Service / Application

Monitoring of Slope Fluctuation by Interferometric SAR Analysis

Development of Analysis Methods Based on the Knowledge of Slope Engineers

NEEDS

Towards the Monitoring of Slopes with Large Disaster Risks that Exist in Vast Numbers

In recent years, earth observation satellites equipped with Synthetic Aperture Radar (SAR), which can observe land areas day and night regardless of weather conditions, have been launched by various countries and are being actively utilized. The observation data acquired by SAR includes information on the intensity of radar reflections from the ground surface as well as phase information of reflected waves*.

SAR Observation Images

Currently, change extraction by aerial laser surveying, which is used for wide-area surveying, has an error of about 30 cm, but the "two-period differential interferometric SAR analysis" can measure the ground surface change with an accuracy of several mm to several cm. Since the interferometric SAR analysis technology can measure minute changes in the ground surface without visiting the site, it enables continuous and labor-saving observation and monitoring of slopes with high disaster risk, which exist in vast amounts in mountainous regions.

Conceptual Diagram of Two-time Differential Interferometric SAR Analysis
Reference:https://vivaweb2.bosai.go.jp/member/ozawa/

*Phase information: Information that indicates the position of radio waves that periodically repeat peaks and troughs. It is expressed by -π to +π.

SOLUTION

Methods to Improve Analysis Accuracy

In addition to phase information caused by ground surface deformation, SAR observation data acquired by satellites also include noise caused by inhomogeneities in the ionosphere and water vapor that exist on the path between the satellite and the ground surface. While the two-time-differential interferometric SAR analysis has the advantage of being able to analyze slope variation using only two SAR observation data, it has a disadvantage where phase information caused by factors other than ground surface variation appears as noise. These noises can be reduced by correction using meteorological data and various filtering methods, but in principle, they cannot be reduced to zero. To solve these disadvantages, there is an analysis method called interferometric SAR time series analysis. This analysis method reduces spatiotemporally random noise such as ionosphere and water vapor by statistically processing a large number of SAR observation data* and enables us to measure ground surface deformation with higher accuracy than two-time differential interferometric SAR analysis. In addition, since a large number of SAR observation data are used, the amount of variation at each observation point can be determined over time.

*At least 15 periods required

POINT

Development of Slope-specific Analysis Algorithms

Interferometric SAR time series analysis is a method for extracting and analyzing areas with high interference* characteristics over time from SAR observation data. In general, interference of man-made structures does not change significantly in their radio wave reflection characteristics, so they are not susceptible to interference degradation. However, natural and cut slopes are susceptible to interference degradation due to vegetation change and other factors. Because of these characteristics, it has been considered difficult to analyze slopes using conventional interferometric SAR time series analysis methods. We have succeeded in measuring slope deformation phenomena such as landslides by using a unique algorithm that enables analysis even in areas with low interferometric properties such as natural slopes and soil structures such as cut slopes. The analysis results are provided as one of the contents of the satellite disaster prevention information service* that we jointly released with SKY Perfect JSAT Corporation and ZENRIN and can be easily viewed on a web browser without using any special software. This system allows us to view analysis results superimposed on maps and landslide disaster warning areas, as well as to view an output time-series graphs of changes at any analysis point, enabling centralized management even when monitoring is conducted over a wide area.

Image of Analysis Result Viewing System
This system is being improved. Displayed items may differ from the actual ones.