Measuring Slopes of Gully Fan Apices Using Digital Elevation Models

Introduction: Martian gullies are defined by [1] to be slope features comprised of an upslope alcove, a midslope channel, and a downslope depositional fan. Due to their relative youth and similarities to waterrelated features on Earth, the Martian gullies have attracted much attention since their discovery by the Mars Orbiter Camera [1]. Many erosional agents for the gullies have been proposed including groundwater from a shallow [1, 2] or deep aquifer [3], brines [2, 4], melting snow [5, 6], melting surface ice [7-10], carbon dioxide supported flows [11], and dry granular flows [12-14]. If some of the gully-forming flows were water-rich, a major outstanding question is whether the water sourced from the sub-surface (an aquifer) or the surface (atmospherically-emplaced). One way to test whether a flow was wet or dry is to look at the channel gradient where deposition initiates. Dry granular flows will tend to deposit at ~21°, the minimum angle of kinetic friction, below which all dry granular flows decelerate (~20.7-22.9°, [15]). Debris flows (sediment-rich), fluidized by liquid water or another fluidizing agent, can deposit on shallower slopes, down to ~1.7-5.7° [16]. Sedimentation from water flows may deposit on still shallower slopes. Note that the apex slope of the bright gully deposit in the Centauri region, which formed between 1999-2004 [17], is 21°. To investigate the likelihood that liquid water was involved in gully formation, we use five digital elevation models (DEMs) derived from High Resolution Imaging Science Experiment (HiRISE) stereo image pairs to measure the channel gradient just upslope of where gully depositional fans “initiate,” i.e. the visible upslope extent of deposition, termed the “apex slope.” The 1 m/post DEMs, produced using the method developed by [18], are located in five different southern hemisphere craters. They include gullies with a range of preservation states and multiple orientations. We assume that the channels above the apices are erosional and that the apex slope is the steepest slope on which deposition has occurred, marking the transition from and erosional to a depositional regime. We also assume that changes in channel width have negligible effects on whether a flow will deposit. Previous studies have looked at slopes in and around gullies using other topographic data, including Mars Orbiter Laser Altimeter (MOLA) interpolated gridded topography data [19-21], individual MOLA topography tracks [22], and parallax measurements performed using HiRISE stereo image pairs [23-26]. These studies have looked at the average slopes within gully alcoves [19, 21], the average slopes of walls hosting gullies [20, 22], the average slopes of gully channels and debris aprons [23], and piece-wise average slopes of gullies, including their alcoves, channels, and debris aprons [24-26]. HiRISE DEMs provide the highest resolution topographic data currently available for Mars, so we select them for our study. None of the previous studies have measured the apex slope. Data: Table 1 lists the HiRISE images, processed using ISIS (Integrated Software for Imagers and Spectrometers) [27], used to make each DEM. The DEMs were created using the area-based automatic matching package of the commercial stereo software SOCET SET (® BAE Systems). The vertical precision of the DEMs is expected to be ~20 cm [18], but the quality of DEMs is limited by differences in illumination, signalto-noise ratio, and amount of hand-editing performed.