率的热仪器具有有限的跨校准。 Reanalysis大气数据可在全球空间网格上提供三个小时的间隔，使其成为现有卫星图像数据的潜在替代方案。本研究探讨了研究和应用的现代回顾性分析，版本2（Merra-2）重新分析数据产品，以预测全球大气（TOA）热红外线，比可用卫星数据更精细地。 MERRA-2数据产品从1980年从1980年每三个小时提供全球大气数据。由于Merra-2数据产品的高时间分辨率，提出了新的研究和应用的机会。虽然Merra-2已用于可再生能源和水文研究，但这项工作旨在利用模型来预测热线。两次接近的接近，即基于物理的方法和监督学习方法，使用Terra / Modis Band 31热红外数据作为参考。基于物理的模型使用前向建模预测TOA热辐射。在施加大气辐射转移模型之前，第二种模型揭示了来自Merra-2大气数据的云的存在。基于物理的模型参数化了以前的模型以最小化计算时间。第二种方法将四种不同的监督学习算法应用于大气数据。该算法包括线性最小二乘回归模型，非线性支持向量回归（SVR）模型，多层Perceptron（MLP）和卷积神经网络（CNN）。本研究发现，与实际的Terra / Modis频段31图像数据相比，多层Perceptron模型总体上总体的误差速率，RMSE为1.22W / M2 SRMUM。该研究进一步旨在表征与每种方法相关的错误，以便任何潜在用户都将有可用的最佳信息，他们希望将这些方法应用于自己的应用程序。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
代理获取

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号