PhobosQuora
Agreed - Mars is great opportunity to hunt for a few molecular remnants of past life (all dead now), but a terrible place to live … or more accurately, DIE.
Earth plants evolved for 1.3 kW/m² Earth illumination - Mars at aphelion (1.67 AU) gets 0.46 kW/m². Indeed, the entire planet gets only 70% more total sunlight than Earth’s small Moon. Estimated 4 million cubic kilometers of impure water ice - maybe - compared to Earth’s 1390 million cubic kilometers of water.
Mars has no magnetic field radiation shield - which is probably why ancient water and atmosphere evaporated away long ago. Hence, humans must live under meters of rock radiation shielding, similar to long term habitation on the Moon. Humans evolved for 9.8 m/s² “one gee” gravity, necessary for health. Mars is 3.7 m/s² … 0.38 gee. That’s unhealthy, but the weight of the rock radiation shielding will still require considerable structural support - and the nearest steel mills are on Earth.
Humans will need to live on centrifuges to make one gee - stepped conical floor sections tilted 20 degrees from vertical, so Mars gravity and centrifugal acceleration add up to 9.8 m/s² at “human perpendicular”.
We might be able to bio-engineer crops that grow in Martian gravity, weak sunlight, and surface-level cosmic radiation, but the crops will still need pressure containment and detoxification of perchlorate-saturated Mars regolith. It would be a lot easier to create a jungle in Death Valley.
All this assumes that we first prospect most of Mars for the ores we need to make machinery and structure and make soil for plants. We know NOTHING about ore location and genesis on Mars, so learning where and how to look will take time. All that said, supporting scientists and robot builders close to Mars will help us explore Mars faster.
Phobos orbits 6000 km from the Martian surface, and has practically no gravity to complicate centrifuge habitats and shielding. A power transmission line from Phobos sunside PV to a (temporarily nightside) habitat will be 7 kilometers long - probably much cheaper than 4 hours of “night” batteries. Eclipses by Mars happen 7.7 hours apart, and last about 30 seconds.
note added after posting to Quora: Phobos circumference of 14 km and 7.7 hour period suggests a PV panel on a trolley on a conductive track around Phobos, moving at 14/7.7 km/hr or 0.5 meters per second. Phobos surface gravity is 600 microgee - the "track" could be wires on gossamer t-bar "telephone poles". Thus only one trolley of travelling PV panels is needed, but more allows redundancy. Perhaps individual trolleys can "demount" and "fall" to the surface below if they "seize up and stop". If the wire track is 5 meters high, the falling panel trolleys will hit the surface at 24 centimeters per second velocity after a 40 second fall.
Adding two comsats at the Mars-Phobos L4 and L5 Lagrange points will give a Phobos settlement line-of-sight connections to robots on more than 80% of the Martian surface (though not the poles), with a shorter path delay than a phone call from San Francisco to Japan. That rapid access can GREATLY accelerate surface exploration.
Arrival delta V to Phobos from a Earth-to-Mars interplanetary trajectory can be reduced by Mars atmosphere aerobraking to orbital capture. For a post-capture orbit periapsis of 3500 km and apoapsis at Phobos radius 9500 km, Phobos “landing” delta V is approximately 600 m/s … which can be spread out over many slowly increasing Mars orbits over a few hundred hours.
Note that Phobos-synchronous comsats and the radio relay base on Phobos is a good idea even before a human mission. It could provide a full time telemetry link for Mars surface robot explorers all over non-polar Mars, though the surface robots might need faster rotating telemetry antennas to track the comsats (moving prograde twice per Martian day, rather than retrograde to Earth once per Martian day).
There are many more reasons to consider Phobos as a “Mars neighborhood destination” at my web page: PhobosPhirst.
After pondering such alternatives for decades, I jettisoned most of my science fiction collection - too unimaginative and non-mathematical. I’m an engineer - I can do for a dollar what any damned fool can do for two, a science fiction author for five, and politically-handicapped NASA for twenty. Elon Musk is not a damned fool; I presume his actual plans are closer to what I write about than what he publicly talks about.