Etude par imagerie radar des pollutions pétrolières

Abstract

Every year, up to a fifth of Russian oil production is lost through leakage. The largest oil spills occur in the Russian North where permafrost is often present. Pipelines are subject to corrosion and cryogenic processes. The risk of rupture increases consequently. The oil spill monitoring is limited by the vastness and the frequent inaccessibility of the pipeline network and therefore, requires remotely sensed data. This work on Usinsk’s disaster, which occurred in 1994, aims at giving some low-level tools for analysing SAR data. Some complicating factors such as the characteristics of the sensor, the sensor look direction, the topography and the speckle, make the SAR data processing difficult. The determination of the characteristics of the target depends on the knowledge of these disturbing effects. This study shows the key role that these corrections play for the interpretation of multitemporal SAR backscattering. The method developed here for the multitemporal study of the Usinsk’s oil spill using 3 SAR images is discussed especially in terms of limits and possible uses as a routine.

Keywords

• pollution
• pipeline
• Arctic
• hydrocarbon
• radar remote sensing