Science News Dec 14

[This is a transcript of the video with references . The NIF nuclear fusion milestone didn't make the cut for this week's news. I'll comment on that next week.]
Welcome everyone to this week’s science news. Today we’ll talk about the 50-year anniversary of Apollo’s blue marble, the Square Kilometre Array, talking to robots, the Yellowstone volcano, nanoparticles that help with carbon capture, a forest bubble – on Mars, a new method of spacecraft propulsion, tests for cancer, earthquake tracking from space, and of course, the telephone will ring.  

Scientists in Germany have simulated the weather of 50 years ago with unprecedented precision. On December 7, 1972, astronauts of the last manned Apollo mission, number 17, took a colour photo of the whole Earth. It was dubbed the “Blue Marble” and became widely used by environmentalists in the 1970s as a symbolic image for the preciousness of our home planet.

Almost exactly 50 years later, researchers at the German Climate Computing Centre and the Max Planck Institute for Meteorology have for the first time successfully built a climate model that reaches the same level of detail as you see in the Blue Marble photo. Have a look at those two images. Which one do you think is the photo and which is the simulation? Pause the video if you want to stare, because I’m about to tell you. The one on the right is the simulation. I’ll explain in a moment how you can tell them apart. But first let me tell you how they did it.

They used a computer model that couples the atmosphere with the ocean and the land and put that on a supercomputer at a horizontal grid size of 1 kilometre. Current climate models typically have a resolution of one hundred kilometres. That requires really a lot of averaging and it just isn’t good enough to resolve a lot of relevant details in the clouds and ocean and over land. At the improved resolution, the calculation for two days of simulation took three and a half days. Here you see a visualization for the wind speed.

So how can you tell the real image from the simulation? If you look back at the comparison you can see that the simulation doesn’t quite get the size of structures right. That’s particularly obvious here where you can see that there are more bigger structures in the photo while the simulation has more smaller ones. So, there’s some work left to do.

If we want to know how the climate will change regionally, scientists must run simulations like this for decades into the future. They can’t do it at the moment because they don’t have the computing resources and they don’t have the computing resources because they don’t have the money. Yes, you heard that right. Climate scientists don’t have enough money to make those predictions.

All they need is a billion dollars for the next 10 years. Compared to a bigger particle collider that’s a pretty good deal, innit. So in case you have a billion dollars to spare, let me know in the comments and I’ll

Hi Elon, I hope you’re calling to donate that billion.

Well I need to know what the wine quality will be like here in 2050.

Only the second richest man in the world. For half a day. Oh dear, I’m so sorry.

Hope you get well soon. Bye.

Construction has finally begun on the Square Kilometre Array, SKA for short. After 31 years of planning and controversy, many astronomers are without doubt breathing a sigh of relief. The telescope’s mission is to detect signals dating back to the dark ages of the universe when stars had not yet been formed, and it can also pick up some signals from exoplanets. When completed, the SKA will be the largest radio telescope array in the world with one part in Australia and one in South Africa. The project will cost about 2 billion Euro and it’s expected to be completed in 2028.

The discovery potential of the SKA has caused some controversy. The reason is that what’s now called the square kilometre array is what was previously called the Square Kilometre Array phase one. But there was supposed to be a phase two that, in many astronomers’ minds, was the reason to justify the investment. The original plan for phase 2 would have had ten times as many receivers, but this plan has silently been dropped. What’s now called the square kilometre array doesn’t even cover a square kilometre. New Zealand dropped out of the project for that reason because they concluded that the down-scaled version isn’t worth the money. It’s one of those things that scientists don’t like to publicly mention at all, so, shhh.

The robotics team at Google has created robots that respond to language in real time.  The robot that they used is a moveable arm that pushes objects around on a table. It was trained on a dataset of hundreds of thousands of annotated trajectories  and learned to follow simple instructions such as “Push the blue triangle to the top left corner” or “place the red star above the blue cube”. The team found that after training, the robots could correctly follow instructions in 93 point 5 percent of all cases. It’s a great step towards building robots that you can naturally communicate with. The team has made their dataset of annotated trajectories openly available so that other researchers can use it too.

I think that’s a remarkable achievement.  A 93 percent response rate to language commands in real time is certainly more than I observe in my household.

A new analysis of seismic data from the Yellowstone caldera found there’s more liquid magma underneath it than previously thought. A team of geologists from the United States and Australia combined data from various deployments in the Yellowstone National Park during the past 20 years. They analysed the propagation of seismic waves to reconstruct the size and location of magma chambers, and to infer how liquid the stuff in those chambers is. While this has been done before, their larger data set gives better results.

They found that the largest concentration of molten magma lies three to eight kilometres below the surface and it’s bigger than previous analyses suggested. But please don’t panic. They explicitly stress in the paper that this doesn’t mean the thing is going to blow up soon. They write “Although our results indicate that Yellowstone's magma reservoir contains substantial melt at depths that fueled prior eruptions, our study does not confirm the presence of an eruptible body or imply a future eruption.”

Is it just me or does that sound concerningly vague?

Engineered nanoparticles might help with carbon capture in the ocean, according to a new paper that just appeared in Nature Nanotechnology. Ocean fertilization is a method to stimulate the growth of water plants, usually by a substance that contains iron. When the plants grow, they take up carbon dioxide. If they die and sink to the bottom of the ocean, that could theoretically permanently remove the carbon dioxide from the air. In practice, this method has turned out to be not particularly efficient, partly because the fertilizer isn’t very readily used and some of it remains floating in the water, and also because the dead plants don’t sink to the bottom of the ocean, as they should, so the carbon dioxide is eventually released again. What it does to fish and seabirds is yet another issue.

The new paper points out that coating the fertilizer with suitably engineered nanoparticles might make it easier for plants to absorb the fertilizer and at the same time add ballast to the dead plants, so they remain at the bottom of the ocean.

The authors review 123 studies and go through the pros and cons of different nanoparticles. Some of those materials are promising candidates for the task, but they can also be toxic in high doses, so more research is needed to find out just what they would do. Nevertheless, the conclusion of the authors is largely optimistic. According to the paper ocean fertilization enhanced by engineered nanoparticles “may be a remarkable carbon dioxide removal approach to fight climate change”.

A new paper in the International Journal of Astrobiology outlines plans for a nature reserve on Mars. The paper comes from Paul Smith at the University of Cambridge. He believes that Mars would be the ideal candidate for the first human colony on another planet.

The plus points of Mars are that its gravitational pull is somewhat less but roughly comparable to that of Earth, and while it’s somewhat further away from the sun than we are, Mars still gets enough sunlight for photosynthesis.

Unfortunately, moving to Mars also has its downsides, would you believe it. The average temperature on Mars is below -50 degrees Celsius, there isn’t much atmosphere, and since Mars has lost its magnetic field, the surface is exposed to ionizing radiation. There’s some water on mars, but very little freshwater, and the soil is tough for terrestrial plants to grow on. I talked about all that in an earlier video.

Nevertheless, Smith thinks that a Mars settlement is possible. Concretely he suggests that the best way to start a colony on Mars would be a “forest bubble” with a diameter of about half a kilometre. It would be shielded by a dome made of a glass-plastic combination that lets useful sunlight in and keeps harmful sunlight and radiation out. He stresses that the aim of this bubble would *not be to recreate an earth-like forest, but an entirely new ecosystem. In his paper he surveys a list of organisms that could flourish in this environment and that could then support humans.

So far, we have been remarkably unsuccessful with creating a self-contained biosphere even on Earth. So maybe hold off buying real estate on Mars for a bit.

But speaking of space-tourism, a group of scientists from the US and Canada have proposed a new method of spacecraft propulsion. In their paper, that was just published in Frontiers in Science Technology, they propose to extract energy from solar wind by using a technique similar to dynamic soaring that is used by some birds.

Dynamic soaring is used by some birds to gain energy by repeatedly crossing the boundary between air masses of different velocities. It is most commonly seen in seabirds such as albatrosses, but it has also been observed in eagles and falcons. It works basically because you can extract energy from a velocity field that has a non-vanishing rotation.

The idea of the new paper is now that a spacecraft could do the same with solar wind. Solar wind contains speedy charged particles that can be used for propulsion, and a number of “solar sails” that exploit this wind have previously been proposed. The authors of the new paper now point out that there are several places in the solar system where there are differences in the solar wind that could be used for dynamic soaring , for example where the solar wind hits the interstellar medium, known as “termination shock”. The new method might thus be a simple and energy-efficient way to speed up spacecraft.

Hi Albert,

Glad you’re calling. Listen, I have a theory of special relatives. It explains why time passes slower during the holidays.

Waistline dilation? You’re a genius! I’ll look into that. Good to talk to you!

Scientists have tested a new method for non-invasive, early-detection cancer tests that can be done with blood or urine samples.

Currently existing non-invasive cancer tests look for circulating free DNA which is then sequenced. These DNA tests are most commonly used to detect cancers that affect blood cells such as leukaemia and lymphoma, and cancers of the digestive system, such as colorectal cancer. That’s because those cancers are often associated with specific mutations or changes in the DNA that one can find in a targeted search. However, other types of cancer, such as breast, lung, or ovarian cancer, may not be detectable with these DNA test, either because they don’t shed large enough amounts of DNA, or because they’re not associated with specific mutations. And some types of cancer, such as brain cancer, don’t leave traces in the blood.

The new method instead looks for traces that are left by presence of cancer cells in certain carbohydrates that may or may not be pronounced glycosaminoglycans. When cancerous tumours are present in the body those carbohydrates may be altered or produced in excess, and these changes can affect the growth and spread of cancer cells.

According to the new paper which was just published in the Proceedings of the National Academy of Sciences, the authors used machine learning to identify what kind of markers to look for. They then analysed samples from somewhat more than a thousand people and found that their method can double the chances of early cancer detection for 14 different cancer types. Though more studies are needed to confirm the tests are reliable, this is a promising development.

Scientists have found a way to improve earthquake tracking from space. In principle, it’s possible to use the Global Navigation Satellite System that the GPS is part of to measure earthquakes. That’s because those systems track the precise position of receivers on the ground that get displaced during earthquakes. In practice, however, this method turned out to be unreliable. That’s mostly because the signals travel about 20,000 kilometres from the surface of earth to the satellites and on the way, they get distorted by the atmosphere.

In a new paper that was just published, geophysicists from the University of Colorado used machine learning on past data to tease apart signal from noise. They used more than 20 years of data which contains signals of 77 known seismic events. The software was trained on 90% of that data and the other 10% was used to check how well it worked. They were able to correctly identify 90% of earthquakes above a magnitude of 4.8 and 98% of those above magnitude 6. This method is still not remotely as good as ground-based seismometers, but it can be useful when those aren’t available or not working.