How can we de­tect slow-or­bit­ing plan­ets around other stars?


He­len Giles is em­ploy­ing a new technique to quickly dis­cover hard-to-spot worlds that take many years to or­bit their stars

Most of the thou­sands of exoplanets dis­cov­ered in re­cent years have been hot Jupiters. The rea­son for this is not be­cause they are more com­mon but be­cause they are much eas­ier to find and quicker to con­firm.

Most exoplanets have been found us­ing the tran­sit method, where they pass in front of the star, pro­duc­ing a mea­sur­able dip in the bright­ness of the starlight. Mas­sive plan­ets in rapid or­bits pro­duce more no­tice­able dips thanks to their sheer size.

NASA’s Ke­pler space te­le­scope also helped us find a num­ber of plan­ets with longer or­bital pe­ri­ods, but that was sim­ply be­cause Ke­pler ob­served stars for four years dur­ing its ini­tial mis­sion. Oth­er­wise, th­ese slower-mov­ing plan­ets have so far mostly been dis­cov­ered us­ing ra­dial ve­loc­ity, where a ground-based spec­troscopy in­stru­ment mea­sures the wob­ble in a star’s light. As­tronomers have been con­duct­ing searches with ra­dial ve­loc­ity a lot longer than we’ve been do­ing tran­sit pho­tom­e­try.

The eye of the be­holder

Com­puter al­go­rithms are used to de­tect sig­nals in Ke­pler data that in­di­cate the ex­is­tence of an exoplanet. I found that ap­ply­ing hu­man judge­ment al­lowed me to find a can­di­date for an exoplanet quite un­like the ma­jor­ity of dis­cov­er­ies, which had the long­est pe­riod of any planet spot­ted by Ke­pler.

My dis­cov­ery came about when I was scan­ning the data from K2 (the fol­low-up mis­sion from Ke­pler) try­ing to find any sort of exoplanet – I didn’t re­ally mind what kind of planet or what its pe­riod was. The com­puter code ranks po­ten­tial exoplanet can­di­dates from the best look­ing to the worst, and I checked every sin­gle one by eye. This planet stuck out like a sore thumb.

It was ranked very low among the pos­si­bil­i­ties, since only one po­ten­tial tran­sit had been ob­served by K2. But as soon as I ap­plied my hu­man eye to look at the data, I saw a beau­ti­fully-shaped dip in the star’s light-curve and thought, “Okay, that looks in­ter­est­ing, let’s in­ves­ti­gate fur­ther.” Tran­sits of most de­tected exoplanets typ­i­cally last be­tween two to 16 hours, and it is stan­dard prac­tice to ob­serve at least three passes be­fore the exoplanet is con­firmed. My exoplanet’s tran­sit lasted 54 hours – al­most two and a half days. Hav­ing to wait three or­bits to con­firm it would take over 30 years at that rate. So in­stead my team at the Univer­sity of Geneva used other ob­ser­va­tions of the star to learn more about it and its po­ten­tial planet. The par­ent star is known by its cat­a­logue num­ber EPIC248847­494. We first used re­cently re­leased data from ESA’s Gaia mis­sion to de­ter­mine that it lies about 1,500 lightyears from us. That in­for­ma­tion, to­gether with the length of the tran­sit, told us that the planet’s dis­tance from its star must be around 4.5 times that of the Earth from the Sun, as well as the length of its ‘year’. Im­por­tantly, we needed to check that this or­bit­ing ob­ject was a planet and not an­other star. We turned to the Swiss 1.2m Leon­hard Euler Te­le­scope at La Silla in Chile to mea­sure the star’s ra­dial ve­loc­ity, or, more sim­ply, its wob­ble. This showed that the newly dis­cov­ered com­pan­ion is less than 13 times as mas­sive as Jupiter, so too small to be a star. Of­fi­cially this dis­cov­ery is still just a great can­di­date for an exoplanet, but a very con­vinc­ing one! To get con­fir­ma­tion us­ing ra­dial ve­loc­ity will take a full or­bit of 10 years, though we should get some es­ti­ma­tion in around half that time. This demon­strates a use­ful technique for find­ing more slow-or­bit­ing exoplanets. Some might al­ready lurk in archived data; find­ing them would be a good cit­i­zen sci­ence pro­ject. But the main fo­cus will be on fu­ture data col­lected by the likes of NASA’s Tran­sit­ing Exoplanet Sur­vey Satel­lite (TESS), which launched in April. TESS will only ob­serve each bit of sky for 27 days, so it should de­tect a lot of sin­gle tran­sits. We look for­ward to dis­cov­er­ing many more exoplanets. In the­ory, we could de­tect a planet as small as Earth, if its star is fairly sta­ble and bright.

We could be dis­cov­er­ing more exoplanets thataren’t hot Jupiters thanks to a new technique that re­quires a hu­man touch

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.