Detecting Ground Deformation in the Built Environment using Sparse Satellite InSAR data with a Convolutional Neural Network

Abstract:

The large volumes of Sentinel-1 data produced over Europe are being used to develop pan-national ground motion services. However, simple analysis techniques like thresholding cannot detect and classify complex deformation signals reliably making providing usable information to a broad range of non-expert stakeholders a challenge. Here we explore the applicability of deep learning approaches by adapting a pre-trained convolutional neural network (CNN) [1,2] to produce a probability map of surface movement and use it to detect deformation in a national-scale velocity field. For our proof-of-concept, we focus on the UK where previously identified deformation is associated with coal-mining, ground water withdrawal, landslides and tunnelling. The sparsity of measurement points and the presence of spike noise make this a challenging application for CNNs, which involve calculations of the spatial convolution between images. Moreover, insufficient ground truth data exists to construct a balanced training data set, and the deformation signals are slower and more localised than in previous applications.

Here, we propose three enhancement methods to tackle these problems: i) spatial interpolation with modified matrix completion, ii) a synthetic training dataset based on the characteristics of the real UK velocity map, and iii) enhanced over-wrapping techniques. Using velocity maps spanning 2015-2019, our framework detects several areas of coal mining subsidence, uplift due to dewatering, slate quarries, landslides and tunnel engineering works. The results demonstrate the potential applicability of the proposed framework to the development of automated ground motion analysis systems for anthropogenic sources of deformation in urban and semi-urban environments. As the dataset is very large, it would not be feasible to manually inspect the entire area at high resolution. Using a probability threshold of 0.5, the method produces some false positives and false negatives. However, the probability values and the sizes of the detected areas can be employed to prioritise further analysis.

[1] N Anantrasirichai, J Biggs, F Albino, P Hill, D Bull (2018), Application of Machine Learning to Classification of Volcanic Deformation in Routinely Generated InSAR Data, Journal of Geophysical Research: Solid Earth

[2] N Anantrasirichai, J Biggs, F Albino, D Bull (2019), A deep learning approach to detecting volcano deformation from satellite imagery using synthetic datasets, Remote Sensing of Environment 230, 111179