Searching for the sources of ice block falls at the martian north polar scarps

Abstract

We searched for the sources of ice block falls at the steep scarps of the martian North Polar Layered Deposits (NPLD). We focused on two study areas at the NPLD, i.e., Scarp 1 (centered at 85.0°N, 151.5°E with a length of ∼20 km) and Scarp 2 (centered at 83.8°N, 235.5°E with a length of ∼12.5 km). Repeat High Resolution Imaging Science Experiment (HiRISE) observations with image scales of up to 0.25 m/pixel were used. Specifically, we identified changes in the shadow patterns at the scarps which attest to block fall activity. An automated change detection method, which searches for the grayscale difference between multi-temporal images, was developed to identify these patches where shadows have changed. The method was evaluated in six areas of Scarp 1 over all detection periods, and achieved an average true positive rate of 97.6% and false discovery rate of 9.4%. The spatio-temporal distribution showed that ice block falls were an active process. Along Scarp 1, about 129 ice block falls occurred per Mars year per km2, mostly at elevations 0–300 m of the 500 m high scarp, where the slope is over 30°. The amount of ice block falls increased with steeper slope angles. Along Scarp 2, about 237 ice block falls occurred per Mars year per km2, resting in the form of smaller blocks and fine particles, or spreading outward in fan-shaped patterns. We compared the detected source areas at Scarp 2 with the fresh ice blocks observed on the underlying Basal Unit (BU) between MY 30 and MY 31 to analyze the dynamic process of ice block falls. Underneath the areas with large numbers of detached ice-fragments, many fresh ice blocks were found on the BU. On the other hand, some regions of Scarp 2 did not show evidence of detached ice-fragments despite fresh ice blocks on the BU areas below. Our comparison proved that both the NPLD and the BU scarps are active source areas of ice block falls. The detections of block fall sources at the NPLD scarps can help to distinguish events from the NPLD and those from the BU and refine the estimates of active polar scarp retreat, which will improve the understanding of the polar cap evolution.