Where NOT to land humans on Mars

Sunday , 2, August 2015

Earlier this month I saw an announcement (PDF) passing by in my email inbox for a scientific workshop with quite an unusual and unexpected subject: the First Landing Site/Exploration Zone Workshop for Human Missions to the Surface of Mars. This workshop will be held at the Lunar and Planetary Institute (LPI) in Houston, Texas from October 27 to 30 this year and is meant for teams who specialize in either Mars science, engineering or In-Situ Resource Utilization (ISRU). The conveners of the workshop are willing to help link teams to bring these specialties together. To attend the workshop you need to submit an abstract proposing a site of your preference. The 25th of August is the deadline for submitting an abstract.

I know about the various landing site selection workshops held for the Curiosity rover on Mars; slowly the list of locations were narrowed down to one: Gale crater. The same is currently happening for the future ESA ExoMars rover mission. But selecting for a human landing site is a different story, and an exciting one! It inspired me to write this article about where NOT to land humans on Mars.

Figure 1. Exploration Zone Layout Considerations

The announcement showed an example of what to look out for when proposing a site (Figure 1). The landing site, with an area of around 25 km², is only going to be a small part of the total Exploration Zone (EZ). The EZ has a radius of 100 km, with the landing site at its center. This therefore means that the astronauts are not expected to travel farther away from the habitat than 100 km in all directions1. Within the EZ there should be various science and resource related regions of interest (RIOs). These locations need to be accessible by the astronauts from the habitat at the center of the EZ. It is to be expected that the science ROIs will be visited multiple times, by multiple crews from different missions and therefore the resource ROIs need to contain enough material to be used for multiple missions.

The conveners prefer a submission of an EZ with multiple science ROIs and at least one resource ROI. Reading through the supplemental background information it becomes clear that water is the number one resource of importance. Next in line are metals, silicon, and structural building materials. But before we can start talking about EZs and ROIs the whole endeavor depends on a safe landing site at the center of the ROI. Figure 2 shows the characteristics of such a landing site.

Figure 2. Characteristics of a plausible Mars landing site.

In this article I will go through each step of Figure 2, more and more narrowing down the possible landing sites. I will conclude by showing an overall picture of which areas are best to avoid.

Latitude and elevation

The landing site needs to be located relatively close to the equator, within 50°N and 50°S, and with an elevation lower than 2 km above the Martian zero elevation level. I used elevation data from the Mars Orbiter Laser Altimeter (MOLA) on board the Mars Global Surveyor orbiter2. The results are visualized in Figure 3 which shows a map of Mars between 50°N and 50°S with the regions in red which are higher than 2 km in elevation. As can be seen, already quite a lot of the Martian surface is going to be off limits.

Figure 3. The map shows gray scale Mars Orbiter Laser Altimeter (MOLA) elevation data, background data between 50°N and 50°S. The red regions have elevations higher than 2000 meter. The map was made using ESRI ArcGIS 10.2. Click on the map to open a high resolution version.

Slope of the terrain

The landing site needs to have a slope of less than 10°. Again MOLA data was used. Figure 4 shows those regions which have steeper than 10° slopes (in red)3. These regions are among the most awesome locations on Mars. Think about 10 km high escarpments with breathtaking views, the landscapes which make Mars the awesome planet it is. They will hopefully one day become part of the collective human experience. Maybe it will be possible to include a breathtaking escarpment at the perimeter of the EZ, but definitely not too close to the landing site. Nevertheless, we will probably first see these views through the eyes of future robots like these ones. And then of course in 3D through virtual reality, which is a long term future goal of the people from SpaceVR.

Figure 4. The map shows in red the areas where the slope is larger than 10°. Regions fully encircled by larger than 10° sloped terrain (for example certain impact crater floors) were considered off limits too and are also shown in red (the ArcGIS Union tool was used). The background map is similar to Figure 3. Click on the map to open a high resolution version.

Landing hazards

There are a variety of landing hazards which should be avoided as best as possible: craters, rough terrain and dunes.


Large and/or closely concentrated craters need to be avoided. The Figure 4 slope map already filters out craters with fully encircling rim slopes larger than 10°. It is important to note that a crater larger than 200 km in diameters is itself a possible EZ location. Big craters were chosen as landing sites before: the Gusev crater (166km in diameter) for the Spirit rover and Gale crater (154km in diameter) for the Curiosity rover

For this exercise, however, I will only exclude craters between 1 and 6 kilometer in diameter, see Figure 54. So why exclude craters between these sizes? I used the crater counting dataset by Robbins and Hynek (2012) which doesn’t go further down in size than 1 km craters. And I chose the 6 kilometer because when you take the 25 km² landing site area to be a circle, it has a diameter of 5.6km. I rounded that up to 6 km5.

Figure 5. The map shows in red the Robbins and Hynek (2012) craters between 1 and 6km in diameter. The background map is similar to Figure 3. Click on the map to open a high resolution version.

Rough terrain

The region shouldn’t be too mountainous, broken or chaotic. Although the Figure 4 slope map already filtered out a lot of this type of terrain I thought it would be best to include roughness measurements as a proxy of how mountainous, broken or chaotic the terrain is. Now it starts to become more complicated, because how do you define roughness and when is the surface too rough?

There are two studies which led to roughness measurements of the Martian surface, both using the MOLA data, but looking at different scales. First there is the work by Kreslavsky and Head (2000). Their data show surface roughness at three scales: 0.6 km, 2.4 km, and 9.2 km. In the paper they propose their own roughness measure which uses the slope of the MOLA data but removes the possible higher scale ’tilt’ of the terrain. The second study is by Neumann et al. (2003) who calculated the roughness at the scale of 75 meter using the width of the MOLA laser pulse6.

For this exercise I decided to only use the Kreslavsky and Head data. Figure 6, in red, shows the locations where the 0.6 scale, or the 2.4 scale, or the 9.6 scale show the highest roughness values7.

Figure 6. The locations in red where the Kreslavsky and Head (2000) roughness data at 0.6 scale, or the 2.4 scale, or the 9.6 scale is high (255).


The United States Geological Survey (USGS) published a map with the dunefields of Mars. Figure 7 shows these dunes in red.

Figure 7. In red the dune fields of Mars derived from the Mars Global Digital Dune Database.

Deposits of dust

Thick deposits of fine-grained dust need to be avoided. Dust can be determined by looking at the thermal inertia and the albedo of the terrain. Dusty regions on Mars show an extremely low thermal inertia and a high albedo. As with the roughness measurements, how do you define extremely low thermal inertia and high albedo?

Putzig and Mellon (2007) used data from the Thermal Emission Spectrometer (TES) on board the Mars Global Surveyor orbiter to derive thermal inertia. Ody et al. (2012) used data from the OMEGA spectrometer on board the ESA Mars Express orbiter to derive albedo. By observing the data histogram I decided upon a threshold value for both datasets: <100 for the thermal inertia and >-8000 for the albedo8. Combining both datasets resulted in Figure 8 showing the dusty regions in red.

Figure 8. In red the dusty regions of Mars.

The combined result

Figure 9 shows the results of Figures 3 to 8 combined (in red). Removing the high altitude regions and dusty regions definitely had the biggest effect, but it seems that there is still a lot of terrain suitable for a human mission to Mars. Of course this is just a first attempt to show which data is available and how a landing site selection can be performed using a GIS. Next, using better, more quantitative estimates, the combined result can be enhanced, especially the dust and roughness estimates. I would therefore like to ask those interested to get in contact with me so we can further narrow down the possible landing sites.

And once ‘where NOT to land on Mars’ has been sufficiently determined the next step is to start narrowing down the possible EZ locations with its various ROIs. Here, I would assume, it is key to combine data sets which favor a specific location, like the map of hydrous minerals provided by ESA. I’m currently looking into this step too but that is going to be the subject for a next article. To be continued.

Figure 9. The combined result summing up Figures 3-8