Radar Interferometry for the Detection and Monitoring of Mass Movement over Southern Red Sea Hills
El Kadiri, Racha1; Sultan, Mohamed1; Becker, Richard2; Chouinard, Kyle1; Emil, Mustafa Kemal1; Bouali, El Hachemi1
1Western Michigan University, UNITED STATES; 2University of Toledo, UNITED STATES

Landslides pose serious threats to settlements and to infrastructures that support transportation, tourism and other economic activities. Across the globe, landslides cause approximately 1000 deaths per year and property damage amounting to 4 billion US dollars per year. The Red Sea Hills in southwestern Saudi Arabia are particularly susceptible to mass movement hazards. The area is characterized by high elevations (up to 2.5 km a.m.s.l) and steep topographic gradients. Rainfall is another triggering mechanism for landslides in the study area. The high precipitation levels (around 550mm/year) are largely related to tropical air masses that reach the southwestern extreme part of Saudi Arabia. Additionally, the Red Sea Hills are characterized by intense structural deformation (e.g., folds, faults, fractures), and numerous planes of weakness of varying orientations. In the past, assessment of the hazards associated with mass movement in the Red Sea Hills involved identifying landslides through field work. This has not allowed a thorough investigation of landslides, especially those that experience slow rates of movement. Some of which have the potential for much higher rates of movement during brief rainfall events. Interferometric Synthetic Aperture Radar (InSAR) techniques provide opportunities to measure mass movements over decadal scales and their applications are especially valued in rugged and inaccessible areas, where field work is difficult, expensive and time consuming to conduct. Our analyses are based on 7 ERS-1/2 SAR and 7 ENVISAT ASAR scenes, that were acquired along the same track and around the same local time (7 a.m.) over the study throughout the time period of 1996 to 2009. We computed the interferometric phase due to topography and removed it from the observed phase using a high spatial resolution (10 meters) digital elevation model (DEM) that was extracted from data provided by the Saudi Geological Survey (SGS). We adopted one of the advanced multi acquisition INSAR techniques termed short baseline subsets (SBAS). This technique was selected given the limited interferometric sets available for the study area and its capabilities in resolving the atmospheric artifacts and the uncompensated topography. Findings from the SBAS includes the following: (1) mapped landslides range in size from 0.6 km2 to more than 4 km2, (2) many failures occur along the edges of hilly mountainous areas within the study area, and (3) areas that are subjected to successive failures and active erosion show continuous displacement away from the satellite, whereas the surrounding areas often show a positive signal indicative of progressive sediment accumulation of the displaced masses. The radar based observations are supported by observations acquired in the field and others extracted from remote sensing data sets. For example, many of the radar-based landslide locations were found to correspond to field verified landslide locations. Results indicate that C-band ERS/ENVISAT data can be successfully used to map and monitor mass movements in the Red Sea hills and points to potential applications of these techniques to develop an early warning system in the area. Even though InSAR cannot be considered a replacement for field based monitoring of active landslides, it provides a significant complementary dataset.