The colonization of Mars can be planned and performed in various ways. This article wants to line out basic colonization strategies with the final goal to establish a sustainable, self reliant Martian colony, that can exist and even thrive independently from Earth.
 Aspect of physical independence
The long term maintenance of complex equipment requires a huge number of persons. At a minimum, they will need to replace or repair critical components, such as life support, medical technology, food production, etc. It is hard to imagine that this can be done without electronics and chemistry. At least some technology must be maintained, for the Martian environmental conditions do not allow people to live naked on Mars. So, there is a critical mass for the number of persons in an autonomous colony.
Even if fully grown, a Martian colony is not considered a closed system without any input or output from and to Earth. It is rather an independent sovereign state, fully in control of its destiny. In that regard, it need not produce all of its needs locally. Even on Earth, no sovereign state would think of eliminating all trade with other nations. However, such a Martian colony can not perform a trade volume that is comparable with any state on Earth, because the shipment costs are bigger by several orders of magnitude. Therefore, the interplanetary commerce between Earth and Mars will be reduced largely to data and services that can be transmitted via a radio link.
 Strategy 1: Independence step by step
An initial colony could start with a few persons. More colonists arrive later. In the beginning it does not supply all of its needs locally. Until critical mass is attained, the settlement will need to buy certain advanced technology. Interplanetary commerce is part of this strategy. It allows starting much simpler and earlier.
 Strategy 2: Independence at once
Due to the risk of an interruption of the colonization program, this strategy aims at the full independence from the very start. The first settlement is built in a very spartan, but nonetheless sustainable way, with all vital supplies produced locally. This first settlement is constructed remote controlled and is fully functional before the first group of settlers head for Mars.
Spartan technology (and hence spartan standard of living) can reduce the critical mass. The inevitable food production is the most critical part. If that can be accomplished with simple technology, the critical mass could be small enough to gain independence at once. However, this includes mining and processing of all needed materials from local resources.
 Aspect of transport and development
To get an idea of the transport costs of a physically independent industrial infrastructure, the current industrial infrastructure on Earth may be estimated as 1 billion workers and 100 tonnes of structure, equipment, and spare parts per worker -- round the total mass budget to 100 billion tonnes. It currently costs $200,000 to land a kilogram on Mars. Additional infrastructure is required for Mars (e.g. pressure vessels and agricultural illumination systems), so double the infrastructure required to 200 tonnes per worker. That comes to 440 million trillion dollars. To reduce this cost by one or two orders of magnitude by creative selection of industrial equipment and workers is probably easy: some of Earth's industry is redundant in terms of self-sufficiency and thus required only for a population of billions. One or two orders of magnitude drop in transport costs may also be possible in the long term. But this only reduces the cost to at least 44 thousand trillion dollars. To reduce these costs to a reasonable sum, i.e. to the range of tens to hundreds of billions of dollars, requires radical reduction in the size of the industrial infrastructure required, which requires radical redesign of the technology (Strategy 1), or it requires further radical reductions in transport costs (Strategy 2), or a combination of both.
 Strategy 1: Minimum transport and intelligent self development
Shipping costs are probably lower for small scale machines than large scale machines, and the financial frame will always be tight. The perfect, but unrealistic, way to colonize Mars is sending a one-kilogram probe with a handful of nanobots, preparing the whole colony, before sending a second one-kilogram probe with a handful of frozen fertilized human eggs, etc. This science fiction scenario is, of course, not realistic, but can serve as an ideal to strive for: minimize the mass and volume that needs to be launched from Earth, both initially and on an ongoing basis, by maximizing the self-sufficiency of Mars' industrial and agricultural infrastructures. This probably requires a radical redesign of almost every piece of equipment, and a radical rethinking of industrial infrastructure generally. Lo-tech (see e.g. pneumatics, hydraulics), small-scale-tech (see e.g. blacksmith, brick, glass), and flexible tech (see e.g. 3D Printer) are promising approaches.
 Strategy 2: Mass transport of ready-to-use technology
A colony needs large machinery for life support and further expansion. All machinery is shipped from Earth to Mars. Plans can be developed for massive colonization ships moving in repeated transfers between Earth and Mars without stopping. Only the cargo and passengers start and stop. Sending a complete industrial economy to Mars is theoretically possible. It just takes a long time, a launch volume much higher than current, or some combination of the two. See cost estimates above. For example, we might spend $100 billion per year for 4.4 million years to set up an independent Mars colony using the same industrial equipment and global-scale economy as Earth, by simply transporting all needed people, structures and equipment (minus one or two orders of magnitude for creative selection of a subset) over this period of time, without any redesign except to account for Martian environmental conditions (low gravity, near-vacuum, etc.) To be anywhere close to being viable, this strategy requires extremely radical reductions in transport costs, possibly from using ISRU-based propellants, suborbital reusable launch vehicles (RLVs) combine with tether-based orbital momentum transfer, and many other theoretically possible strategies. But then again, how can we do ISRU on such a massive scale without lots of equipment already in space? Catch-22.
 Strategy 3: Develop industry on Luna to support Mars colonization
This is inspired by the 1970's idea of Gerard O'Neil to build massive solar power plants out of lunar materials. Developing Luna first can take advantage of Luna's unique advantages to provide a massive Mars colonization effort that may not be possible otherwise.
Luna has a lower gravity well than Mars or Earth and a high vacuum at surface level. Because of this there is a potential for a low cost electric launching of cargo from the moon to orbit that Mars and Earth will never be able to match. The cost was estimated as "pennies-per-pound" on the Mike Combs Space Settlement web site FAQ, the "Why not build solar power stations on the moon?" section. The actual cost per pound of such cargo launch will be heavily dependent upon the size of the market for launching cargo.
Luna has one face constantly turned toward earth so that a stationary antenna there could provide continuous real time communications with less than a three second round trip delay. This allows remote control industrial development from Earth without developing artificial computer intelligence to operate the industry 45 minutes without instructions on the latest developments as would be necessary for remotely controlled industry on Mars. The axial tilt with respect to the ecliptic is 1.5 degrees. So, there is a potential in the course of development of Luna's resources to establish a set of solar power stations within 46 kilometers of either pole linked in a grid such that there is always one of them in sunlight. The ambient vacuum makes efficient thermal insulation as easy as packing fine grains sifted from the regolith. The big advantage of industry on Luna is that it does not need to be a complete infrastructure that can support human beings for every task. Remotely controlled machines are being used on Earth with a growing range of complexity and function. If as much effort were put into adapting machines to a lunar environment as has been put into adapting machines to an undersea environment, the designs for machines to develop lunar industry would likely be in hand. People on Luna will be few and do only those things for which people are especially capable. Association with Luna can bring Mars into the market for building structures in Earth orbit before there is even a colony on Mars. Building an industry on Luna will be handicapped by the scarcity of volatiles on Luna but this can be remedied by establishing automated industry on Mars to distill the atmosphere, mine the permafrost at 50 degrees latitude and ship the products to orbit with one system and on to Luna with another. This is probably a simpler problem then building a complete industrial base for a colony, and it would be paid for out of returns from building structures in Earth orbit. Industry started on Mars without Luna would not be able to sell solar power plants to Earth to pay for the investment. It would have to rely on selling hypothetical patent rights of Martian colonists' inventions, gold shot out of an electric cannon to Earth if any is found on Mars, real estate sales and such.
Benefits to Mars colonization that would result from full industry on Luna would include the ability of Luna to ship tanks of liquid oxygen to low Earth orbit for 40 cents a pound to use as rocket fuel. This low cost of transportation would allow considerably higher production costs for oxygen on Luna than on Earth while still having a much lower total delivered cost. The same is true for materials to construct a Mars colonization ship in orbit and a space based solar power plant to be sent to operate in orbit of Mars. If industrialization of Luna is successful, these benefits could be provided to Mars in trade for volatiles to Luna more cheaply than they could be provided by any conceivable arrangement of trade between one region on Mars and another.
Some criticisms of this approach:
Removing the task of industrial development to the moon mostly just switches the location of the problem. The strategy above is simply to move the problem of developing self-sustainable space industry, given high launch costs on earth, to a new locale then introduce trade between Mars and the new locale. But we can divide industry and trade between places on Mars too: if trade between Mars and the moon could solve the problem so could trade between different places on Mars. The supposed benefits of teleoperation are hypothetical and there is no strong reason to assume it would allow the needed radical reduction in the mass of industrial infrastructure to be launched. No designs for teleoperated machines are presented that would have this effect of radically reducing the mass of the industry that needs to be launched from earth. If they were, they would form part of the general alternative to radical lowering of launch costs, namely radical redesign of the industrial infrastructure. Automation indeed is probably an important element of the radical redesign of industrial technology that will be needed.
To get to the lunar surface from earth still costs within the same order of magnitude as to get to the surface of Mars (currently about $100,000/kg to the lunar surface vs. about $200,000/kg to the Martian surface). Indeed the industrial development problem on the moon is worse because it is scarce in volatiles which are crucial and voluminous inputs to industry. On earth where voluminous water is needed industry locates next to the water. Trying to get them out of Mars' gravity well all the way to the moon may not be economically viable, any more than reversing the course of a river by putting the water into a large fleet of trucks and driving back to the top of the mountains where the river comes from would be an economically viable way to conserve water. It is also easier to think about and solve the self-sufficiency problem on a location like Mars where volatiles and other industrial inputs are both readily available.
The mass driver upon which the effort depends is only a design. It has never been tested as a complete device. If things proceed as supporters of raising an industrial base on Luna by its bootstraps hope, it will take a long time, probably in the neighborhood of fifty years before first financial return. Importing volatiles from Mars requires that both bodies be industrialized at the same time with remotely operated equipment. That causes more complex planning requirements. Dealing only with Mars development avoids having to import elements which are scarce on Luna. Mars has all elements necessary for industrial development available in usable concentrations and quantities.
 Strategy 4: Genetically engineer crops to grow on Mars
The amount of equipment needed to be transported to Mars could be reduced if crops could spread themselves on the natural surface of Mars. The lichen which grows in extreme conditions on Earth could perhaps be modified to grow on Mars in the current conditions of the 50 degree latitude region. It could incorporate an internal antifreeze such as polyethylene glycol. Various sorts of crops could be designed to incorporate various useful substances in their tissues. When these crops have spread themselves around Mars, colonists would come and harvest them to support their colony. This might greatly reduce the need for industrial infrastructure as a prerequisite to support agriculture.
No plant or lichen genetically engineered or otherwise has ever been grown in vacuum or in near-vacuum conditions like those on Mars, and it is far from clear that this is possible. The chemical and radiation environments on the Martian surface probably also present major hurdles for this approach. This strategy would take many years to first develop crops to grow on Mars and then to plant them and let them spread. The plant released into the wild might mutate into an undesirable form that would not support a colony.
A variation of this strategy is to genetically engineer a hollow tree that can serve as a pressure vessel for smaller life forms inside it. A problem with this approach is how to protect its leaves while giving them sunlight. The coconut tree Protects its seed from the harsh environment of the open ocean with a hard case. By combining genetic engineering with selective breeding, it might be possible to develop a plant that grows a little green house for each leaf and each seed. Seeds would need to be packed with a considerable amount of food and water to get a new plant started on Mars, as coconuts are packed with food and water to start new coconut trees. Starting plants in the most Mars-like environment that they can survive, then developing plants to grow in more and more Mars-like conditions, people might be able to develop a plant that could grow on Mars.
While Earthly plants store sugars and starches in their tissues for use during times of darkness, and they store complete nutrition for their offspring in seeds; they do not store oxygen. They use it from the reservoir of the Earth's atmosphere. Plants that could grow in the ambient Martian conditions would need special tissues to store oxygen. They would also need to develop a flexible water impermeable skin with a sort of window that while remaining impermeable to water would allow carbon dioxide to diffuse through, another sort of window to allow oxygen to be excreted and another sort of window to be translucent. The roots would need to totally engulf particular grains of soil in order to absorb minerals from them without losing water. The adaptations for plants' survival on Mars would be many, but the result would be plants whose seeds are a source of food, water, and oxygen.
 Aspect of finance
Frontier settlements are capital investments from which investors expect some utility. Rarely small amounts are donated to altruistic causes (e.g. expanding humanity). Governments invest small sums in science and larger sums in national security. Most commonly, investors demand a profitable return from their investments. The sooner a colony becomes financially self-sufficient, the less investment is required, and thus the sooner an investment is likely to be made in the first place. Since radical elimination of all imports is probably impossible in the short run (see above), a colony is much more likely to be financed if it can generate exports that match or exceed the costs of imports. Since imports are costly, the exports must be valuable. They must also be affordably transportable to Earth: high value and low mass.
 Initial Capital Investments
The colony requires equipment sufficient equipment to sustain agriculture to feed and clothe the personnel, to generate exports sufficient to pay for imports, and to make at least the heavy replacement parts for the equipment itself. The colony should also preferably grow from an early capital investment rather than continuing to require massive equipment transports from earth in order to expand operations.
Treating the colony as a single financial entity, its ongoing expenditures will be equal to the cost of ongoing imports.
Treating the colony as a single financial entity, its ongoing revenues will be equal to the ongoing proceeds from exports. Profits in a given period will be equal to revenues minus expenditures in that period. In the single financial entity model the profit of the colony in a period thus equals proceeds from exports minus the cost of imports during that period.
 Net Present Value (NPV) and Internal Rate of Return (IRR) of Investments
Given a series of cash flows, for example annual expenditures and revenues over a period of thirty years, and an interest rate, the NPV function computes the net present value of these cash flows at the start of the series. It is in other words the net present value of profits to be expected by the investors. NPV should exceed the initial capital investment (or alternatively, if capital investments are counted as cash flows, NPV should be positive).
In NPV analysis risk is represented by increasing the interest rate (the "risk premium"). If risk is not fairly evenly distributed over time NPV is less accurate and real options analysis is required for both for accuracy and for designing better strategies. However for most purposes NPV is fine, and it's also easier (you can use a spreadsheet, whereas real options analysis requires more sophisticated software).
In the model treating the colony as a single financial entity, the cash flows in a period are just the proceeds from exports (positive) and the costs of imports (negative) in that period. The net present value of the ongoing exports and imports should exceed the initial investment given an interest rate that includes a risk premium reflecting the risks of the financial failure of the colony.
The internal rate of return (IRR) of an investment gives us the rate of return as an output instead of the interest rate as an input. The initial capital investment is treated as a cash flow in the first period(s).
For more information, see your spreadsheet's help pages for "NPV", "IRR", and related functions.
 Golden Mars scenario
We need concrete strategic and financial scenarios to work with. Since colonizing Mars is currently far from economically viable, we have to make some hypothetical, but plausible, assumptions, in order to develop scenarios that are financially and otherwise strategically viable. Here is one, the Golden Mars scenario:
(1) The same geological processes that formed gold ores on Earth once operated on Mars and have left there concentrations of gold on its surface not seen by humans since they first started finding the easy pieces on Earth c. 4000 BC. In particular, 1,000 kg of equipment on the Martian surface can find and ship to the Mars spaceport 10 kg of gold nuggets and flakes per year. Some technological goals to strive for: 10,000 kg of imported equipment (or 100,000 kg of native equipment, because these would be the bulkier parts) requires one person to operate and maintain it, and that person requires another 10,000 kg of equipment to support him. Some of aforementioned equipment should be made on Mars if that increases the economic return, which requires further labor, otherwise it should be imported.
(2) Because of the development of ISRU-based propellants, the costs of transport from Earth have been reduced by two orders of magnitude (to $2,000/kg) and the costs of transport from Mars surface to Earth surface are $1,000/kg.
(3) Plausible assumptions can be made about making things from Martian raw materials, as long as every part (use parts lists) and every material can be accounted for all the way back to through the supply chain (really supply tree, it keeps branching at every step) to Martian mines.
(4) The price of gold on Earth is about the same as today: $1,100/oz. * 1/28 oz/g * 1,000 g/kg = $39,000/kg. The market for gold production on Earth is about $100 billion/year, and the above-ground inventories are in the trillions of dollars, so you can produce at least $50 billion/year worth of gold before you start saturating the market and the price drops. (To be more precise about this, look up research on the supply/demand curve for gold, I'm sure economists must have researched this many times).
Can you design this Mars colony to be profitable, and thus attract investors?
 Make vs. Import Tradeoffs
 Initial Capital Investments
 Net Present Value of Investments
- Another Contin(g)ent Concept
- The Human Mission to Mars — Colonizing the Red Planet (The Greatest Adventure in the History of Humanity Edited by Joel S. Levine, Ph.D., NASA, Co-Chair, Human Exploration of Mars Science Analysis Group (HEM-SAG) of the Mars Exploration Program Analysis Group (MEPAG).; Rudy Schild, Ph.D., Center for Astrophysics, Harvard-Smithsonian In Association and Collaboration with the Mars Society