Every term I try to go on the Ge136 field trip. This class, a student seminar based cornucopia of learning, has taken me on trips throughout the American south west. This time was no exception.
On many of the classes we've driven straight up the Owens valley towards Lake Tahoe or similar. This time, we drove a lot less (only ~800 miles) and enjoyed some of the places slightly closer to home.
The trip began at 5:30am on Saturday. Surprisingly, everyone was there on time. We loaded the trucks and headed out via the 210, 5, and 14 as the sun roared over the horizon. Before long we were through Mojave and up onto the Garlock fault, where I gave my talk on the Rand Schist. We proceeded throughout the day, stopping for lunch and talks at the Alabama Hills and Lone Pine Fault, eventually reaching our campsite at Convict Lake around sunset. The glacial valley terminated in a coloured spire of rock displaying a particularly nice metamorphic roof pendant.
It was incredibly windy. Windy enough that talking was difficult. We lashed our tents to relatively immovable objects and huddled around a beaten fire. Dinner, of chili with a secret ingredient, was filling and excellent. A and I walked down to the lake to attempt photography, but constant spray and tripods being unable to stay up in the wind led us to give up. We returned to the campsite and organised a mercy dash to a nearby hot spring. Two cars departed, but only ours managed to find a sufficiently empty hot spring in time. We relaxed under the open sky and discussed terraforming and other sciency stuff. A few meteors flew overhead.
At length, we returned to the campsite, I warmed my sleeping bag near the fire, then retired to the camp stretcher and bundled up against the cold. Taking a few long exposures required enough crunches to ensure I was thoroughly warm by the time I went to sleep. Having slept only 3 hours the night before and driven all day, I passed out extremely quickly.
The next morning, I woke, inverted my previous actions resulting in a dressed, upright Casey, packed my stuff, ate some breakfast, and we were on the road. We talked about dating of glacial moraines and eventually drove as far as Bodie, a ghost town north of Mono Lake. Booming and busting in 5 years, Bodie's remaining buildings (much of the town burned down historically) are remarkably well preserved. By this time the wind had died enough to allow an attempt at drone flying. We checked out the museum and some of the houses, while trying to imagine how it had been in 1880, bullets flying everywhere, gold dust in the air, constant noise and the thrum of possibility.
Later that day, we drove back to Mono Lake, shivered during the sunset while hearing talks on hydrology, history, volcanoes, microbiology, and so on. Soon after we departed through half-dead forests for our campsite nearby, set up and cooked spaghetti for dinner. The first sign of trouble came when cleaning off the tables we encountered patches of nice. That night, the temperature dropped to -8C/20F, which was so cold I put on some socks with my sandals. Waking up frequently to shiver afforded terrific opportunities to admire the stars, galaxies, meteors, and quality campfire singing. Eventually the sun rose, two dozen or so frozen icicles stumbled from the tents and trembled their way to the coffee pot and lightly burnt toast.
We packed and returned to Mono Lake, then some of the surrounding volcanic craters, and gradually made our way south via the Bishop tuff, some tasty road cuts, canyons, and another glacial valley with awesome marble metamorphic rocks and tungsten mines, and some hot springs. Soon it was time to head back and we zoomed down the 395 to get home by 9pm.
This article originally appeared in the California Tech on Tuesday 21 October 2014.
Hong Kong is one of the most vibrant, prosperous cities in the world. Rising to power and prominence under British colonial rule, it was formally handed over to the People's Republic of China in 1997. Despite being part of China, Hong Kong has always enjoyed a privileged administrative status under the One Country, Two Systems model with a substantially different economic structure and a vibrant political life. Hong Kong is ruled by a Chief Executive (something like a governor) elected by an Election Committee of (today) 1200 entities, representing individuals, districts, large business interests and other organizations, and serve up to two consecutive 5 year terms. Not unlike the governing structure of a modern multinational corporation, economic power carries real political power.
In 2007, the then Chief Executive Donald Tsang published the Green Paper on Constitutional Development which was subsequently ratified by the National People's Congress Standing Committee, the relevant governmental body in mainland China. Since the end of British rule there have been only 3 Chief Executives. They preside in a responsive way over the rapidly developing economy and political climate, leading to a much more dynamic constitutional framework than that to which we are accustomed in the USA. This Green Paper was seen as a compromise between mainland China and the Pan-Democratic movement which enjoys the support of about 60% of the population, as it ruled out the possibility of a general election by universal suffrage in 2012, but allowed for the possibility of election of the Chief Executive and Legislative Council (something like the US Senate) by universal suffrage in 2017, then a decade away.
Fast-forward to 2014, and it is now time for the National People's Congress (the Chinese governing assembly) to decide how to implement the recommendations of the Green Paper. To the original wording they added the stipulation that the Election Committee must preselect two or three candidates who 'love the country and love Hong Kong', and that any election by universal suffrage must contain institutional safeguards for this purpose. While all historically elected Chief Executives have been Nonpartisan (pro Beijing), a general election may upset this balance and lead to the legitimization of separatist policies. Beijing worries that this would set a dangerous precedent. The proposed vetting of candidates by >50% approval from the (mostly pro Beijing) Election Committee largely obviates this concern, as does the process by which Beijing must officially appoint the Chief Executive for the duration of their term.
It is thus important to realise that the resulting protests are not pro-democracy in the sense that Beijing is removing pre-existing rights, but that they are punting promised progress on the issue. Hong Kong has never had anything like representative democracy. It is also far from clear that a populist majority government would naturally serve the best interests of the people, who enjoy prosperity borne by the freewheeling economic nature of the city, a prosperity certain to be harmed in any chilling of relations with the mainland.
The resulting protest is not a singular group, but actually involves representatives of four distinct groups. The first participants in the protest were student led groups who announced and executed a weeklong class strike. Comprised of Scholarism, representing secondary school students, and Hong Kong Federation of Students, representing tertiary students, they began their protest by ditching school for a week to really get their point across. Of course, no longer in school, these students sought to occupy public spaces and subsequently had antagonistic encounters with police. While tame by the standards of the Arab Spring or even Occupy Wall Street, the use of pepper spray and tear gas against secondary school students galvanized the movement, leading to the accelerated involvement of the third group, Occupy Central with Love and Peace, a non violent local branch of the Occupy movement. The fourth movement, the pan-democracy camp, is an alliance of the many pro-democracy political factions in Hong Kong. They must be the most patient advocates of democracy in existence.
The student-led groups have stated their goals are
Universal suffrage.
Resignation of the incumbent Chief Executive CY Leung, partly in response to perceived disingenuous statements and actions as the protests proceeded.
The withdrawal of the decision of the National People's Congress.
The submission of a new electoral reform plan that includes civil nomination of candidates for the Chief Executive.
This is not your garden variety protest. International coverage focused on the largely peaceful nature of the protests. To date there has not been a single death, and a relatively tiny fraction of injuries. The protests, which have consisted of disruptive occupation of central arterial streets, have been exceptionally well organized, with guest lectures, homework, decentralized distribution of food, water, clothing, shelter, and the development of mobile medical facilities. Communication has also stepped up - lessons learned during Egypt's short-lived revolution now translate to dedicated apps for encrypted peer-to-peer mesh networks, although many participants' phones were also compromised by phishing attacks distributed by SMS. The protests became known as the Umbrella Movement due to the innovative use of umbrellas to deflect tear gas cannisters. Cannisters, protesters noted, often manufactured in the US.
As usual, mainstream media coverage has downplayed the size of the protests, especially in mainland China, but the protests regularly drew around 100,000 participants, easily occupying several blocks of the city with multi-leveled roads recognizable in the astonishingly prescient film Pacific Rim. Confrontations between police typically escalated to the firing of tear gas and arrests of dozens (but not hundreds) when protesters or police managed (often inadvertently) to surround each other. Arrested protesters have consistently been released soon after detention. As the protest has developed, protesters and police have generally interacted with a great deal of mutual respect, as the politically unaligned police clearly work long hours to keep the peace and have largely eschewed the earlier tactical error of appearing in riot gear and wielding batons. Perhaps the greatest strategic error and outbreak of violence occurred when police unfamiliar with social media were brought in from the mainland and beat up a bunch of school children, all recorded and instantly shared on the internet. This blunder is probably the single greatest contributor to the current detente, especially in contrast with the overworked Hong Kong police, who have also protected the pro-democracy protesters from violent attacks by pro-business or Triads-linked counter protesters. As of this week, many protesters have returned to school, but still return to the streets in substantial numbers in rapid response to progress or lack thereof during ongoing negotiations.
What is the outcome? In many respects the protests have been atypical. Protesting about a proposed political process three years in advance is unprecedented. The key stakeholders, consisting of entrenched business interests aligned with Beijing, as well as a Beijing extremely wary of local movements for separatist or autonomous political innovation, are unlikely to budge. On the other hand, a new generation of youths educated with relatively unfiltered access to the internet and personal communication tools of extreme efficiency have a strong interest in gaining unfettered access to the political process both within Hong Kong and China more generally. Although the state has a jealously guarded monopoly on violence, it is likely that any suppressive response is likely to spawn matching unrest on the mainland, with the possibility of future Tiananmen Square-like trouble. Although the main sources of endemic unrest on the mainland - perceived economic inequality - are less relevant in Hong Kong, the possibility of a forged common cause is likely to lead to some tense meetings in the halls of power. At the least, token concessions carefully measured to avoid encouraging disruptive protest action are none-the-less likely to materialise in order to appease ongoing frustration in the youngest generation. Perhaps the least terrible hypothetical outcome is the eventual formation of a second, lower house of government, a chamber of the people, elected by universal suffrage, to complement a senate-like body of the Election Committee and Legislative Council, where the Chief Executive becomes in essence a powerful Secretary of the Treasury. Meanwhile a Secretary of State elected by the lower house leads the government and bargains hard for Hong Kong's ongoing semi-autonomy and ensures that Hong Kong's continued prosperity is adequately shared. If moves in this general direction are not made officially, it is only a matter of time until app-based consensus gains enough legitimacy to claim a seat at the table of real power, with or without the approval of the existing governance structure.
NB: This article originally appeared in the California Tech, the Caltech student newspaper, on Monday October 13 2014.
Starting in 2009, NASA began a process of selecting private companies to provide astronaut transport capabilities to low Earth orbit. Coinciding with the retirement of the space shuttle, the program is intended to maintain US crewed space capability, reduce dependence on the Russian space agency, cut costs, promote innovation, and allow NASA to focus on crewed deep space exploration, should they ever receive a mandate to do so.
Despite a substantial under-provision of funds from congress, the program has proceeded, with yearly reviews, designed in collaboration with the key players, to assess ongoing performance via 'Milestone goals' and keep the program focused. The biggest prize by far in this program is the development of the spacecraft, and by the beginning of 2014, three main contenders were still in play.
Boeing, the giant of US aerospace, has a long history of building rockets and space systems, including the Atlas V launch vehicle with United Launch Alliance. With tens of thousands of spare engineers and a deep understanding of the bid process, they proposed the CST-100, a very conservative capsule design based on proven technology.
SpaceX, the startup space company based in Hawthorne, has managed to bend the CCtCap process around their own development of the Dragon V2 capsule, a design with significantly more capability than called for during the process. With incredible design and manufacturing innovation, SpaceX is making waves in every part of the launch business, including their high-profile argument with ULA over sharing the military satellite launch business.
Sierra Nevada Corporation, a smaller and private aerospace electronics and systems company had branched out into the launch business with the development of the Dream Chaser lifting body spaceplane, as well as the new Virgin Galactic rocket motor. The Dream Chaser is designed to launch atop the Atlas V, then re-enter and land on a regular runway, whereas SpaceX and Boeing's capsules are designed to land with parachutes and/or rockets on land. All are designed to carry up to 7 people and to be reusable.
Company
Boeing
SpaceX
Sierra Nevada Corp.
Spacecraft
CST-100
Dragon V2
Dream Chaser
Bid
$4.2b
$2.6b
$3.3b
Gross mass to LEO
20,000 lbs
16,600 lbs
25,000 lbs
Milestone goals
17/20
13/17
8/13
The latest decision on capsule development was expected by September 2014. NASA was widely expected to pick two of the three designs to proceed, based on progress and assessments of future risk.
At this point it was anybody's race. Boeing had asked the most to develop the least capable capsule, but had experience and resources. SpaceX had asked the least to develop a futuristic capsule design, to be launched on their own Falcon 9 rocket. Sierra Nevada had also proposed a relatively cheap development for the Dream Chaser, but would have to launch it with ULA (and hence Boeing's) Atlas V rocket. Adding to the complexity, the Atlas V is not presently rated for human flight. Rating the Atlas V is a complex program that Boeing understandably would be charging contractor rates for if its CST-100 capsule was not selected. If that wasn't enough, the Atlas V uses the Russian built (and thoroughly badass) RD-180 rocket on its first stage. ULA has admitted that they have a limited supply of engines and uncertain capability to build copies themselves. Recently they contracted Blue Origin, Jeff Bezos's private rocket company, to build a replacement engine (the BE-4) powered by methane and oxygen.
On September 16, NASA announced that SpaceX and Boeing would proceed, while Sierra Nevada would not receive further funding. In response, Sierra Nevada pledged to continue development with internal funds, cooked up a version of the Dream Chaser which could launch from the Stratolaunch plane, and filed a formal protest, citing similar progress to Boeing but a lower cost. The result of the protest will be determined by January, but the result was anticipated fairly widely.
NASA is leery about space plane designs, after the experience with the shuttle. The promise of a cheaper, reusable rocket was belied by the reality. The shuttle, for all its benefits, was also a technological nightmare, extremely expensive, not particularly safe, and spent a very large proportion of its overall power lifting wings, wheels, brakes, and empty space into orbit. When a really good rocket can get 3% of its launch mass into orbit, the last thing you want is a spacecraft which comes with substantial weight penalties for features of limited utility. While the shuttle could land on a runway, there were only a few runways long enough in the whole world, requiring precise targeting and good weather. In contrast, the Russian workhorse Soyuz capsule can land anywhere, including once famously on a frozen lake.
On Saturday, Aviation Week obtained and published a copy of NASA's report into the decision. This report states in part "[We] consider SNC's design to be the lowest level of maturity, with significantly more technical work and critical design decisions to accomplish. The proposal did not thoroughly address these design challenges and trades. [The proposal] has more schedule uncertainty. For example, some of the testing planned after the crewed flight could be required before the crewed flight, and the impact of this movement will greatly stress the schedule."
The Dream Chaser looks great and promises a lot, but it is clear from this quotation that large aspects of the design are still to be hashed out. Given that uncertainty, it is not surprising that NASA chose to exclude them from the final stage of the program.
Every year the Caltech Y, a Caltech associated community involvement organization, runs a week-long hike in the Sierras for incoming students just before orientation and rotation. Normally at this time of the year, I am involved in International Student Orientation, but this year I decided to lead the hike instead!
About 25 people total were involved, split into three groups according to degree of ambition to achieve altitude sickness. I was assigned the 'beginner' level group, which was pleasantly cruisy, walking 3-5 miles (5-8km) per day. Normally arriving at camp by midday, we had plenty of time to explore, climb, look for bears, filter water, and shiver.
For a change I used a GoPro instead of my regular camera. It is much more wide-angle and more versatile, though nearly useless for night-time photography.
After a week all three groups met at Lodgepole campground and compared notes and blisters. Noone died! Success all around. I had only previously hiked on the eastern escarpment, so it was nice to explore the interior of the range.
What started in the mid 80s as a small group of friends gathering on a beach to burn stuff and party has now grown, in its ~28th year, to one of the most extraordinary and bizarre events in history. During the last week of August, approximately 80,000 people leave their drab, wretched lives and travel to the Black Rock Desert in Nevada to build a city, do art, music, dance, and have a terrific time.
Having lived in the American south west for nearly 4 years, it was high time for me to go. Planning had been underway for more than a year, but at the last moment my designated transport, a tiny Cessna 152 plane, became unavailable. Plans awry, improvisation began.
S, O, and I met and crammed our stuff in S's Ford Taurus one exquisite Sunday afternoon in the recent past. By 6pm we had set out and laden with all the food, water, costumes, sparkles, and fairies we could possibly need, we sped up the 5, through Sacramento, Reno, and out into the deserts of Nevada. By 4am we had joined the queue of vehicles waiting to get in through the gate into Black Rock City. The place I had heard so much about yet never seen, never been, never lived, never experienced.
My curiousity turned to frustration and ambivalence as shortly after our arrival a series of unusual thunderstorms dumped 6 inches of water onto the playa, or dried up lake bed, on which we were parked. Instantly the surface turned to adhesive devil's cake and all vehicles were halted until the surface dried. When it became apparent that watching playa dry was like watching paint dry, and about as fast, the 15,000 or so of us stuck in line did what we do best - brought the party with us. By noon the surface was dry enough to walk on, and around us (the line was about 15 lanes wide and a few miles long) people cracked open their RVs, assembled bikes, changed into playa clothes (often not much of anything), fired up their barbecues, and even held spontaneous improvisational concerts on their roof of their RV, or got their DJ equipment out. A series of dark clouds spun overhead but the heavens held, and by 7pm we were on the move again.
At the wheel yet limited to 5mph, we met our greeter who provided a kit of information and most valuable advice: "Pee clear!". We dropped O at her camp; Basshenge, occasionally of the silent 'B', sporting a 100,000W sound system, then headed in to 5 o'clock and Esplanade. S's camp and my artistic nexus was Phage, a camp of mostly scientists and their sympathizers positioned in prime location on the front row. Black rock city is arranged as a series of concentric circles from A through L, and radial streets named after times on the clock from 2 through 10, with a gap between 10 and 2 for art and dancing.
Coming upon Phage, S unpacked her gear and I headed for the Tesla Coil. A project I'd been involved with since the beginning, my main (though minor overall) contributions were a less lethal capacitor bank and the external design of some of the structure. I was thrilled to see the structure standing (as opposed to in tiny pieces when last I saw it) and the coil in the middle. A ladder underneath supported the main builders, D and M. Oh no, the coil is here but doesn't work properly! My fears were unfounded. We dropped the ladder and fired up the midi controller. Soon gigantic bolts of lightning were sending 120dB shivers down my arms and spine. With some trepidation I approached the janky midi keyboard and proceeded to play what little piano music I could remember. The sound, light, and smell was overwhelming. Hordes of brightly lit burners milled around like moths to a candle flame - the sound was audible from the eponymous Man, a 105 foot wooden structure in the center of the city, not yet burning.
S and I drove out to the airport, where I was camping, and set up my camp. Despite the lack of a convenient wing I was able to raise my shade and hammock, driving rather disposable stakes into the living playa with a series of mallets of steadily increasing size and lethality.
I crawled into my trusty hammock and passed out.
The following morning I volunteered to be an oar-less galley slave and helped produce enough hash browns, veggies, sausage, and so on for a hundred people who were camping at the airport. When I had eaten enough to last until dinner I hiked into the city where, for the first time, I took a look under the light of day. Phage was located on a street corner with containers, several trucks, parked cars, shade tubes containing tents, a galley, dining area, steam yurt, generator with trenched 3-phase power, fresh water hose, grey water disposal, art installations, flags, a dome, and two art cars, Dr Malthus and Dr Brainlove.
Art cars, or mutant vehicles, are the only powered vehicles allowed in the city. Limited to a maximum speed of 5mph, they must disguise their vehicular origins and look like something else. The photo album has a few photos of various art cars, but in total there are over a thousand, ranging from tiny single person vehicles up to large articulated multi-story party complexes sporting millions of computer controlled LEDs and flame throwers. One art car, called Robot Heart, features a sound system with a mere 85,000 watts of undistorted subwoofer and traditionally the best DJs in the world, though the lineup is never announced. This is a video someone else made about Robot Heart.
Dr Malthus belongs to one of the sister camps of Phage, and is styled as a military half track with outdoor and indoor sitting areas as well as a roof lounge/DJ table. The indoor space features squashy leather arm chairs, a fireplace, and a mounted moose head. The driving area is surrounded by a cage to comply with liquor open-carry laws, and believe it or not, the vehicle is road legal.
Dr Brainlove is Phage's new art car. Started only 6 months ago, she is a converted school bus. A large geodesic brain is mounted on top of the bus, modeled on an MRI scan of one of the designers. Each geodesic node features a lighting node with dozens of computer controlled LEDs that, combined, fire like neurons, and are controlled by an EEG cap worn by a willing volunteer. As you may have gathered, Phagelings are somewhat geeky.
Brainlove was the very opposite of road legal, being GIGANTIC in size. It had to be assembled on playa, and at this point the lighting system was being put together. Composed of 3D printed parts, they needed to be aligned, the wiring checked, then ziptied into submission and position. Only 178 nodes were required for completion.
Slightly more complete than before, Brainlove was loaded with a kaleidoscope of people and taken for a drive out on the playa to rehearse safety procedures during the day. What do you do if, while driving, a horde of hippies jumps on the structure and begins climbing? I sat on the pre-frontal cortex and kept an eye out for people trying to get run over.
Back at camp preparations were underway for the Tuesday night party, requiring the running of optical fiber to run the coil under the esplanade. A trenching machine cut a hole, we threaded 100 feet of conduit, and laid fiber. Only took a few hours! After a quick break for dinner, I returned to phage in time for the party, then walked out to 10 o'clock. At night, the playa is completely dark save for art, cars, and people festooned with LEDs, EL wire, flame throwers, and lasers. One camp had a trio of very powerful lasers shining over the city onto a nearby mountain to display the time. At the far end of the city I was seriously tired and returned to phage and the airport. On the way back I fell asleep on my feet several times, waking up to find I was still walking but in the wrong direction, or with flaming art cars zooming by.
The following day I spent mostly walking around taking it easy, meeting new people, checking out the big art installations on the playa (most of them at least 3 stories high), and resting. That morning the Black Rock City ultramarathon is run, and a few of them were passing the camp. Here's a video from 2013:
I had brought a pair of welding goggles dark enough to look at the sun, and found that after about 10 minutes, your eyes adjust and you can see other things through them. Walking around where people are silhouettes against the sky and everything is a funny tinge of green did not seem out of place at all. In the afternoon I made my way back to the airport and gave the ACME Bomb Company cocktail hour talk on Cosmology, an hour or so long speech describing this history of the universe in reverse chronological order. A large and appreciative audience asked useful questions and were highly amused by my descriptions of metallic element nuclear synthesis, oxygen isotope variance, interplanetary meteorite exchange, the cosmic microwave, neutrino, and gravitational wave backgrounds, and so on. In the background, a couple of powered paragliders were looping and rolling close enough to the ground that they regularly touched it with a wing tip. After grabbing some dinner, I hitched a ride on a perilously dangerous ice bike back to Phage and gave a second talk, this time on black hole collisions. The kicker there is that at the moment of greatest gravitational wave emission the signal actually vanishes because the common black hole event horizon grows and swallows all its own screams.
We got the Tesla coil working and managed to attract some talented pianists, whose memorable performance still rings in my ears. I walked around the inner playa with S and discussed business, space launch, and a few other interesting ideas before stumbling back to the airport and passing out.
The following day I strategised and got a ride on the skydiving plane. Burning Sky Camp runs a skydiving plane all week and allows divers and riders to get a free trip up and around Black Rock City. That morning, a record had been set for the largest simultaneous nude jump (15), so all were in high spirits. At the zenith, 10,000 feet off the deck, the last of the divers leapt out and the pilot cut the power and put the plane into a steep, zero G dive. We were back on the ground in just over a minute.
In the heat I walked slowly around to 7 o'clock and K where I met some Caltech types, chilled in the shade, and had a chat. By this point I had realised that Burning Man is not really an alternate reality. The culture is not so different to the culture in any university or musical society, or at least parts of it. I had experienced substantially more culture shock in many foreign countries. Burning Man is overwhelmingly affluent white people from San Francisco. Cultural dislocation is not really the point.
Accidents do occur. Legend has it that very few people die at Black Rock City because most people live long enough to get on the medevac plane, after which they are declared dead in Reno. The previous night, someone fell from and was crushed by an art car, the first such death since 2003. From conversations with various rangers, there were many more near-misses, in which injured people came back to life in the nick of time.
Back at Phage, the talk was of a new art installation. A post-hole digging auger had become stuck out in the deep playa, so a decision was made to convert it to an artwork entitled 'The Lobster Trap'. A set of handcuffs was ziptied to it, a large sign saying 'do not touch', and a set of unrelated keys. A day later, someone rode out to check there weren't any skeletons attached and found the ziptie broken and the handcuffs gone - incredibly, someone had locked themselves to it! Later, the identity of the trapped individual was discovered - we had caught a fellow Phageling - who had had to get the handcuffs taken off by the police.
M and I spent a pleasant hour climbing all over the Phage dome attempting to install a new lighting system, then back at the airport a cocktail hour talk on deep brain electrical stimulation, followed by an excelsior dinner of chicken, salad, and curry. That evening I helped do the dishes, but made it back to phage in time for the tail end of a talk on the IBMI, a group dedicated to the improvement of Burning Man by turning it into a two day networking event for 300 people. The coil was once again on fire until an unidentified fault (strategic power cable disconnection by grumpy person) gave me the evening off and I went to explore center camp, which had a sweet Jazz bar and an even better Tango dome with funky LED triangles and nice music.
The following day, my fourth on the playa, was windy and dusty. Around noon, K and I went on an extensive bike adventure. At this point I realised not bringing a bike to BM was a serious oversight. On the playa, visibility was down to 10 feet, so we rode aimlessly, discovering art, lost people, and burned stuff at random. We made our way to 9 o'clock, picked up some mail, and spent a while finding its recipients and delivering it. On the way, of course, meeting lots of interesting people. At 3 o'clock, we visited another post office, called the BRC3PO (Black Rock City 3 o'clock Post Office), with a Star Wars Theme, obtained 3 more letters, and delivered them after a detour to the deep playa wherein we found yet more cool widely separated art and nice people.
Back at Phage around 6pm, it was time for 'Ask a Drunk Scientist', in which hordes of unfulfilled PhDs swarmed the Esplanade armed with megaphones, labcoats, signs, and attitude, haranguing passers by to ask us anything. "Is a PhD worth it?" "How much debt do we have?" "Can science help you find love?" "Why are sunsets red?" and so on. A great opportunity to apply knowledge and patiently explain stuff in an understandable way!
That evening, M gave a talk on the Tesla coil, answering the three most commonly asked questions: "How does it make sparks?" "How does it make music?" "Why doesn't it kill hippies?". After that we were infested with people who couldn't play very well, which is a problem when everyone within half a mile has to shout over the noise of yet another terrible rendition of Fuer Elise...
By this point the Institute LED sculpture was up and running - a pair of LED encrusted spheres with a bewildering display of patterns, together with some pretty good dance music. I alternated grooving with cerebral chats on the Moveable Feast, gave a few massages, and then passed out.
It is Saturday! The day of The Burn! All week we have been working for The Man, and soon he must be burned. With Burning Man closing only 3 days later, it is time to strike! First the Phage Dome was unbolted, parted, and packed. Then a gigantic crane lifted the coil, we detached the legs, then the rest of the structure, power electronics, and so on. Rugs were rolled, bikes corralled - I managed to salvage an abandoned Huffy Cranbrook (new and already falling apart) that became my ride for the next few days. Shade tubes were collapsed. Piles of stuff 10 feet high formed around various trucks, trailers, and containers. Hexayurts were disassembled.
Come 5pm K and I rode to the airport for the cocktail party beneath the Star Port, enjoyed a terrific band, climbed the observation tower, watched the planes, made an incredible salad, and then ate absurd quantities of paella, pork, beans, bacon, and salad. Several of my airport friends worried that I still had not been awarded a playa name. Some people said it can take a decade to get one. One pilot told a story about a ride he had gifted to a couple in which the man proposed, and then the woman revealed she was pregnant! Much excitement!
We rode back to Phage, parked, then walked out into the playa. The Man was surrounded by a Colosseum of art cars all pumping music into a crowd of 70,000 people mostly seated between them and an inner fire containment circle. A few thousand fire twirlers performed, followed by fireworks and gigantic fireballs as the Man was lit. The facade burned away quickly, the inner structure of very large lumps of timber took longer. Tornadoes of smoke, flame, and embers rose downwind of the sculpture and at times we were showered with glowing cinders. After about 90 minutes the ankles burned through and the structure leaned, collapsed, and fell. Back at phage D had finished packing the Moveable Feast so we said goodbye. At the airport, I was still buzzed so I rode around the perimeter fence, a distance of about 11km, seeing all sorts of odd things in the dark. Already, a line of vehicles had started to leave the event. Back at camp, I had my daily wet-wipe bath, was re-covered in dust from the shade structure, and prepared for bed.
Before coming to Burning Man, I had been warned that my habit of wearing sandals everywhere would be my undoing, due to a mysterious and scary condition known as playa foot. Allegedly, playa dust is super alkaline and, like laundry powder, will burn your exposed skin, leading to dryness, cracks, infections, and probable amputation. Now, I was relatively certain that small amounts of dust in foot wrinkles would have their pH neutralised by sweat and then form a protective layer, but ever the servant of science I decided to clean my left foot thoroughly (thus destroying any protective layer while also removing the offending material) and leaving my right foot as the control. I also brought a supply of vinegar in case of emergency pH adjustment. After 5 days of steeping in playa, neither foot showed more than a mild sunburn, which annoyed my inner scientist no end. My left foot had a couple of scrapes from being trodden on, but that was it.
Sunday. The man, which had previously towered over the whole event, was gone. Packing up continued. I rode out to the deep playa and found the art I'd previously missed, including the zollotrope and the last outpost. The zollotrope is a gigantic suspended wheel with sculptures attached. At night, it spins in time with a strobe and the objects appear to be moving. The last outpost was a very interesting, dark artwork. An interactive building set in the near future during a time of societal breakdown due to aliens or zombies or whatever, it conveyed a sense of paranoia and black humour. Detailed journals and various signs were quite immersive, and made a nice contrast to the usual love/harmony-centered artistic zeitgeist.
Back at Phage, Brainlove was being dismantled. I clambered all over the structure for hours pulling zipties, nodes, wires, optical fibers, thread protectors, and throwing them all into labelled bags. Finally, the bare structure was ready for disassembly. Two teams of two climbed to the peak and began impact drilling the nodes apart. To ensure safety we had safety gloves, safety helmets (optional) and safety whiskey (compulsory). Obviously you had to check what you were standing on, since large parts of the structure became fluid once certain bolts were removed. After only a few hours most of the structure was disassembled, and by the following day, packed. In one casualty, a freed strut swung around and smacked my foot - a small amount of blood was readily clotted by some playa dust I found in my pocket and work proceeded. Needless to say I strongly advocated major design changes for any future brain.
Drank a lot of water, then grabbed a rake with L and proceeded to moop. Moop, or Matter Out Of Place, is bad news on the playa. The camp has to be thoroughly mooped, or de-mooped, before going home. Any foreign object must be picked up and packed out. After an hour or so, L and I had demooped the area formerly under the Tesla coil and the Phage dome, collecting more than a liter of hair, feathers, sequins, metal shards, bolts, sawdust, wood chips, contaminated dirt, thread, fiber, string, paper, plastic, and so on.
Back at the airport I disassembled my own shade. Getting said disposable stakes out of the ground required vice grips and long bits of metal for levers. Dinner was pasta, salad, meatballs, and mushrooms, followed by a ride back to the playa for the temple burn.
The temple is a structure designed and built each year along the same axis as the Man, roughly at 12 o'clock. Within it people write or leave tributes to deceased friends or relatives. When I visited earlier during the week, there were numerous tributes to Robin Williams, as well as thousands of others of the dearly departed. The day after the Man is burned, the temple is burned. The structure is incredibly beautiful and ornate, but ultimately serves to celebrate the finite nature of life. In contrast to the previous night, tonight the circle of art cars was silent, the people sat silently and watched. The structure was lit on the down-wind side and the flames quickly spread. With its high surface area, the temple was soon a skeleton which fell with a half-twist. People approached the flames and watched as its last parts turned to hot gas and rose into the sky. Here's a video of the burn, the collapse happens just after 9 minutes:
Back at Phage everything was packed, so P and I sat on 3-phase power adapters and discussed Burning Man romance in terms of semi-conductor theory before returning to the airport and sleeping.
Monday I struck my own camp, then returned to Phage for the last time and mooped for hours, finding all sorts of good stuff. We did our best to return the surface to its original state, including crushing and filling trenches, tire tracks, and so on. K was my ride home, and after finding R, we loaded my stuff, said goodbye at the airport, and left at about 1pm. Traffic was slow out of BRC as far as Gerlach, then picked up. We fueled at Carson City, and drove south on the 395. The sun set as we pulled into the Mobil at Lee Vining, where I conspired to eat a reasonable yet light dinner and marveled at the bathroom with running water. K, R, and I were in no hurry to get home, so we took a detour to Mono Lake and explored the tufa towers by moonlight, then continued on to Mammoth.
After a few false starts, we found a Mammoth hot spring around midnight and soaked under the stars and meteors for nearly 2 hours. At last, a place to get clean! The following morning I found the one place I had neglected to scrub - the back of my hands - still carried a patina of playa dust. I climbed into my carefully packed clean clothes and K drove us south through the Owen's valley until just before dawn, we explored the Fossil Falls by flashlight. I drove some more, passing through Red Rock Canyon at dawn, with Venus and Jupiter clearly visible in the east. R came to life near Mojave, so we ducked into a Denny's, I ate a decent meal, and then R drove us the rest of the way to Pasadena, arriving around 10am.
I had heard of, read about, and imagined Burning Man for years. It surpassed all my positive expectations and failed to meet my negative ones. The climate, culture, and conditions were less extreme and much more friendly than I had anticipated. The diversity of artistic expression exceeded my expectations by about a factor of 20. That is, if you try and imagine all the things you think are possible at Burning Man, you will be lucky to get to 5% of what is there. I can and will recommend at least one visit to anyone who thinks they might like to go. I don't think it's the universal panacea or the only place I can ever be myself, but I am privileged to live an existence which ordinarily suits me quite well!
Here are two slightly NSFW videos which capture at least a fraction of what Burning Man is about.
Some of my loyal readers may not know that I just got back from a 10 day trip to Spain! Summer in LA is a time of heat and getting research done, but I was toying with a trip to Europe some time during the next year to catch up with some friends when two events occurred to bring the schedule forward. First, my old housemate/chem lab partner B finished his other PhD, and second another old friend T narrowly escaped death and dismemberment, and was recovering on the sunny northern coast of Spain. Meanwhile B decided the best way to celebrate finishing a few years of pain and drudgery in grad school at Oxford would be to spend a month walking across Spain on the Camino de Santiago, or French Way, an ancient pilgrimage route leading to the alleged remains of St James.
Cut to last Friday, 2am. I can barely sleep, which is lucky, because an airport van picks me up and whisks me through the LA streets to LAX, where I wander about desperately trying to remember how to walk long distances. The previous week's training sessions were cancelled when I came down with a nasty cold, but dedicated sleeping and doing nothing seems to have taken care of the worst.
The first leg, to JFK, enjoyed a 100kt tailwind courtesy of the jet stream. The second leg, from JFK to Brussels, involved talking programming with a retired COBOL programmer/grandmother and watching Wrath of the Titans - excellent aeroplane TV. In Brussels, I had a 6 hour layover, so caught the excessively swift and smooth train to downtown, and walked around looking for bread, water, and an ATM. At 8am on Saturday, Brussels isn't exactly the most alive place, but I eventually found my way to another train station and returned to the airport in time to negotiate 18 flights of escalators and 400 chocolate shops for my final flight to Bilbao, in Spain.
This flight featured a nearby child with the most piercing scream I have ever heard, and I've heard a lot of children scream, for unrelated reasons! I slept soundly. Bilbao delivered me to a sequence of buses and before long I was in Pamplona, meeting B and finding the Xarma Hostel, a mere 31 hours after leaving my home in Pasadena, and only 47 hours since I'd last been horizontal. So I was wide awake. B and I went shopping and managed to find enough ingredients for a vegetarian, gluten free dinner, which I consumed with all the enthusiasm of someone who had lived on airline food for a few days.
The following morning we rose before dawn, packed, and got started. Walking through downtown Pamplona (site of the famous 'running of the bulls') we located the Camino, marked with shell symbols and yellow arrows, and began our journey westward. I should note that B had already walked for 4 days from the edge of France, and had waited in Pamplona for me to arrive. Most days the Camino ends in a rather small village which can be hard to get to - as I discovered a few days later.
About half way through the first day I'd finished my first loaf of bread just as we climbed a ridge bedotted with gigantic wind turbines. We arrived in Puenta la Reina at around 1pm. B's feet were rather blistered though mine in my terrible sandals were as yet holding up fine. Puenta la Reina was an ancient walled town built to support pilgrims in the 11th century, with wide arched doors to ground floor stables, walls, and an ancient arched bridge. Every shop was closed until the evening so we relaxed. B was carving a walking stick with an emblem for each day of the trip. That evening we found a bunch of Irish pilgrims (and thus the best food in town), got yelled at by a drunk homeless pilgrim, and took an early night.
The following day we got up around 5:30, obtained mass quantities of carbs, and walked out of the city. B realised he'd forgotten his stick (not the first or last time, I might add), but after retrieving it we continued on through the Spanish countryside. I had brought my video drone, but sadly the transmitter had range issues, so flying it too far away tended to result in abrupt catastrophe. Fortunately I had many spare propellers! Video from the drone can be found here (https://www.youtube.com/watch?v=SvwG6Z0cO8A), though watchable footage is only really possible if there is no wind at all.
I ate my own weight in Spanish tortilla for lunch, and by 1:30 we reached the next town, Estella. B and I found a piano shop in Estella, where we played music for a while. Both of us have not practised much in the last year, and it shows. In most hostels people are up most of the night drinking and being noisy. In the pilgrim albergues, many people were fast asleep by 4pm! Nearly everyone sported gratuitous blisters and quite a few were strapping up various joints. Most people were carrying a lot more weight than they needed! Including me! I could have done without the first aid kit, the cooking pot, the drone, the raincoat, and the solar panel. I helped modify B's flip flops to make them more useful, with a pocket knife and string. They were works of art.
The following morning came with the sad news of the death of Robin Williams, one of my favourite comedians. We struggled on, passing a fountain of free wine, then remarking on the relative lack of pilgrims on the road. Later that evening, we met a few of them who had filled their water bottles with wine and thus achieved a slower but less painful pace! An incredible sunrise illuminated a distant cliff and golden haystacks. Toward midday, our pace slowed and we were passed by numerous pilgrims, including two Russian students in their late teens. I haven't spoken that much Russian in quite some time! Soon after we arrived in Los Arcos and checked into the Austrian albergue, which had a nice kitchen and a small pond in which to soak feet. During the night a strong wind blew up and in my attempt to retrieve some washing before it blew away, I found we'd been locked in - a sobering reinforcement to my usual ritual of identifying two fire/emergency escape routes before bed in any unfamiliar place. What B lacked in mobility he made up for with excessive gregariousness and we made about 20 new friends that afternoon.
The following day was my last day on the Camino. We walked through the stone old-town, many buildings abandoned and derelict, and then popped out into the countryside. The route typically goes through three or four towns a day, with ample opportunities to water and feed oneself. The last town before our destination was Viana, which seemed to suffer from being slowly consumed by the adjacent and much larger Logrono. It had a cool church which had partially collapsed. In Logrono we saw a good example of predatory hostel locating, in which a new hostel positions itself just down the road from an older one, so that sore footed pilgrims drop their pack at the first possible moment. I left B to carve his stick and walked across town to the bus station, where I bought a ticket to Gijon the following day, then back via a supermarket and a few other neat places. That evening B and I went out on the town and found a good place where we ate awesome food and, later, bought some sandals for B so he could join the sandal-wearers cult. Back at the hostel, got involved in a lively discussion about best practises for blister treatment. A lot of people drained them, cut them, or even sewed threads through them overnight! No-one seemed on board with my suggestion of draining and fixing the gap with a small amount of superglue. As a member in long standing of the sandal wearing cult myself, I have often had cause to patch up the odd hole and superglue works perfectly, if you don't overdo it. Sometimes you can even stick your fingers together.
The next morning B left early and I stuck around, walking around the old town of Logrono, buying some breakfast, and eventually finding my bus to Gijon. It took about 6 hours to get to Gijon - quite a long drive, through impossibly rugged mountains dotted with half-abandoned villages, cows, and sometimes huge flocks of eagles. I walked around the beaches of Gijon, stalked by pigeons who recognised my by now high bread content. After a magnificent dinner, I located my couchsurfer. A rather nice dentist who rode a speedy motorcycle, we immediately got on well, and that evening walked into town for a giant fireworks display in celebration of the city's festival.
The following morning I woke, packed, and walked across town to meet up with T, my other mystery friend! I last saw her in late September 2010, so it was great to catch up and see for myself how not-dead she was, despite recent injuries. After a magnificent lunch involving cider poured from a great height, we said goodbye and I walked back into town.
My original plan had been to catch a bus to Santander and enjoy the beach and stay with couchsurfers. My couchsurfers had all fallen through, however, and the coastal weather was windy, cloudy, and cool, so I instead decided to return inland. All the direct buses were booked out, so at 4pm I took a bus to Leon, then a train to Palencia, then another train to Miranda de Ebro, arriving at about midnight. I was relatively certain that B's feet would prevent him making it to the nearby Santo Domingo de la Calzada until the following day, so I found a local hotel and prepared to, you know, roost. But wifi was my friend and after 3 days of radio silence, B got a message through confirming he was, at that minute, already IN Santo Domingo. If I had arrived the following day at 10am or so and walked the wrong way to intercept him, we might never have met. Instead I jumped in a taxi and by 1:30am was in Santo Domingo.
At about 10pm the last hotel closes its doors, so I hunkered down in a doorway where, despite wearing ALL my clothes, by 3am a cold breeze made not moving difficult. I walked around town with my phone out and eventually managed to locate the holy grail - unsecured wifi! I took advantage of several hours of unscheduled time to check all the internet stuff I needed to do, and then walked to the four corners of the town to stare at the stars and talk to sleeping ruminants and count road tiles. At around 6am I found B's albergue and walked straight in, which he found pleasantly surprising. Sandals it turned out had worked for the better and he was now making excellent time. I took a quick nap on the kitchen table while they all packed, then we set out for the next town. It felt good to be back on the Camino, and I was rather wired rather than tired.
By that afternoon, however, 28km of walking plus failing to sleep/lie down had taken its toll and my feet were a little sore. We picked the best of Belorado's five albergues (quattro cantones) and spent the afternoon hanging out in their back yard, which featured a greenhoused pool, a lot with chickens, peacocks, geese, and rabbits, and several adjacent derelict houses. I caught up on laundry, and together with a german vegetarian Z cooked a gigantic dinner. We met a Basque fellow A who was an expert on conflict resolution and had a long and interesting chat about the nature and history of the rebels in northern Spain. Apparently a cease-fire was signed only 2 years ago.
The following day I fried the left-over dinner and we left early, quickly leaving the towns behind and walking the longest stint on a rather terrible trail, perhaps 4 hours between towns. B, Z, and I sang a lot of Disney and other songs on the trail, and eventually we arrived in San Juan de Bottega, another town with one non-abandoned building, a gigantic church containing the remains of several saints and some cool architecture, and a dusty combine harvester. 3km further on we found our destination, Ages. Unlike the medieval architecture of Los Arcos and Puenta La Reina, Ages and Belorado featured rows of larger places in an un-walled setting, perhaps benefiting from nearby larger towns and almost certainly much later construction. Like the other towns, especially ones away from the Camino, Ages was also mostly abandoned, with about a dozen modern renovated homes between many more derelict and abandoned older places. At this point I realised that despite the relaxing nature of life on the Camino, it was, today, a highly privileged existence to spend a month involved in unnecessary physical labour! The majority of pilgrims were not religiously motivated. After a terrific dinner I walked to the outskirts of the town and set up a timelapse with my camera, which can be viewed here: https://www.youtube.com/watch?v=7JOPbhqoU6s
The next morning I retrieved my camera, where the dawn's light revealed a field full of thistles, explaining my previous discomfort! We set out but failed to find a town with a bakery, leaving me rather hungry, for a change. Eventually, on the outskirts of Burgos, I ate a semi-microwaved tortilla and we pressed on. The last 2 hours of the walk was through the outer sprawl of Burgos, one of the most depressingly awful things I've seen in quite a while! Fortunately the old town was lovely. I located the bus station, bought a ticket, and took a shower. Not long after, I said goodbye to B and Z, and 2 hours later was in Bilbao, footsore and exhausted.
So I walked for an hour across the city, taking in the Guggenheim museum and various Calatrava edifices, meeting a couchsurfer at the city hall. With him I walked through a giant fair ground to his place at the top of the tallest hill in Bilbao. Not long after, we returned to the city, climbing down about a thousand steps into the old town, where more than half a million people had gathered to celebrate the city's festival. It was out of control!
I had an early flight the following day so narrowly caught the last bus to the airport. At the airport (another Calatrava design) we were herded into a rather cool room to stay overnight next to the car rental place. At 2am they sent a floor polishing unit around to help us sleep. I guess they could start the buses a bit earlier in the morning! A moderate amount of yoga and armrest limb threading and I could sleep flat, although by 4am I was shivering too much to sleep well.
Later that morning I was in Brussels once more, this time with a 9 hour layover. I boarded a train to Ghent, where I met an old colleague E, who I had not seen since accidentally running into them in early 2011 in San Jose! We spent two hours catching up on our respective news, discussing how best to navigate the academia/industry divide, and being shown the awesomeness of downtown Ghent. All too soon it was time to return to the station and the airport, and take a flight to Washington DC. On the flight I watched the Dark Knight Rises, listened to the Beatles, thought about the Stokeslet Green function, and watched a few other silly films. My neighbours were veteran electrical engineers, specialising in low-voltage switching. That is, less than 11kV! At DC passed through about 15 layers of immigration and security, then boarded the last flight to LA, on which I mostly slept. I got back to my place at around midnight, 38 hours after leaving Burgos, and immediately stayed up for 5 hours collating photos and otherwise having a terrific time.
It is nice to have the freedom to occasionally zoom around the world and indulge in pointless physical exercise for days at a time. Something to consider with career development!
Some time before 28 March, 2011, my dear great aunt R asked me to do a blog post on what my actual research is. What is it that I actually do on a day-to-day basis?
The reasons for the absurd delay in writing this are many, but I think it's time to get on with it, since I finally published something. It can be found at http://arxiv.org/pdf/1406.7029v1.pdf .
In this post, I'll talk generally about what my day-to-day existence is like when I'm not doing exciting stuff I write blog posts about, and then I'll give an executive summary in plain English of what the paper is about.
Here's a photo of me in my office.
I spend a LOT of time here. On the window providing some shade is a gigantic chart of all the isotopic nuclides. It's like a periodic table, but with isotopes too. The story behind the mask is too detailed to relate here.
I'm wearing a shirt I had printed with the Maxwell equation written in exterior differential form - it's a very compact way to write the equations which completely describe electromagnetism. To my right on the desk can be seen some toy magnets, a titanium spork, a spec sheet for an orbital telescope I worked on, some 3D printing stuff, a rocket design manual, and a giant spanner. To my left on the desk is a pile of papers on true polar wander (the process by which the earth tips over), multiple Rubiks cubes, a jar of vegemite, a toothbrush, a pile of books on Chinese, Russian, Indian history, and geophysics. Hanging from my laptop is one of my small drones charging on a USB cable.
Concealed behind me is a gigantic pile of papers dealing with my actual research, which is on gravitational wave propagation algorithms. Plural, because all but one don't work. Proof by exhaustion states that there is a large space of possible algorithms, and the correct one is in there somewhere, and all you have to do is optimize on every possible dimension until you find it. Unfortunately the number of dimensions, and thus potential causes of errors, is practically infinite.
Don't think for a second, though, that simply getting a nerdy shirt and a cluttered desk will result in (possibly correct) algorithm generation. Many, many long hours I've frittered away my youth experimenting with various schemes that not even I think might work. When I started this post, more than three years ago, I'd just started out, and the problem had not even formed. A year later, I hard started on the problem plus a bunch of preliminary work. Two years after starting, I discovered that the general consensus was that the problem was impossible to solve, at about the same time that I had a scheme that seemed like it was thinking about working - enough to present a poster at APS. Since then, every last part of the algorithm has been chopped and changed, like a car that, over its lifetime, becomes an entirely different car through part substitution. Part of the fun of using an object oriented code is that this is even possible with some degree of confidence.
The code is written in C++, one of the most powerful and most widely used high-performance programming languages, and one of the hardest to master. After about 3 years of using it, I can say with confidence I still have not plumbed the depths of my ignorance on the topic. Common knowledge suggests a decade of incidental use will result in reasonably good understanding. I must emphasize, however, that even the most bizarre, difficult to use programming language is still orders of magnitude more simple than the simplest natural languages, even made-up ones like Esperanto.
All that said, what does the paper actually talk about? It's not the sort of scientific paper that announces a new result, or some great discovery. It's more representative of the bulk of scientific papers, representing incremental progress. It's also a methods paper, in that it establishes a new method for solving a known problem. It's also a general paper, in that future papers (probably from me, maybe from other people) will refer back to it as establishing the basic method, and laying out the formalism in modern notation.
The paper contains a lot of equations. None of the equations are particularly new, in the sense that I did not have to discover 'the equation of gravity' in order to do this work. That equation (The Einstein Field Equation) has been known for nearly 100 years. Any new equation would not be correct, because that is the correct equation. But it's actually a horrendously complicated equation, and we have to slice it up into (about 50 different) manageable parts to solve on the computer. People have been trying since the 1960s to solve this equation on the computer, but only succeeded starting in 2005. Many things can make computers go wrong when solving equations, and this equation can and does produce all those problems, plus some new ones. For example, when simulating black holes, the simplest choice of coordinate system is called inertial, which means the coordinate system you'd get if each grid point was in freefall. It's simple because there are no accelerations, and it's the coordinate system you get in space when there's no apparent gravity. But a blackhole quickly swallows that coordinate system, which is very sad.
Fortunately, I work with an existing code base of more than 100,000 lines, written by many collaborators in my group over the last decade or so. This provides a LOT of functionality which just doesn't exist in many other codes or programming languages. This allows me to focus on the algorithms themselves, instead of getting bogged down with a lot of low level functions.
The paper starts with a general discussion of the problem that my work directly addresses. The existing code is tremendously powerful and successful, and simulates black holes that inspiral, merge, and ring down, emitting gravitational waves in the process. We hope to detect those waves on earth, which will provide us with new insight into these peculiar and high-energy regimes. The simulation volume is typically similar in size to earth's moon. Even at its outer edge, the grid points are very close to the action, and the coordinate system is shaken by the gravitational waves. This makes unique determination of the gravitational waves difficult, because it's difficult to see what's going on when everything is shaking around.
Unfortunately, even a little bit of error is unacceptable if we want to have a hope of getting useful physics out of gravitational wave astronomy. So there's an equation which allows extraction of gravitational radiation from the simulations, by simulating the space between the simulation and infinity. Unfortunately, the equation is so complicated that there is only one other code that does it. Written in the 90s, it is now too slow. Moreover, it does not exploit the advances that have been made with the code we use in our group.
Often in science progress is made when a fresh face (i.e. me) approaches a problem they don't know everyone else thinks is impossible. Or at least extremely technically annoying. And the first 10 or so attempts at solving the problem didn't work. At first, I tried to understand the process on paper, and later by prototyping using a simpler computer code and a simplified problem. Eventually I built up my understanding of the tools and system behaviour, which allowed better guesses at avenues for exploration. At the same time I was learning how the code worked and how to write new parts of it.
There were two central technical tricks I developed and exploited for this research. To the best of my knowledge, they've never been employed together in this fashion, and for the well being of grad students everywhere, I hope they never are again!
The first trick is one that just shouldn't work on computers, but somehow it does. The problem arises because the equations explicitly diverge on both sides of the equals sign. Eg, we want to solve for Q, where F(Q) = RHS(other stuff). Q is finite in the domain, but F(Q) goes to infinity as radius goes to infinity. Using some moderately complicated maths it's possible to invert F and write Q = F^(-1)(RHS(other stuff)), but getting it to work on a computer, where numbers can't be infinite and so on, took some finagling. I explain this in the paper in great detail, but suffice to say, without the code's ability to perform extremely accurate and quick radial integrals, this method would not work at all. All hail the Fast Fourier Transform!
The second trick is both better established in the literature and less well known among my advisers, and it's called the Magnus expansion. As it happened I worked out the basics for how the method would work, tried to devise a name for it, googled the name, and found someone else (Magnus, as it turns out) had already discovered this rather obvious technique and, what's more, had written a paper doing all the hard sums for me. In this case, the Magnus expansion was applied to a matrix which represented some complex numbers. The equation had an annoying part which could be thought of as a phase. The Magnus expansion allowed this phase to be set to zero, and ignored until it could be dealt with in a consistent fashion. Again, incredibly seat-of-the-pants numerical methods, and I'm just lucky that when all was said and done I'd only wasted about 14 bits of precision, leaving me with perhaps 6 orders of magnitude to play with. For people who know the difference between single and double precision floating point numbers, the difference makes ALL the difference.
Okay, I had the basics of this method finalized a year ago, and permitted myself a shy smile. There were a bunch of other technicalities covered in the appendices related to translating information from the simulations into the form where my algorithm could deal with it. I crushed that problem on only the third try. At this point I could have published yet another theory paper, but if you want the method to be used, or for anyone to take notice, you need to actually prove that it works. Especially when the underlying methodology is so left-field and the apparent efficiency increase (hundreds or thousands of times faster) so unlikely seeming.
To prove that it works you need to show that its stable by feeding it garbage and watching it resolutely refusing to explode. You also need to show that it converges. That is, as you increase the resolution, it gets closer and closer to a particular value in a predictable and most importantly, extremely rapid way. The prescription for these sorts of graphs is well known, and its in this phase that cryptic bugs are eventually discovered. Sometimes they even have a comment nearby saying "\\this will cause problems I bet...".
When the code behaves itself, you finally need to compare it to results produced with another similar code. Afterall, it could converge to garbage, which would be useful to noone. In this case, it was compared to the code it was designed to replace. First I had to convince this no-nonsense older code to do what I wanted it to do, and then get the parameters right to prevent its own menagerie of stability and convergence issues. Principle problems include it never being designed to do what we had in mind. Similarly, my code was never designed to operate with parameters that could compare to the older code. After a few decades of wasted CPU time and more than a few false starts, we generated comparable data sets. By this time, enough of the problems had been worked out that they were close enough to publish, as you can see in Figure 8 of the paper. This figure shows that over the absurdly long time scale we ran this code, both codes were pretty accurate to begin with. Moreover, when the resolution was increased, the error dropped by a certain, predictable amount. Lastly, the difference between the two codes is also really tiny. Of course, there are always a few problems. For instance, the older code begins to lose convergence toward the end. Also, the codes resolve the junk radiation to differing extents.
Junk radiation is nonsense signal at the beginning caused by the fact that the simulations have to start somewhere. This insult to well behaved computers causes unphysical fake waves to bounce around inside the simulation and sometimes even stop it working properly. We think my code might be able to stop or reduce junk radiation in the future. Stay tuned!
It bears mentioning that this work was a team effort! I don't name names on this blog, but the most overt contributors are listed on the paper. For a more complete list, you'll have to wait for my thesis! But you know who you are, and without you, there is a good chance we would have never known this problem was, indeed, possible to solve.
