Evaluating subsidence in the Nile Delta, Egypt using Persistent Scatterers Interferometry
Bouali, El Hachemi¹; Sultan, Mohamed¹; Becker, Richard²; Cherif, Omar³; Zahran, Khalid⁴; Chouinard, Kyle^1
1Department of Geosciences, Western Michigan University, UNITED STATES; ²Department of Environmental Sciences, University of Toledo, UNITED STATES; ³National Authority for Remote Sensing and Space Sciences, EGYPT; ⁴Geodynamic Department, National Research Institute of Astronomy and Geophysics, EGYPT

The Nile Delta of Egypt is being threatened by two continuous and relatively slow hazards: encroachment from the Mediterranean Sea as sea levels rise and land subsidence of the delta itself. These two additive phenomena – due to natural and anthropogenic forces – will create a dire situation for the future of the population of deltaic inhabitants, particularly along the Mediterranean coast. The magnitude of sea level rise has been actively monitored in stations across the Mediterranean Sea, while the subsidence of the Nile Delta, as a whole system, has not been adequately quantified; instead, previous studies have focused on measuring local ground movements.

One purpose of this study is to measure the second half of the equation: subsidence rates across the entire Nile Delta. The overall study objective is three-fold: (1) to measure and map the spatial and temporal variations in subsidence rates across the entire Nile Delta, (2) to identify the nature of factors (natural and/or anthropogenic) controlling modern subsidence rates with the use of third-party datasets, and (3) to predict the deltaic response to these forcing factors along with the impact of sea encroachment over the next century. To investigate subsidence we have employed the Synthetic Aperture Radar Interferometry (InSAR) technique, specifically Persistent Scatterers Interferometry (PSI). An extensive dataset of 23 ERS-1, 157 ERS-2, and 106 ENVISAT Single Look Complex (SLC) scenes acquired between 1992 and 2010 have been received from the European Space Agency. These 286 scenes cover four descending tracks (#164, #207, #436, and #479) and four ascending tracks (#200, #243, #429, and #472). Multiple combinations of these scenes have been processed to validate and optimize results.

The Nile Delta is highly populated, yet quite large and densely vegetated, which creates a complex system of high-coherent and incoherent pixels. Therefore, subsidence of the Nile Delta may be measured within its urban centers – where PSI works well – and trends can be mapped across the regions of dense vegetation by interpolation techniques. This limitation of PSI can become beneficial: the Nile Delta is too large and complex to adequately describe subsidence rates with one large range of values – as has been done in previous studies – and so constrained coherent pixel locations force a local-scale focus, which requires a detailed analysis of the delta in a piecewise fashion. Therefore the approach is to calculate subsidence rates on a city-by-city basis: (1) the relevant urban areas are chosen and the combination of full SLC scenes is cut to an approximately 25 square kilometer area; (2) the scenes are processed multiple times, using difference scene and input parameter combinations (spatial baseline, temporal baseline, Doppler centroid differences) in order to best optimize Persistent Scatterer (PS) abundance and ground displacement calculations; (3) results are then compared to field geodetic GPS data, from which true ground motions is calculated, in order to calibrate and validate PSI results. Time-series displacement graphs are created over areas of interest, which allows for high temporal resolution, where ground motion is observed between each input scene, and high spatial resolution, where subsidence rates observed at a city-scale can be compared to rates across the entire Nile Delta.

Results indicate the following: (1) optimum subsidence rates of high coherent pixels were observed over urban areas; (2) cities proximal (<20 km) to the Mediterranean coast (i.e., Damietta and Port Said) display larger average subsidence rates (as much as 8-10 mm/yr) than those more distal (>20 km) to the coast (i.e., Mansoura and Al Mahallah Al Kubra, with rates as low as no subsidence); (3) cities near (<3-5 km) present-day meanders and tributaries of the River Nile, where younger sediment is located (i.e., Damietta and Mansoura) display a higher (>5 mm/yr) average subsidence rate than cities farther (>5 km) away (i.e., Damanhour) (<5 mm/year); (4) most cities show a wide range of subsidence rates (i.e., Port Said subsidence rate range is approximately 9 mm/yr and Mansoura subsidence rate range is approximately 7 mm/yr); (5) given the fact that subsidence rates for coastal cities is estimated to be as high as 8-10 mm/yr and the regional rise of the Mediterranean Sea is about 5-10 mm/yr, then the relative sea level rise is almost 20 mm/yr along the coast of the Nile Delta.

Current work include: (1) continuation of subsidence rate mapping across the Nile Delta cities; (2) statistical correlation of subsidence rates with sediment thickness and sediment age; (3) graphical representation of the spatial distribution of all the subsidence rates across the entire Nile Delta; (4) identification of subsidence anomalies and identifying the controlling factors; (5) generating a prediction model of the subsidence rates across the entire delta.

Although the Nile Delta is highly vegetated, the PSI technique is still applicable in measuring subsidence rates over a long period of time. The success of this project attests to the possibility of calculating subsidence rates across other regions of high vegetation where sea encroachment is an ever-increasing threat.