SBIR Phase I: Autonomous Interferometric Synthetic Aperture Radar (InSAR) for surface deformation monitoring

该小型企业创新研究（SBIR）第一阶段项目的更广泛影响是实现全球自主检测表面变形。测量地球表面变形对于检测和分析由于人类活动引起的表面和地下变化至关重要，具有无数的工业应用，包括监测石油和天然气开采领域和储存库、采矿作业、二氧化碳封存和/或基础设施完整性。分析干涉合成孔径雷达（干涉合成孔径雷达）数据的市场预计将在5年内翻一番，这说明了其用途的经济和社会影响。除了干涉合成孔径雷达的巨大经济增长之外，其广泛的应用具有广泛的社会和科学影响，特别是在自然灾害和气候变化方面。 干涉合成孔径雷达处理的进步和信噪比的提高将转化为对地震活动、滑坡、供水、森林砍伐、洪水、冰盖等的更好监测。这一小企业创新研究（SBIR）第一阶段项目旨在解决干涉合成孔径雷达处理缺乏自动化的问题，并提高干涉合成孔径雷达时间序列分析的检测阈值。虽然该技术可以测量几天到几年时间内毫米级的变形变化，但大气影响会引入可能掩盖微小表面变形的误差，从而对重复通过干涉合成孔径雷达判读造成严重破坏。这些影响主要是由于对流层中的压力、温度和相对湿度变化造成的，可能导致比大多数感兴趣的变形信号更大的误差。目前的算法不适合在没有先验数据的情况下进行自动化的大规模监测，因为它们需要耗时的人工干预，并且最终产品需要详尽的专家解释。通过开发机器学习和人工智能方法，该项目的目标是：通过使仍然依赖大量和昂贵的人工干预的处理流程的某些关键部分自动化，进一步自动化和加速干涉合成孔径雷达时间序列的处理;以及（ii）开发新的方法来生成干涉合成孔径雷达时间序列，这是鲁棒的噪声，并允许更精细的时间和空间分辨率相比，国家的，该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this Small Business Innovation Research (SBIR) Phase I project is to enable the global autonomous detection of surface deformation. Measuring Earth surface deformation is fundamental to detect and analyze surface and subsurface changes due to anthropogenic activity, with a myriad of industrial applications that includes the monitoring of oil and gas extraction fields and storage reservoirs, mining operations, carbon dioxide sequestration, and/or infrastructure integrity. Illustrating the economic and social impact of its uses, the market for analyzing Interferometric Synthetic Aperture Radar (InSAR) data is expected to double within 5 years. Beyond the dramatic economic growth of InSAR, its far-ranging applications have broad social and scientific impacts, in particular related to natural hazards and climate change. Advances in InSAR processing and improved signal-to-noise ratios will translate into improved monitoring of earthquake activity, landslides, water supplies, deforestation, floods, ice sheets, etc.This Small Business Innovation Research (SBIR) Phase I project aims at tackling the lack of automation in InSAR processing and improving detection thresholds in InSAR time series analysis. While the technique can potentially measure millimeter-scale changes in deformation over periods of days to years, atmospheric effects can wreak havoc on repeat-pass InSAR interpretation by introducing errors that may mask small surface deformations. These effects, which are fundamentally due to pressure, temperature and relative humidity variations in the troposphere, can lead to errors that are larger than most of the deformation signals of interest. Current algorithms are not suited for automated, large-scale monitoring without a priori data because they require time-consuming manual intervention, and the final product requires exhaustive expert interpretation. Through the development of machine learning and artificial intelligence methods this project aims at: (i) further automating and accelerating the processing of InSAR time series, via the automation of some key sections of the processing pipeline that still rely on extensive and costly human intervention; and (ii) developing a new methodology to generate InSAR time series, that is robust to noise and allows for a finer temporal and spatial resolution compared to the state-of-the-art.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.