A novel and theoretically consistent method for correcting systematic errors in earth observation data and earth system model results

一种新颖且理论上一致的方法，用于纠正地球观测数据和地球系统模型结果中的系统误差

• 批准号: FT130100545
• 负责人: A/Prof Valentijn Pauwels
• 金额: $ 42.44万
• 依托单位: Monash University
• 依托单位国家: 澳大利亚
• 项目类别: ARC Future Fellowships
• 财政年份: 2014
• 资助国家: 澳大利亚
• 起止时间: 2014-01-06 至 2018-01-05
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/FT130100545
• 关键词: novel; theoretically; consistent; method; correcting

For a correct interpretation of satellite-based earth observation data and/or Earth system model results, it is very important that these data are free of systematic errors, commonly referred to as bias. It is well known that both these data sources are prone to a significant bias, which is currently neglected in many environmental impact and prediction studies. This project will present a method to develop models for these biases. A state update technique, the Ensemble Kalman Filter, will be adapted to correctly take into account bias in the merging of the two data sources. The project outcomes will be of high importance for long-term environmental studies, since these strongly rely on physically-based models and remote sensing data.

为了正确解释基于卫星的地球观测数据和/或地球系统模型结果，这些数据不含系统错误，通常称为偏见非常重要。众所周知，这两个数据源都容易出现重大偏见，目前在许多环境影响和预测研究中都忽略了这一点。该项目将提出一种开发这些偏见模型的方法。状态更新技术，集合卡尔曼过滤器，将被调整以正确考虑两个数据源的合并中的偏差。对于长期的环境研究，该项目的结果将非常重要，因为这些强烈依赖于基于物理的模型和遥感数据。