Digital Image Correlation of Google Earth Images for Earth’s Surface Displacement Estimation

Luigi Guerriero,Domenico Calcaterra,Diego Di Martire,Mirko Francioni

doi:10.3390/rs12213518

Luigi Guerriero, Domenico Calcaterra + Show 2 more

Open Access

https://doi.org/10.3390/rs12213518

Copy DOI

Abstract

An increasing number of satellite platforms provide daily images of the Earth’s surface that can be used in quantitative monitoring applications. However, their cost and the need for specific processing software make such products not often suitable for rapid mapping and deformation tracking. Google Earth images have been used in a number of mapping applications and, due to their free and rapid accessibility, they have contributed to partially overcome this issue. However, their potential in Earth’s surface displacement tracking has not yet been explored. In this paper, that aspect is analyzed providing a specific procedure and related MATLAB™ code to derive displacement field maps using digital image correlation of successive Google Earth images. The suitability of the procedure and the potential of such images are demonstrated here through their application to two relevant case histories, namely the Slumgullion landslide in Colorado and the Miage debris-covered glacier in Italy. Result validation suggests the effectiveness of the proposed procedure in deriving Earth’s surface displacement data from Google Earth images.

Highlights

In the last decades, an increasing number of satellite platforms are persistently acquiring Earth observation data in the form of optical, multispectral, hyperspectral, and radar products (e.g., GeoEye, WorldView, Sentinel, Jilin, Radarsat, Envisat)
The suitability of the procedure is demonstrated here through its application to two relevant case histories, namely the Slumgullion landslide in southwestern Colorado, USA (Figure 1a; [29,30]), which is consistently moving since at least the last three centuries, and the Miage debris-covered glacier in Italy, which is subject to seasonal displacement (Figure 1b; [31])
The analysis indicates that the middle and lower sectors of the Miage debris-covered glacier consistently moved between 29 August 2009 and 28 May 2011 (637 days)

Summary

Introduction

An increasing number of satellite platforms are persistently acquiring Earth observation data in the form of optical, multispectral, hyperspectral, and radar products (e.g., GeoEye, WorldView, Sentinel, Jilin, Radarsat, Envisat) Such products form a very important database for Earth surface control and mapping, offering a general medium to high resolution (0.3 to >50 m single-sided pixel dimension) and a revisiting time of a few days to a few weeks. Due to the characteristics of the sensors, it is able to work in all-weather and day and night conditions This approach is not always suitable due to limitations related to the maximum detectable velocity [11] and potential high cost, especially in the context of high-resolution analyses (e.g., persistent scatterer approaches, [12,13]). Similar to SAR imagery, there are major limitations, where optical images are not suitable for fast mapping and fast process understanding, because of their relevant cost especially for very large areas and the time necessary to access the data

Methods

Discussion

Conclusion