Long-Term Orbital Monitoring of Oxia Planum, Mars: Active Dust Devils at the ExoMars Rosalind Franklin Rover Landing Site

Introduction

The solar-powered ESA ExoMars Rosalind Franklin (EMRF) rover will land in Oxia Planum in 2030. Given the importance of dust-lifting events in clearing solar panels for energy production on previous in situ missions [e.g. 1,2], it is vital to understand the likely frequency of such events at Oxia Planum. A global machine learning study found “3 small and 3 large” dust devils in the Oxia region over 8 Mars years [Conway, 2025]; of these dust devils, only one intersects the landing ellipses for the EMRF rover. Our earlier study in Oxia Planum found 2 active dust devils in a single High Resolution Stereo Camera (HRSC, [3]) image, and 4 active dust devils using a ratio of 2 Context Camera (CTX, [4]) images [5]. The aim of this study is to expand previous studies to identify active dust devils in Oxia Planum using all available orbital data, to determine the spatial and temporal distribution of possible dust lifting events in the landing site for the EMRF rover.

 

Data

Our analysis covers the time period May 2006 to April 2024, corresponding to Mars Years 28 to 37. Our study uses visible wavelength image data from instruments on three different spacecraft: HRSC on Mars Express, CTX (49 images) and High Resolution Imaging Science Experiment (HiRISE, [6]) on Mars Reconnaissance Orbiter, and Colour and Stereo Surface Imaging System (CaSSIS, [7]) on Trace Gas Orbiter.

The main study site was selected on the basis of maximizing the temporal range available for identifying dust devils in a representative part of the EMRF landing site. As such, our study focuses on the center of the landing ellipses, and contains the majority (69%) of the 1s landing ellipses as positioned at the time of this study. In addition, we used the InSight mission landing site as an area to test our methods in the same time periods used in a previous study using higher spatial resolution data [8].

 

Method

The main novelty in our approach is the rigorous coregistration, and, through the use of a stereo Digital Terrain Model (DTM), orthorectification of many overlapping CTX images. At the main Oxia study site, we produced 49 co-registered CTX orthoimages. At the InSight test site we produced 2 co-registered CTX orthoimages. We first identified changes between orthoimages by creating image ratios. Additional dust devils were also identified using an image ‘flicker’ approach. At Oxia Planum, we searched HRSC (11 images), CaSSIS (37 images), and HiRISE (50 images) image data for the presence of active dust devils. For HRSC and CaSSIS, we manually georeferenced images to our CTX orthoimages, and exploited the multispectral capabilities to help identify any active dust devils (e.g. [e.g. 9, 10]). For HiRISE, we used stereo pair anaglyphs to look for evidence of change between images. Although our main goal throughout the study was to identify the presence of active dust devils, we also traced changing surface albedo features that correspond to dust devil tracks and/or wind streaks when possible [e.g. 5, 8].

 

Results

We identified a total of 37 active dust devils in the Oxia Planum study area. Of these dust devils, 32 were in CTX images, and 5 in HRSC images. We did not identify any active dust devils in our Oxia Planum study site in CaSSIS or HiRISE images. Dust devil diameters range from 30 to 200 m, with a mean and standard deviation of 110 and 48 m respectively. The velocity of dust devils identified in HRSC images ranged from 10 to 23 ms^-1, and were moving in roughly NE or SE directions. Dust devil activity in Oxia Planum was concentrated in two seasonal periods, peaking at both L_s ~ 15° and 180°. Peak activity occurred in Mars Year (MY) 33 and 36, with 13 and 14 dust devils respectively. We identified almost 800 new dust devil tracks across 2 different time periods, which had mean directions of 92° and 94°, and median lengths of 1.9 and 1.8 km.

Figure 1. Examples of active dust devils identified in Oxia Planum in CTX images.

 

Implications

Using the mean dust devil diameters, and all track lengths from 2 different time periods, we can estimate the formation rate of dust devils in Oxia Planum. Inverting this formation rate gives us the clearing interval, as calculated at other landing sites on Mars [e.g. 11]. We estimate solar panel clearing intervals for the EMRF rover of 3708 to 4218 sols. For comparison, Spirit experienced a solar panel clearing interval of 100 – 700 sols [1].

 

References: [1] Lorenz R.D. & D. Reiss (2015) Icarus, 248,162-164. [2] Lorenz R.D. et al. (2021) PSS 207, 105337. [3] Jaumann R. et al. (2007) PSS 55, 928-952. [4] Malin M.C. et al. (2007) JGR 112, E05S04. [5] Favaro E.A. et al. (2021) JGR 126, 2020JE006723. [6] McEwen A.S. et al. (2007) JGR 112, E05S02. [7] Thomas N. et al. (2017) Space Sci. Rev. 212, 1897-1944. [8] Perrin C. et al. (2020) GRL 47, e2020GL087234. [9] Stanzel C. et al. (2006) GRL 33, L11202. [10] Rangarajan V.G. et al. (2023) Icarus 394, 115443. [11] Reiss D. & R.D. Lorenz (2016) Icarus 266, 315-330.