There is a new top candidate for the best place to search for life outside our Solar System. Announced in Nature on April 20th 2017, LHS 1140b is a super-Earth in the habitable zone of a nearby small star. This is the story of that discovery and the role PEST played.
29th July 2016. Joey Rodriguez, whom I know from working with KELT, emailed Ivan Curtis (another Australian amateur astronomer) and me to ask if we would be interested to observe a candidate from MEarth, with a predicted transit on 7th August. The candidate planet was a super-Earth with a suspected ~25 day period. Right up the PEST alley, so I said yes.
Shortly after, Jonathan Irwin (MEarth) sent me target information. The star was LHS 1140. At first glance, the star seemed too dim to get good precision. V = 14 is equivalent to a candle 400km away. But look again; Ic = 11. Well that’s more like it! The star is much brighter in infra-red than in visible light, so that’s what PEST will observe in.
“We have three slightly iffy partial transits from MEarth-South that predict this ephemeris, the main aim here is confirmation that there is a transit in the predicted window. The ephemeris is the main missing piece of the puzzle holding back further progress (it’s already radial velocity confirmed)…”
The MEarth telescopes had first seen a possible transit on 15th Sept 2014. At the time, this tiny dip in the light at the start of the transit caused the MEarth computers to immediately trigger more intensive observations than the normal frequency of 30 mins per target star. But at the time, no human noticed the transit. Many triggers are due to wind or other changes in atmospheric conditions, so this event was lost in the crowd.
It was only a year later, when Jason Dittmann (lead author of the paper, then a graduate student at Harvard) applied a new machine learning technique to the data that this transit popped up as being credible. Of course seeing one possible transit does not confirm it is real. But it was credible enough that MEarth commissioned radial velocity measurements using the HARPS instrument. HARPS, run by the European Southern Observatory and located in Chile, is the world’s premier planet hunting spectrograph which measures the wobble of stars such as caused by an orbiting planet.
HARPS intensively observed LHS 1140 starting in Nov 2015. On 19th June 2016 it happened again – two MEarth telescopes caught another transit. Armed with transit as well as radial velocity data Dittmann and team identified 3 potential periods for the planet. One of these, 25 days, meant that a transit observable by MEarth would have taken place on 23rd Dec 2014. Indeed, inspection of data from that night showed a few points suggestive of a transit – but it wasn’t definitive. To confirm this 25 day period, another transit would have to be caught. The next one predicted was 7th August 2016, observable from Australia.
The Perth weather forecast for 7th August: “Very high chance of showers with a cold front coming through”. Ivan Curtis in Adelaide did not have much luck either – “The sky is not looking good right now, and the forecast is poor”. Irwin wrote back to say, “I’ll look into possibilities for the next event, which is 1 September”.
On 29th August I wrote to Irwin, “After some bad weather the last few weeks, there might be a break with 1st Sept forecast to be ‘mostly sunny’.” But the 1st Sept transit had a predicted mid-point of 9:50 pm Perth time, whereas the star would be high enough for me to observe only at 10 pm. This meant that if the transit was on time, I would only be able to capture half of it.
My location was not ideal. Telescopes in eastern Australia or even further east, would be better placed. Irwin had made alternative arrangements, “We’ve organised a campaign from Mt Kent, Siding Spring and Hawaii to try to target most of the window, however if you did still want to try observing this a robustly detected egress may be enough”. In other words, six telescopes were lined up but he would take what he could get. Dittmann remembers this as being a “frantic night”.
The night of 1st Sept turned out to be perfect with clear skies. PEST was programmed that evening to turn its stare towards LHS 1140 as soon as it was high enough, and to observe through the night. Data trickled into my home office computer as I slept.
After work the next day I worked through the data. This chart, each dot representing one measure of the brightness of the star, was the result.
Most of the points were at about the same brightness level – except for about a dozen points at the start. The difference was minute – just 0.5%, but consistent with what MEarth had seen. Could this be real? Sometimes other effects masquerade as a signal. I poked at the data, changing parameters, to see if the signal could be made to disappear. The signal stubbornly persisted.
An email went back to Rodriguez and Irwin with the tell tale light curve and,
“Transit egress seen at ~HJD +7633.12. Depth about 5 mmag.”
If PEST had started to observe just half an hour later, I would have missed the transit. We had it – we had confirmed the period of LHS 1140b to be 24.7 days. None of the other telescopes lined up to observe were able to, mostly because of weather.
The period of the planet was now in hand. MEarth observed two more transits, later that September, and in October 2016. All the photometric and radial velocity data converged towards LHS 1140b being a rocky planet in the habitable zone of a small, placid, red star. The faint blue fringe in the artist’s impression above captures our hope for a life-sustaining atmosphere.