InSAR observations of incremental fault growth from the 2005-2010 Dabbahu (Ethiopia) rifting episode.
Hofmann, Barbara1; Wright, Tim J.1; Paton, Douglas A.1; Rowland, Julie V.2; Vye-Brown, Charlotte3
1University of Leeds, UNITED KINGDOM; 2University of Auckland, NEW ZEALAND; 3British Geological Survey, UNITED KINGDOM

It has long been recognized that faults develop as a result of a series of successive slip events but direct observations of growth are few and far between. Published fault growth models are therefore commonly based on statistical properties of cumulative fault displacement patterns. Here we use InSAR observations of fault slip during the 2005-2010 Dabbahu (Afar) rifting episode to test fault growth models directly. The recent Dabbahu rifting episode, which commenced in 2005 within the Afar Depression, has provided a unique opportunity to study progressive fault growth. The initial dike intrusion caused deformation along the entire length (60 km) of the segment comprising horizontal opening of up to 8 m, subsidence of up to 3 m in a 2-3 km wide graben and uplift on the flanks of up to 1 m. Displacements during the initial event and the subsequent 13 smaller dike intrusions have been observed with InSAR using data from ENIVISAT, with 4 of the dikes also being imaged by ALOS. InSAR measurements can provide accurate maps of surface deformation over large areas (10s of km) with a precision in the range of mm and a spatial resolution of meters to tens of meters. For dike intrusions the majority of the deformation is centered above the dike in the area of extensive faulting where unwrapping of the short (5.6 cm) wavelength C-band ENVISAT interferograms is very challenging. We have therefore focused our analysis on the 4 dike intrusions observed by the longer (23.6 cm) wavelength L-band ALOS. However, without precise local digital elevation models (DEM) relating this deformation to individual surface features is impossible. We therefore combine the ALOS InSAR data with a high-resolution DEM, derived from an airborne LiDAR survey, to process the ALOS data at full resolution and reduce the error caused by the relative large baseline between the ALOS acquisitions and to precisely locate slip on individual surface faults. The high-resolution airborne LiDAR survey was carried out in October 2009 covering the central section of the Dabbahu segment. The resulting DEM covers 800 km2 with, on average, one return every 4 m2 , but including areas with 1 return per 0.25 m2 . The height accuracy of the DEM is ~10 cm. We identify the slipped structures by calculating the gradient of the unwrapped deformation. Once the structures have been identified and roughly picked we apply our new algorithm to pick hanging and footwall cut-offs along the surface faults and extract their displacement-length profiles from the LiDAR. At the same time we automatically extract the incremental line-of-sight fault offsets from the InSAR data, converting these into vertical throw using an average sense of motion on the faults. By comparing these two measurements we can directly measure how faults are growing, and test models of fault growth and linkage. During each event we observe reactivation of faults along the entire length of the dike with several 10s of fault segments involved in each case. We can further see that the deformation is not just located along the obvious surface faults but that a considerable amount is located on buried structures.