Overview of Active Measures to Mitigate Global Warming.

A lot of science points to the fact that human created global warming (GW) is real and accelerating. Currently there is a lot of talk about passive methods to reduce GW. Passive methods mostly involve reducing the amount of CO2 and methane emissions. But if GW gets sufficiently bad, we may need to explore and implement active methods to reduce GW. Active methods are sometimes also called “climate engineering” and it’s a quite controversial topic for various reasons. The main idea I want to outline here is that various technologies are progressing at a rapid rate. As these technologies get better and better they could work together more powerfully in ways that could actively (not passively) mitigate GW.

One of the key technologies that allows us to research GW is computer climate models (CCMs). As computer hardware and CCM software gets better and better we’ll gradually have more confidence in our ability to model what will happen to the climate under various scenarios. For example, one scenario is that our carbon dioxide (CO2) output goes up. Another is that our CO2 output goes down. CCMs can give insights as to what may happen to the future climate under these two different scenarios. The CCMs and the hardware they run on are both getting better all the time. That means we’ll have increasing confidence in the models because (as they improve) they will give results that increasingly match historical data (both recent climate data and more ancient data).

The key concept here is with better CCMs we can see what happens if we take active measures to mitigate GW. This is very important because if GW starts to accelerate then we may decide that we need active measures to mitigate it. If we take active measures we need to clearly understand the effects of these measures so we’re going in the desired direction. So very good CCMs are like the headlights on a car. Only with the headlights on can we see the road ahead and drive safely at night. Good CCMs are also like using GPS to successfully drive thru a complex city. The destination is a global climate that supports the simultaneous flourishing of the natural environment and human societies.

CCM’s will of course never be perfect predictors of future climate, but they’re undoubtedly getting better and better.

Climate engineering is the science of how to change the climate in a controlled scientific manner for a desired goal. More specifically, it is the science of how to change the climate so GW doesn’t get out of control and seriously harm our global environment. So some climate engineering scientists explore things we can do to change the climate in a way to slow down or reverse GW.

In one sense we’re already doing climate engineering on a massive scale by releasing so much CO2 and methane into atmosphere. The important point is this unintentional climate engineering is already being done without full consideration of the negative consequences. Every time you drive your car you’re actively participating in a climate engineering experiment entitled “Let’s see what happens when we pump massive amounts of CO2 into the atmosphere in a very short period of time.”. So don’t think that climate engineering by humans hasn’t been done before, we’re doing it now on a massive scale.

The most obvious measures to mitigate GW are to reduce the amount of fossil fuels we’re burning and work to make human societies more energy efficient and use more renewable energy (mostly solar and wind power, and possibly fusion in the future). Hopefully we’ll soon work to take these obvious steps. But these measures may not be enough. These measures don’t reduce the amount of CO2 in the atmosphere, they just slow down the rate at which we’re emitting CO2. This is an important point. These are passive measures, not active measures. There have been various global agreements to reduce GW, and the efforts of the world’s nations to work together on this issue is great, but the agreements have done very little to actually slow down worldwide CO2 emissions. Why? Humans are apathetic and big, powerful global corporations get massive profit from the status quo.

The biggest reason why active climate engineering may be needed is that we may already be beyond the point of no return. It may be that we could totally stop tomorrow putting any CO2 into the atmosphere and it wouldn’t make much difference. We could already be locked in to massive GW even if we stop all CO2 emissions tomorrow. That’s why we may have to do active climate engineering to prevent massive GW.

To be clear I am 100% supportive of passive measures to mitigate global warming. Reducing use of fossil fuels and making vehicles and buildings more energy efficient is great. Using more renewable sources of energy is fantastic. But these passive measures may not be enough.

So here are some active measures we can take to slow down and reverse GW. These ideas have been around for a long time and have been researched by scientists, but it’s helpful to introduce them so there’s an understanding of our options to mitigate GW.

1. First Active Measure: Stratospheric calcite aerosols and other alkaline metal salts. These aerosols reflect sunlight. They would be injected into the stratosphere to reflect some of the incoming sunlight back out into space. This would reduce the amount of sunlight reaching the surface of the earth. This would have an overall cooling effect on the planet, hopefully helping to reverse the melting of the poles, glaciers and ice shelfs. Since they’re alkaline they could also help strengthen ozone layers because they would help to remove acids from the atmosphere. Earlier concepts involved using reflective acidic aerosols but the potential side effects are too much (harming of ozone layers, ocean acidification, etc.).

This is not an active measure in the sense of capturing CO2, but is it is active in that we would deliberately increase the albedo (reflectivity) of the earth.

2. Second Active Measure: Phytoplankton blooms. Phytoplankton are very small and very abundant plants that float in ocean currents. About 50% to 80% of our oxygen comes from phytoplankton. They need certain amounts of minerals to thrive. In large parts of the ocean there is not enough dissolved minerals in the seawater to support a large population of phytoplankton. The idea is to spread a fine powder of these minerals (like fine iron dust) onto the surface of the ocean. Then the phytoplankton will rapidly multiply (bloom) and capture a great deal of CO2 as it does (just as plants on land consume CO2 and give off oxygen). Thus this would reduce the amount of CO2 in the atmosphere. After the phytoplankton use up the minerals they will either die and float to the bottom, or get eaten by other sea creatures large and small. Marine scientists have researched how this can be done in a way that is beneficial to the ocean environment.

It goes without saying that both of these active measures need more research and, ideally, small experiments would be done to verify their efficacy, safety and side effects.

Before I discuss the 3rd active measure I need to point out that computer artificial intelligence (AI) is rapidly getting better and better. The same is true with robotics. There is a field of AI called Deep Learning that uses “neural networks” to solve problems that computers could not solve before. In a very simplified fashion, neural networks mimic the layers of neurons in brains. Deep learning uses neural networks and other techniques to solve difficult problems that can’t be solved with traditional computer programming techniques. For example, it is deep learning systems that help self-driving cars to operate safely on public roads. Deep learning allows computers to teach themselves (program themselves) to solve difficult, real-world problems without the need for people to write large computer programs. Other examples are how deep learning makes possible near real time translation of Skype calls, and very accurate recognition of tumor growths in mammograms. Many other examples exist.

There are different types of deep learning techniques, and researchers are exploring how they can be combined synergistically to perform even more complex tasks. Just like the brain has different “modules” that perform different tasks. In addition to performing different tasks, the modules communicate with each other so the whole system is more effective than just the sum of the various modules.

Deep learning is now being applied to improve the predictive power of CCMs. Deep learning techniques are an intense area of study by computer scientists. Deep learning methods are improving rapidly and will undoubtedly help CCMs become more and more accurate.

Also robotics (machines that can move and operate autonomously in the real world) are getting better and better. And, of course, the most important part of any robot is the AI that is controlling the robot. The AI is what makes the robot autonomous and truly useful and productive. Advanced AI and deep learning will allow robots to do tasks that only humans can do now. For example, teams of advanced robots could do labor intensive jobs like build energy efficient homes or cultivate large organic farms. Organic farming is wonderful but it is very labor intensive.

Some people may laugh and say such robots are decades away, but from what I can see they will arrive sooner than we think. I would say they’re already here in the form of self driving cars that are safer than human drivers (at least in limited road conditions). A self driving car is just an autonomous (or semi-autonomous) robot that looks like a car. And self driving cars need a lot more development before they can handle all road conditions as good as a human.

So how would a humanoid robot learn to successfully manipulate objects in the real world? One simple way is to have humans wear an instrumented body suit. The suit would then accurately record the detailed motions of many different tasks. After the motion capture data set gets large enough, it’s fed into a deep learning system. The systems learns the motion patterns and use the knowledge to control a humanoid robot in real world tasks. Research labs are exploring this method. Some of the obvious tasks are to move objects in a warehouse, clean a house, help with the elderly, harvest crops that need to be harvested by hand, and many more.

Another way is to have a deep learning system view some of the billions of YouTube videos that are available. This, of course, is a huge data set that contains many examples of all possible human motions and activities. The deep learning system watches the videos and slowly learns to understand the motions, the intent of the motions, and how to control a humanoid robot to successfully do the desired tasks in the real world.

Likely a combination of both methods (and other methods) will be used to teach humanoid robots how to successfully manipulate objects in the real world. The research is in the early stages.

Just as deep learning systems in a short period of time mastered and beat human world champions at the extremely complex game of Go, so to this approach would likely lead toward successful humanoid robots in a relatively short period of time. An example of a “successful humanoid robot” is one that could load and empty a dishwasher, properly fold a load of laundry, vacuum the whole house including the stairs, harvest produce from a garden or farm, and (most importantly) understand the verbal instructions give to it by a human.

It’s very important to understand that these robots do not need to be sentient, conscious or self aware like humans. They will be built with enough intelligence to perform the desired simple tasks. This level of intelligence does not require sentience, self awareness or consciousness. If we now have robot self driving cars that can drive more safely than humans then why not robots that can do much of the simple physical labor that humans now do? Self driving cars that drive better than humans have no need at all for self-awareness or consciousness. Same with autonomous humanoid robots. How many humans will want to do the back breaking work of organic farming or reforestation if it can be done more efficiently by advanced robots?

An important point is that one of the biggest shortages we have is the shortage of intelligent, cheap labor. For example, why don’t we have more organic farms? Because they’re very labor intensive and paying for human labor is expensive. Intelligent humanoid robots could help make organic food abundant and much cheaper. Many people would love to operate an organic farm, but when they learn how much hard labor is involved, many drop out.

Keep in mind the economic incentives and opportunities for robots that can do human level work are massive. Many existing companies and smart entrepreneurs realize the huge economic benefits of human capable robots. So I can easily imagine the scenario where the amount of research and investment money going toward this will increase dramatically. Deep learning has recently triggered great advances in AI, and I see the scenario where rapidly advancing AI will help human capable robots come to market faster than most people realize. Imagine if Amazon or Wal-Mart could have human capable robots handle much of the massive amount of physical labor needed to run their operations. Thus you can easily see the economic incentive to pour money into researching this technology.

Two other technologies getting better and better are solar power and large scale water purification plants (LSWPP). Solar panels are getting cheaper to build and more efficient. The same is true with LSWPP (turning seawater into fresh water). For example, Israel is now getting a large portion of its fresh water from LSWPP. Israel is now the only country in the middle east with a stable and sufficient supply of fresh water. Israel also is the best in the world at recycling its fresh water.

Also, a new technology using graphene sheets could make the large scale purification of salt water much cheaper. This is being intensely researched.

OK, now I can discuss the 3rd active measure.

3. Third Active Measure: Greening of deserts and arid regions. What is preventing us from greening the desert regions? Simple. Lack of fresh water and lack of cheap labor. Imagine we have advanced robots that do most of the difficult labor to build a LSWPP and install nearby a large array of efficient solar panels. Now imagine we build this LSWPP on the coast of a desert or arid region that gets lots of sunlight (many desert and arid regions exist on or near an ocean or other large body of salt water). Now imagine the solar powered LSWPP is up and running converting ocean water into millions of liters of fresh water. Now imagine that this fresh water is being efficiently used to turn that desert area into grassland, forest and organic farms. And the vast majority of the hard labor is being done by robots. This green land will now attract rainfall and will capture many, many tons of CO2.

Keep in mind the area that could be converted to lush green land is large, up to many hundreds of square miles. Now imagine this is being done in many other arid and desert areas in the Middle East, India, Australia, Asia, South America and North America. Imagine how much CO2 is getting captured if this is done. And how much healthy food could come from the robotic organic farms, orchards and forests. The more desert and arid areas get greened the more stable the earth’s climate will become and the more the climate will support ecological and human flourishing.

With accurate CCMs, the location of the newly greened areas around the world could be optimized. One obvious question question for a global ecologist is, “If we can green large sections of dry and desert land, which would have the most benefit?” For example, if a large part of some dry area is greened, it could help to improve weather patterns so that other dry areas get more rainfall. Remember that the distribution of dry land and green land help to influence weather and rainfall patterns.

Related to this is the possibility of genetic engineering plants that can thrive with less water. This, of course, will make it easier and more efficient to re-forest and re-green dry land.

Of course, if we had advanced robots there are many other things that can be done like massive amounts of composting, large water retention projects, cleaning up landfills, and much more organic farming, hydroponics and aquaponics. The main reason we don’t do these now is lack of cheap labor. Doing these things on a much larger scale would also be a big help in reducing GW. There are many other ways advanced robots could help slow down and reverse GW, too many to mention here.

I fully recognize that (as of now) there is very little chance that world leaders would agree to implement these actions to mitigate climate change. But the most important concept I want to communicate is that there seems to be ways we humans could safely reverse GW. Once again, the key to this whole process is good computer models of the global climate. As I mentioned at the beginning computer hardware is always getting more powerful. And computer models of the climate are steadily getting better and better which means we can have more and more confidence in their predictions. That means we take only those actions in which we have good scientific confidence will help mitigate GW. As mentioned earlier; we’ll be driving the car of climate engineering with the headlights on high beam. Or, to use another analogy, we’ll be driving the car of climate engineering with an excellent GPS system that shows us how to get to our destination.

Some people may object that these active measures are too expensive. The good news is that there is plenty enough wealth in the world to pay for these measures. The bad news is that most nations have policies and laws that allow rich and powerful people to capture much of the wealth produced by human labor and natural resources. For example, Africa has an amazing abundance of natural resources. But corrupt, greedy government leaders and heartless corporations have taken the majority of that wealth out of Africa and into private offshore banks and other offshore assets.

Most wealthy and powerful people have little interest in helping humanity. They have great interest in using their power and wealth to influence governments so they can become even more powerful and wealthy. It’s estimated that between $10 and $20 trillion dollars are hidden in offshore banks, assets and tax havens! So remember there is plenty enough wealth to fund these measures to solve global warming. This is a very important point that most people don’t realize.

So we can probably solve the problem of global warming and reverse climate change. Great news! But, unfortunately, as we all know, we haven’t yet solved the problem of human greed and apathy. So if those problems don’t get solved the planet may well be doomed for climate disaster. And we’re doomed with possible solutions right in front of us.