Assessing Sensitivity of Hyperspectral Sensor to Detect Oils with Sea Ice

Abstract

The reflectance of two commonly used oils, crude oil and diesel, is measured under various conditions: on a water surface, among pack ice, and on/beneath compact ice. The spectral characteristics of each oil are analyzed using the results from these measures. In conjunction with estimated noise thresholds of the sensor environment, the theoretical potential to identify oil is assessed for the hyperspectral Hyperion. The hyperspectral sensor is more sensitive to the crude oil than to diesel under all conditions. The visible and infrared bands, from 468 nm to 933 nm, are more suitable to identify the crude oil. In addition, when the background is pack ice, the infrared region from 1134 nm to 1326 nm is another potential useful zone. Through the visible-to-infrared bands, the sensitivity to the existence of diesel is inferior to that of crude oil. Relatively, the bands greater than 1134 nm have the potential to separate diesel from the water or sea ice. These characteristics and sensitivity of oil film in terms of ice and oil type can be effectively used to select suitable bands to distinguish oils from sea water and sea ice.