1 00:00:06,519 --> 00:00:03,350 well good morning or good afternoon 2 00:00:10,480 --> 00:00:06,529 everyone depending on your time zone 3 00:00:14,150 --> 00:00:10,490 it's morning back here on the west coast 4 00:00:17,810 --> 00:00:14,160 I'm Carl Pilcher and I am going to 5 00:00:20,659 --> 00:00:17,820 introduce our speaker today for the NAI 6 00:00:23,120 --> 00:00:20,669 director seminar our speakers Drake gain 7 00:00:24,710 --> 00:00:23,130 from Goddard Space Flight Center a 8 00:00:27,439 --> 00:00:24,720 member of the Goddard Space Flight 9 00:00:29,179 --> 00:00:27,449 Center team I've known drank for a lot 10 00:00:32,319 --> 00:00:29,189 of years and I'm really glad that he was 11 00:00:34,670 --> 00:00:32,329 able to give this but this talk because 12 00:00:39,290 --> 00:00:34,680 there have been some very exciting 13 00:00:42,440 --> 00:00:39,300 recent results particularly based on the 14 00:00:44,510 --> 00:00:42,450 spitzer space craft data and Drake is 15 00:00:46,400 --> 00:00:44,520 also going to tell us about how those 16 00:00:49,540 --> 00:00:46,410 kinds of results might be extrapolated 17 00:00:52,460 --> 00:00:49,550 to what would be possible with jwst 18 00:00:56,080 --> 00:00:52,470 Drake received his PhD in astronomy in 19 00:00:59,000 --> 00:00:56,090 1976 from the University of Illinois at 20 00:01:00,170 --> 00:00:59,010 champaign-urbana and spend some time at 21 00:01:01,940 --> 00:01:00,180 the astronomy department at the 22 00:01:06,279 --> 00:01:01,950 University of Maryland came to Goddard 23 00:01:09,169 --> 00:01:06,289 in 1979 and he has been there ever since 24 00:01:11,560 --> 00:01:09,179 working on infrared observations of 25 00:01:14,749 --> 00:01:11,570 solar system objects and obviously he's 26 00:01:16,310 --> 00:01:14,759 expanded his purview to observations of 27 00:01:18,529 --> 00:01:16,320 planetary objects and other solar 28 00:01:20,419 --> 00:01:18,539 systems so today Drake's going to tell 29 00:01:22,249 --> 00:01:20,429 us about infrared spectra of extrasolar 30 00:01:27,099 --> 00:01:22,259 planets and I will turn it over without 31 00:01:29,480 --> 00:01:27,109 further ado to Drake okay Thank You Carl 32 00:01:31,249 --> 00:01:29,490 first I want to acknowledge my 33 00:01:34,999 --> 00:01:31,259 collaborators in this work Jeremy 34 00:01:38,899 --> 00:01:35,009 Richardson was a NASA postdoctoral 35 00:01:41,270 --> 00:01:38,909 fellow at Goddard and actually I wasn't 36 00:01:43,069 --> 00:01:41,280 his advisor he was Jeremy worked in a 37 00:01:45,010 --> 00:01:43,079 couple of different laboratories and his 38 00:01:47,389 --> 00:01:45,020 advisor was filled Angie over in the 39 00:01:49,879 --> 00:01:47,399 exoplanet and stellar astrophysics 40 00:01:51,289 --> 00:01:49,889 laboratory Karen Horning is an 41 00:01:53,239 --> 00:01:51,299 undergraduate student at Florida 42 00:01:56,779 --> 00:01:53,249 Institute of Technology and she was our 43 00:01:58,639 --> 00:01:56,789 summer astrobiology intern last summer 44 00:02:01,219 --> 00:01:58,649 and Karen did a lot of the really hard 45 00:02:02,090 --> 00:02:01,229 work that's going to make the analysis 46 00:02:04,219 --> 00:02:02,100 that I'm going to show you today 47 00:02:06,050 --> 00:02:04,229 possible Sara Seager is our team's 48 00:02:08,570 --> 00:02:06,060 theorist and in the early stages of this 49 00:02:08,779 --> 00:02:08,580 work she was up to the Carnegie node and 50 00:02:11,630 --> 00:02:08,789 she 51 00:02:13,520 --> 00:02:11,640 since moved to MIT and joe harrington is 52 00:02:17,420 --> 00:02:13,530 our collaborator at the University of 53 00:02:19,759 --> 00:02:17,430 Central Florida and this this talk is 54 00:02:21,740 --> 00:02:19,769 really divided into a series of topics 55 00:02:23,899 --> 00:02:21,750 first since i'm talking about spectra of 56 00:02:27,470 --> 00:02:23,909 extrasolar planets i'm going to just 57 00:02:29,569 --> 00:02:27,480 review that very briefly the status of 58 00:02:32,929 --> 00:02:29,579 that field the the known extrasolar 59 00:02:34,610 --> 00:02:32,939 planets extend from hot Jupiters to in 60 00:02:37,879 --> 00:02:34,620 one or two cases some planets that could 61 00:02:40,039 --> 00:02:37,889 be arguably called potters and I'll 62 00:02:41,899 --> 00:02:40,049 mention the possible detection of 63 00:02:44,780 --> 00:02:41,909 biomarkers I'm not going to show you any 64 00:02:47,589 --> 00:02:44,790 actual biomarkers these are just our 65 00:02:49,699 --> 00:02:47,599 future goal but I'll try to relate the 66 00:02:51,860 --> 00:02:49,709 spectra that we that i'm going to show 67 00:02:55,909 --> 00:02:51,870 you for hot Jupiters to the eventual 68 00:02:57,949 --> 00:02:55,919 detection of biomarkers now in order to 69 00:03:01,069 --> 00:02:57,959 show you that we've measured spectra of 70 00:03:04,009 --> 00:03:01,079 two extrasolar planets i first have to 71 00:03:06,440 --> 00:03:04,019 show you that we can detect photons from 72 00:03:09,470 --> 00:03:06,450 extrasolar planets at all so i'll 73 00:03:11,720 --> 00:03:09,480 briefly describe the first spitzer 74 00:03:15,409 --> 00:03:11,730 detections and follow-up detection using 75 00:03:19,159 --> 00:03:15,419 spitzer using photometry to just measure 76 00:03:20,509 --> 00:03:19,169 the planets in a photometric mode and 77 00:03:22,399 --> 00:03:20,519 then i'll talk about measuring the 78 00:03:24,969 --> 00:03:22,409 spectra of giant planets with spitzer 79 00:03:27,379 --> 00:03:24,979 i'll show you two extrasolar planets 80 00:03:29,649 --> 00:03:27,389 specter of two extrasolar planets one 81 00:03:31,699 --> 00:03:29,659 measured by our group and another by 82 00:03:33,740 --> 00:03:31,709 another group that we work in parallel 83 00:03:35,869 --> 00:03:33,750 with and I'll finish up by saying how 84 00:03:38,479 --> 00:03:35,879 this might eventually lead to this 85 00:03:43,399 --> 00:03:38,489 spectra of what we call hotter or close 86 00:03:46,670 --> 00:03:43,409 in extra solar terrestrial planets okay 87 00:03:48,170 --> 00:03:46,680 well since 1995 the Doppler groups have 88 00:03:51,050 --> 00:03:48,180 detected the bulk of the extrasolar 89 00:03:53,240 --> 00:03:51,060 planets and the way to remind you how 90 00:03:55,460 --> 00:03:53,250 the this works is the Doppler groups 91 00:03:57,199 --> 00:03:55,470 will look at a stellar system and they 92 00:03:59,030 --> 00:03:57,209 won't detect the planet directly that is 93 00:04:00,860 --> 00:03:59,040 they won't see photons from the planet 94 00:04:03,050 --> 00:04:00,870 but they'll see the Doppler reflex of 95 00:04:06,649 --> 00:04:03,060 the parent star as it orbits the center 96 00:04:10,280 --> 00:04:06,659 of mass of its planetary system one of 97 00:04:11,749 --> 00:04:10,290 the earliest surprises of the results 98 00:04:13,369 --> 00:04:11,759 from this group was the fact the very 99 00:04:16,460 --> 00:04:13,379 first planet that was detected was 100 00:04:18,379 --> 00:04:16,470 orbiting the star 51 pegasi and it was 101 00:04:20,659 --> 00:04:18,389 very it was a giant planet of nearly 102 00:04:22,290 --> 00:04:20,669 jupiter-mass and it was in very close 103 00:04:25,770 --> 00:04:22,300 point 05 a you 104 00:04:27,659 --> 00:04:25,780 that's 20 times closer to the to its 105 00:04:29,219 --> 00:04:27,669 star than the earth is to the Sun and 106 00:04:31,170 --> 00:04:29,229 these planets have been called by 107 00:04:32,850 --> 00:04:31,180 various names I prefer to call them hot 108 00:04:35,460 --> 00:04:32,860 Jupiters because they're jupiter-mass 109 00:04:36,960 --> 00:04:35,470 but they're in close to the star and in 110 00:04:39,990 --> 00:04:36,970 that close to the star they're going to 111 00:04:42,149 --> 00:04:40,000 be very strongly irradiated on one side 112 00:04:45,149 --> 00:04:42,159 as as envisioned here by Greg Laughlin 113 00:04:46,499 --> 00:04:45,159 and James Cho by stellar irradiation 114 00:04:50,430 --> 00:04:46,509 where they're going to be heated to very 115 00:04:52,589 --> 00:04:50,440 high temperatures now of course 116 00:04:54,719 --> 00:04:52,599 ultimately what we want to do in the 117 00:04:56,939 --> 00:04:54,729 field of astrobiology has concerns 118 00:04:58,950 --> 00:04:56,949 extrasolar planets is we would like to 119 00:05:01,830 --> 00:04:58,960 measure the spectrum of an earth-like 120 00:05:03,689 --> 00:05:01,840 planet or something at least close to an 121 00:05:05,670 --> 00:05:03,699 earth-like planet being in the habitable 122 00:05:09,260 --> 00:05:05,680 zone and this is some material from 123 00:05:11,610 --> 00:05:09,270 vikki meadows showing model spectra of 124 00:05:13,499 --> 00:05:11,620 first like planets where you could see 125 00:05:16,969 --> 00:05:13,509 an ozone signature near nine and a half 126 00:05:20,850 --> 00:05:16,979 microns methane near 7.8 microns 127 00:05:23,719 --> 00:05:20,860 possibly co2 at 15 microns and to 128 00:05:25,890 --> 00:05:23,729 measure the spectra that the 129 00:05:27,089 --> 00:05:25,900 technologically difficult but 130 00:05:30,059 --> 00:05:27,099 conceptually the most straightforward 131 00:05:32,969 --> 00:05:30,069 way would be to make it a highly 132 00:05:34,649 --> 00:05:32,979 specially resolved image of a star 133 00:05:36,779 --> 00:05:34,659 containing the planet and then to 134 00:05:39,240 --> 00:05:36,789 isolate the planet and observe it with 135 00:05:40,980 --> 00:05:39,250 some very advanced space mission and if 136 00:05:43,439 --> 00:05:40,990 you can separate stellar and Planetary 137 00:05:45,029 --> 00:05:43,449 photons by some high technology method 138 00:05:47,040 --> 00:05:45,039 like this then you can just take these 139 00:05:48,510 --> 00:05:47,050 planetary photons you can put them 140 00:05:50,519 --> 00:05:48,520 through your spectrograph and you have a 141 00:05:52,469 --> 00:05:50,529 spectrum and it's a simple proposition 142 00:05:55,110 --> 00:05:52,479 but technologically this is very 143 00:05:58,350 --> 00:05:55,120 difficult and it will be very expensive 144 00:06:00,480 --> 00:05:58,360 and it hasn't been done yet but for 145 00:06:02,999 --> 00:06:00,490 those of us who are interested in seeing 146 00:06:04,800 --> 00:06:03,009 this done we can't wait so we're 147 00:06:08,209 --> 00:06:04,810 pursuing other techniques and those 148 00:06:11,159 --> 00:06:08,219 other techniques were used transits um 149 00:06:13,619 --> 00:06:11,169 since many of these planets are close in 150 00:06:15,659 --> 00:06:13,629 to their star a closed-end planet has a 151 00:06:18,899 --> 00:06:15,669 high probability to transit its star 152 00:06:22,379 --> 00:06:18,909 that is to cross in front of its star as 153 00:06:24,360 --> 00:06:22,389 seen from the earth the transit 154 00:06:27,180 --> 00:06:24,370 probability if you calculate it is is 155 00:06:30,059 --> 00:06:27,190 the radius of the star divided by the 156 00:06:31,889 --> 00:06:30,069 orbital radius of the planet and for the 157 00:06:33,460 --> 00:06:31,899 closed-end hot Jupiter type planets 158 00:06:37,180 --> 00:06:33,470 that's about ten percent 159 00:06:38,710 --> 00:06:37,190 and early on in the history of this 160 00:06:41,140 --> 00:06:38,720 field when there are about a dozen such 161 00:06:44,530 --> 00:06:41,150 planets known were in clothes from the 162 00:06:45,940 --> 00:06:44,540 Doppler surveys statistically one of 163 00:06:48,310 --> 00:06:45,950 them should transit and that one that 164 00:06:52,900 --> 00:06:48,320 was discovered to transit was hd2 or 945 165 00:06:54,760 --> 00:06:52,910 8b which mrs. Hubble observations of the 166 00:06:57,820 --> 00:06:54,770 transit of that made from space very 167 00:06:59,560 --> 00:06:57,830 high quality very low noise photometry 168 00:07:02,440 --> 00:06:59,570 the transit depth the amount of light 169 00:07:04,710 --> 00:07:02,450 that the planet blocks is about one and 170 00:07:07,210 --> 00:07:04,720 a half percent one-point-six percent 171 00:07:09,970 --> 00:07:07,220 there are now 14 of these known 172 00:07:11,020 --> 00:07:09,980 transiting bright solar-type stars and 173 00:07:12,580 --> 00:07:11,030 of course the transit is very 174 00:07:15,280 --> 00:07:12,590 interesting the depth of this transit 175 00:07:17,680 --> 00:07:15,290 tells us tells us the amount of stellar 176 00:07:20,320 --> 00:07:17,690 life that the planet blocks but we're 177 00:07:22,540 --> 00:07:20,330 not detecting light from the planet this 178 00:07:25,750 --> 00:07:22,550 way in order to detect light from the 179 00:07:28,810 --> 00:07:25,760 planet we have to use what's called the 180 00:07:31,480 --> 00:07:28,820 secondary eclipse the transit is the 181 00:07:33,610 --> 00:07:31,490 primary eclipse planet canít planet 182 00:07:35,710 --> 00:07:33,620 passing in front of the star that tells 183 00:07:37,480 --> 00:07:35,720 us the size of the planet and in 184 00:07:39,280 --> 00:07:37,490 principle we can see and in fact it has 185 00:07:40,810 --> 00:07:39,290 been done radiation from the star 186 00:07:42,580 --> 00:07:40,820 transmitted through the planet's 187 00:07:44,409 --> 00:07:42,590 atmosphere so in fact the first 188 00:07:47,230 --> 00:07:44,419 detection of an atmosphere was made this 189 00:07:49,810 --> 00:07:47,240 way but we can't see photons emitted by 190 00:07:51,909 --> 00:07:49,820 the planet in order to do that we have 191 00:07:54,640 --> 00:07:51,919 to go to the secondary eclipse when the 192 00:07:56,560 --> 00:07:54,650 planet passes behind the star then if we 193 00:07:58,690 --> 00:07:56,570 look in the combined light of the system 194 00:08:00,630 --> 00:07:58,700 with no attempt to specially resolve the 195 00:08:03,040 --> 00:08:00,640 system we will see thermal radiation 196 00:08:06,010 --> 00:08:03,050 from the planet disappear and reappear 197 00:08:07,540 --> 00:08:06,020 and in so doing we will learn particular 198 00:08:09,969 --> 00:08:07,550 about the atmosphere of that planet 199 00:08:12,280 --> 00:08:09,979 because that thermal radiation that we 200 00:08:13,719 --> 00:08:12,290 receive emerges ultimately from the 201 00:08:15,610 --> 00:08:13,729 atmosphere of the planet will also learn 202 00:08:17,409 --> 00:08:15,620 about its orbit and other things but I'm 203 00:08:20,200 --> 00:08:17,419 going to focus here mainly on the 204 00:08:24,219 --> 00:08:20,210 spectra of two of these planets which 205 00:08:26,950 --> 00:08:24,229 tell us about their atmosphere here are 206 00:08:28,630 --> 00:08:26,960 the first detection of photons from 207 00:08:30,550 --> 00:08:28,640 extrasolar planets they were made with 208 00:08:33,659 --> 00:08:30,560 the Spitzer Space Telescope almost 209 00:08:36,940 --> 00:08:33,669 exactly two years ago this month and 210 00:08:39,610 --> 00:08:36,950 there were in fact done one of the great 211 00:08:42,240 --> 00:08:39,620 coincidences of astronomy is that the 212 00:08:43,800 --> 00:08:42,250 two detection papers were submitted 213 00:08:45,840 --> 00:08:43,810 on the same day two different journals 214 00:08:48,960 --> 00:08:45,850 completely uncoordinated the two groups 215 00:08:50,820 --> 00:08:48,970 had known no we're not courting the 216 00:08:54,570 --> 00:08:50,830 coordinating their measurements our 217 00:08:57,660 --> 00:08:54,580 group detected ht-29 458 be using the 218 00:08:59,190 --> 00:08:57,670 using spitzer at 24 microns this shows 219 00:09:02,430 --> 00:08:59,200 the secondary eclipse where the light 220 00:09:04,230 --> 00:09:02,440 from the planet goes goes away then the 221 00:09:06,570 --> 00:09:04,240 radiation comes back again after Eclipse 222 00:09:08,670 --> 00:09:06,580 and then a group led by Dave Charbonneau 223 00:09:11,040 --> 00:09:08,680 at Harvard detected the planet called 224 00:09:12,780 --> 00:09:11,050 trace one at eight microns and they also 225 00:09:16,070 --> 00:09:12,790 detected it at four and a half microns 226 00:09:18,540 --> 00:09:16,080 which isn't isn't shown here now the 227 00:09:20,160 --> 00:09:18,550 eclipse depth of the secondary eclipse 228 00:09:22,320 --> 00:09:20,170 is just the depth of the it's the 229 00:09:24,210 --> 00:09:22,330 projected area ratio this is really just 230 00:09:25,470 --> 00:09:24,220 the depth of the transit and in the 231 00:09:27,090 --> 00:09:25,480 rayleigh jeans limit there's an 232 00:09:29,070 --> 00:09:27,100 additional factor which is the ratio of 233 00:09:32,490 --> 00:09:29,080 temperatures and this is something like 234 00:09:34,560 --> 00:09:32,500 one fifth row one quarter so these these 235 00:09:36,000 --> 00:09:34,570 secondary eclipses are less deep than 236 00:09:37,850 --> 00:09:36,010 the transits and that's why there of 237 00:09:40,830 --> 00:09:37,860 course much noisier than transits and 238 00:09:43,050 --> 00:09:40,840 these initial detection were you know 239 00:09:44,850 --> 00:09:43,060 you can see that the data have a fair 240 00:09:46,530 --> 00:09:44,860 amount of scatter an individual point 241 00:09:49,740 --> 00:09:46,540 wouldn't really be sufficient to detect 242 00:09:51,420 --> 00:09:49,750 these eclipses but in aggregate each of 243 00:09:54,780 --> 00:09:51,430 these eclipses is about a six sigma 244 00:09:58,140 --> 00:09:54,790 detection and the temperature is about 245 00:09:59,670 --> 00:09:58,150 eleven hundred Kelvin now although these 246 00:10:01,410 --> 00:09:59,680 were you know sort of two independent 247 00:10:04,800 --> 00:10:01,420 groups and we both got about the same 248 00:10:06,240 --> 00:10:04,810 result on to two different planets using 249 00:10:07,590 --> 00:10:06,250 different Spitzer instruments 250 00:10:09,990 --> 00:10:07,600 nevertheless if you were a really 251 00:10:12,000 --> 00:10:10,000 skeptical person you might say well at 252 00:10:14,490 --> 00:10:12,010 all six sigma maybe I don't really 253 00:10:16,530 --> 00:10:14,500 believe that well I will in that case I 254 00:10:18,930 --> 00:10:16,540 I welcome that skepticism i will show 255 00:10:24,090 --> 00:10:18,940 you an additional follow-up detection 256 00:10:26,460 --> 00:10:24,100 because a later in 2005 an additional 257 00:10:28,430 --> 00:10:26,470 planet was detected that has an even 258 00:10:31,110 --> 00:10:28,440 greater transit dip and an even greater 259 00:10:33,360 --> 00:10:31,120 secondary eclipse depth due to the fact 260 00:10:36,420 --> 00:10:33,370 that the star that it that it orbits is 261 00:10:38,820 --> 00:10:36,430 much smaller to the ratio of planetary 262 00:10:41,730 --> 00:10:38,830 flux stellar flux is much better and 263 00:10:43,680 --> 00:10:41,740 that's HD 189733 B and here's its 264 00:10:45,600 --> 00:10:43,690 secondary eclipse observed by Spitzer 265 00:10:47,970 --> 00:10:45,610 although you know maybe you could be 266 00:10:49,380 --> 00:10:47,980 skeptical over first detection is no 267 00:10:51,960 --> 00:10:49,390 one's going to be skeptical of this one 268 00:10:53,210 --> 00:10:51,970 this is a 32 sigma detection of this 269 00:10:55,879 --> 00:10:53,220 planet at 16 270 00:10:58,340 --> 00:10:55,889 Brown's man this is the unbend or all 271 00:11:00,019 --> 00:10:58,350 the original spitzer measurements all of 272 00:11:03,740 --> 00:11:00,029 them every 10 seconds and this is a bin 273 00:11:05,780 --> 00:11:03,750 plot just to show it more clearly again 274 00:11:08,150 --> 00:11:05,790 the deputy clips depth is the ratio of 275 00:11:10,420 --> 00:11:08,160 the projected area of the planet to 276 00:11:12,860 --> 00:11:10,430 start times the temperature ratio and 277 00:11:14,119 --> 00:11:12,870 what I wanted to point out with this is 278 00:11:17,210 --> 00:11:14,129 that as you go down the main sequence 279 00:11:18,949 --> 00:11:17,220 two smaller stars this is a case star if 280 00:11:21,619 --> 00:11:18,959 you go to even smaller stars if you go 281 00:11:23,689 --> 00:11:21,629 to M dwarf stars this this becomes even 282 00:11:26,540 --> 00:11:23,699 more favorable situation in the infrared 283 00:11:30,889 --> 00:11:26,550 and the reason is that this this term 284 00:11:33,100 --> 00:11:30,899 this this area ratio was the dominant 285 00:11:34,910 --> 00:11:33,110 term whereas this term is the 286 00:11:36,920 --> 00:11:34,920 temperatures of the planets and the 287 00:11:38,689 --> 00:11:36,930 stars don't vary all that much the 288 00:11:40,460 --> 00:11:38,699 temperature of the planet is the 289 00:11:42,949 --> 00:11:40,470 temperature of the star in thermal 290 00:11:45,889 --> 00:11:42,959 equilibrium times the angular diameter 291 00:11:48,199 --> 00:11:45,899 of the star as seen from the planet and 292 00:11:50,780 --> 00:11:48,209 the square root of that so this varies 293 00:11:53,840 --> 00:11:50,790 relatively mildly where is this term 294 00:11:55,610 --> 00:11:53,850 very predominantly and as you go to 295 00:11:58,280 --> 00:11:55,620 smaller stars this becomes more 296 00:12:00,199 --> 00:11:58,290 favorable so lower main sequence stars 297 00:12:02,030 --> 00:12:00,209 are very good for for high 298 00:12:05,780 --> 00:12:02,040 signal-to-noise planet detection using 299 00:12:07,129 --> 00:12:05,790 secondary eclipses so I think by this 300 00:12:12,499 --> 00:12:07,139 I've convinced you that we're seeing 301 00:12:14,629 --> 00:12:12,509 photons from extrasolar giant planets so 302 00:12:17,240 --> 00:12:14,639 now I want to talk about dissecting 303 00:12:19,790 --> 00:12:17,250 those photons and and looking at their 304 00:12:22,610 --> 00:12:19,800 at their spectrum the previous 305 00:12:26,119 --> 00:12:22,620 detections were photometric and now when 306 00:12:28,850 --> 00:12:26,129 we begin dividing the photons into 307 00:12:30,679 --> 00:12:28,860 smaller wavelength bins it's a more 308 00:12:33,079 --> 00:12:30,689 challenging problem but nevertheless 309 00:12:36,199 --> 00:12:33,089 there are two transiting planets HD 310 00:12:39,590 --> 00:12:36,209 189733 that I just showed you an hd2 a 311 00:12:41,420 --> 00:12:39,600 945 a which are bright enough that we 312 00:12:43,369 --> 00:12:41,430 can measure their spectra using the 313 00:12:46,730 --> 00:12:43,379 relatively modest aperture Spitzer's an 314 00:12:48,740 --> 00:12:46,740 85 centimeter telescope and both of 315 00:12:52,189 --> 00:12:48,750 those detection zwarst earlier this year 316 00:12:54,860 --> 00:12:52,199 our detection was in nature there's a 317 00:12:58,160 --> 00:12:54,870 group headed by Karl grill mayor at the 318 00:13:00,350 --> 00:12:58,170 ipaq at the Spitzer Science Center and 319 00:13:03,319 --> 00:13:00,360 Dave Charbonneau collaborated with him 320 00:13:04,819 --> 00:13:03,329 who detected HD 189733 and I'll show you 321 00:13:05,540 --> 00:13:04,829 their results also there's another 322 00:13:07,970 --> 00:13:05,550 analysis 323 00:13:10,190 --> 00:13:07,980 done by mark Swain who reanalyzed our 324 00:13:13,759 --> 00:13:10,200 data and i won't i won't discuss marks 325 00:13:17,870 --> 00:13:13,769 result unless someone asks here's the 326 00:13:19,579 --> 00:13:17,880 principle that we use and i emphasize 327 00:13:20,900 --> 00:13:19,589 this is the principle this is not what 328 00:13:23,360 --> 00:13:20,910 we actually do but this is the 329 00:13:25,550 --> 00:13:23,370 conceptual basis of what we do before 330 00:13:28,190 --> 00:13:25,560 the secondary Eclipse we use spitzer to 331 00:13:30,980 --> 00:13:28,200 measure the combined light the spectrum 332 00:13:32,780 --> 00:13:30,990 of the star plus planet now we know when 333 00:13:34,850 --> 00:13:32,790 the planet is going to be in Eclipse so 334 00:13:38,060 --> 00:13:34,860 during that time we observe the stellar 335 00:13:40,069 --> 00:13:38,070 spectrum Spitzer's in a very stable 336 00:13:44,000 --> 00:13:40,079 heliocentric orbit the instrument 337 00:13:46,550 --> 00:13:44,010 doesn't change very very much and so we 338 00:13:49,660 --> 00:13:46,560 can subtract the two and drive the 339 00:13:53,000 --> 00:13:49,670 planet spectrum so it's a very simple a 340 00:13:54,410 --> 00:13:53,010 simple idea we'd practice it gets more 341 00:13:57,740 --> 00:13:54,420 complicated and I'm going to show you 342 00:14:00,290 --> 00:13:57,750 some of the actual you know nitty gritty 343 00:14:02,930 --> 00:14:00,300 dirty laundry of how we do this analysis 344 00:14:05,300 --> 00:14:02,940 here's an actual here's an actual 345 00:14:07,100 --> 00:14:05,310 imaging frame from Spitzer and I'm 346 00:14:08,990 --> 00:14:07,110 tempted to say this is raw data that the 347 00:14:11,510 --> 00:14:09,000 Spitzer people would correct me this is 348 00:14:13,850 --> 00:14:11,520 what they call BCD or basic calibrated 349 00:14:16,910 --> 00:14:13,860 data it's an image of the spectrum of 350 00:14:19,400 --> 00:14:16,920 hd2 0945 8v here's the eight we observed 351 00:14:22,430 --> 00:14:19,410 between a rap about seven and a half 352 00:14:24,050 --> 00:14:22,440 microns to about 14 microns here this 353 00:14:27,400 --> 00:14:24,060 street this bright streak is the 354 00:14:30,260 --> 00:14:27,410 spectrum of the star plus the planet and 355 00:14:32,630 --> 00:14:30,270 you can see all you know bad pixels on 356 00:14:34,340 --> 00:14:32,640 the array the hot pixels over at the 357 00:14:36,319 --> 00:14:34,350 right side of the detector array there 358 00:14:39,310 --> 00:14:36,329 what are called the peak up arrays which 359 00:14:42,380 --> 00:14:39,320 are just regions which which receive 360 00:14:45,110 --> 00:14:42,390 which are not dispersed so the star 361 00:14:47,389 --> 00:14:45,120 first the telescope places the star here 362 00:14:49,430 --> 00:14:47,399 to center it and then offsets it to the 363 00:14:50,930 --> 00:14:49,440 slid to the spectrograph we actually can 364 00:14:52,670 --> 00:14:50,940 use the peak up arrays for various 365 00:14:55,220 --> 00:14:52,680 things they're very bright here because 366 00:14:56,660 --> 00:14:55,230 they're not seeing this first light 367 00:14:59,510 --> 00:14:56,670 they're looking at the at the background 368 00:15:02,210 --> 00:14:59,520 radiation from our zodiacal background 369 00:15:04,880 --> 00:15:02,220 directly but we actually observe two 370 00:15:07,689 --> 00:15:04,890 full eclipses of this planet each 371 00:15:10,250 --> 00:15:07,699 eclipse was a six-hour sequence of 372 00:15:12,860 --> 00:15:10,260 observation each of one is let each one 373 00:15:14,730 --> 00:15:12,870 of which is like this these were made in 374 00:15:16,960 --> 00:15:14,740 july 2005 375 00:15:18,850 --> 00:15:16,970 shortly after we did the photometric 376 00:15:20,850 --> 00:15:18,860 protections we immediately jumped on 377 00:15:23,200 --> 00:15:20,860 this spectroscopic technique our 378 00:15:25,360 --> 00:15:23,210 observations are basically 60-second 379 00:15:27,820 --> 00:15:25,370 exposures there are two hundred and 380 00:15:29,500 --> 00:15:27,830 eighty of them her eclipse and we have 381 00:15:32,140 --> 00:15:29,510 two eclipses so we have a total of five 382 00:15:34,780 --> 00:15:32,150 hundred and sixty data frames that look 383 00:15:36,400 --> 00:15:34,790 like this there is a telescope nod at 384 00:15:38,650 --> 00:15:36,410 the center of the eclipse we move the 385 00:15:40,690 --> 00:15:38,660 spectrum so that we can subtract the 386 00:15:43,630 --> 00:15:40,700 background and what's under the spectrum 387 00:15:45,690 --> 00:15:43,640 or more accurately the spectral 388 00:15:48,220 --> 00:15:45,700 resolving power is about a hundred and 389 00:15:50,650 --> 00:15:48,230 that's just sort of barely what a 390 00:15:52,480 --> 00:15:50,660 spectroscopy would call a spectrum but 391 00:15:54,160 --> 00:15:52,490 it is a legitimate that's a high enough 392 00:15:56,530 --> 00:15:54,170 spectral resolution it's not photometry 393 00:15:59,410 --> 00:15:56,540 it's spectroscopy now we can immediately 394 00:16:02,350 --> 00:15:59,420 estimate what signal-to-noise we expect 395 00:16:04,480 --> 00:16:02,360 on the planet in combined light the 396 00:16:07,330 --> 00:16:04,490 signal to noise on this spectrum Star 397 00:16:10,860 --> 00:16:07,340 Plus planet is about 100 in a given 398 00:16:16,300 --> 00:16:10,870 pixel a pixel meaning a given wavelength 399 00:16:18,760 --> 00:16:16,310 / spectrum since we have 560 spectrum 400 00:16:20,830 --> 00:16:18,770 about half of which are in Eclipse and 401 00:16:22,390 --> 00:16:20,840 half of which are out of Eclipse to 402 00:16:23,530 --> 00:16:22,400 detect the spectrum of the planet we 403 00:16:26,200 --> 00:16:23,540 have to basically at each wavelength 404 00:16:28,870 --> 00:16:26,210 define two levels and in Eclipse level 405 00:16:30,940 --> 00:16:28,880 and an out of Eclipse level each level 406 00:16:33,430 --> 00:16:30,950 is defined by there are five hundred and 407 00:16:35,590 --> 00:16:33,440 sixty total spectra each level is 408 00:16:37,420 --> 00:16:35,600 defined by half of those are about 280 409 00:16:39,220 --> 00:16:37,430 so the signal Foyle is on each level 410 00:16:41,890 --> 00:16:39,230 improves by a hundred times the square 411 00:16:44,650 --> 00:16:41,900 root of 280 but then since we have to 412 00:16:46,600 --> 00:16:44,660 compare two independent levels we lose 413 00:16:50,020 --> 00:16:46,610 the square root of two and if you work 414 00:16:52,150 --> 00:16:50,030 that out in our combined data at each at 415 00:16:54,250 --> 00:16:52,160 each wavelength we would have a signal 416 00:16:57,490 --> 00:16:54,260 to noise on the combined system of about 417 00:16:59,350 --> 00:16:57,500 1,200 on Star Plus planet the planet is 418 00:17:01,840 --> 00:16:59,360 about three-tenths of a percent of this 419 00:17:03,550 --> 00:17:01,850 total therefore our signal to noise on 420 00:17:07,300 --> 00:17:03,560 the planet at a given wavelength will be 421 00:17:09,310 --> 00:17:07,310 about 4 which is low signal-to-noise we 422 00:17:14,829 --> 00:17:09,320 are limited by the number of photons 423 00:17:20,480 --> 00:17:18,559 now in order to get to that photon limit 424 00:17:23,319 --> 00:17:20,490 we have to deal with some some 425 00:17:25,549 --> 00:17:23,329 instrument and telescope on 426 00:17:27,769 --> 00:17:25,559 uncertainties some systemic effects 427 00:17:29,899 --> 00:17:27,779 there are two principal systemic effects 428 00:17:34,460 --> 00:17:29,909 first of all there's something we call 429 00:17:37,759 --> 00:17:34,470 the ramp and it what happens is when you 430 00:17:39,350 --> 00:17:37,769 first put light on on the specter on the 431 00:17:41,299 --> 00:17:39,360 detector in the infrared spectrograph 432 00:17:43,700 --> 00:17:41,309 but also on some of the other Spitzer 433 00:17:46,269 --> 00:17:43,710 detectors see the same effect the 434 00:17:48,470 --> 00:17:46,279 intensity that comes out of the detector 435 00:17:51,049 --> 00:17:48,480 even though the source isn't changing 436 00:17:53,570 --> 00:17:51,059 the intensity gradually ramps up with 437 00:17:57,049 --> 00:17:53,580 time that asymptotically approaches some 438 00:17:59,840 --> 00:17:57,059 constant value over many hours and this 439 00:18:02,330 --> 00:17:59,850 is recently the physical cause of this 440 00:18:04,399 --> 00:18:02,340 as has recently been identified by the 441 00:18:06,230 --> 00:18:04,409 effect by the Iraq group at the Center 442 00:18:07,850 --> 00:18:06,240 for Astrophysics and they believe this 443 00:18:10,399 --> 00:18:07,860 is due to charge trapping in the 444 00:18:13,039 --> 00:18:10,409 detector material the first photons that 445 00:18:14,899 --> 00:18:13,049 come in don't produce electrons instead 446 00:18:17,570 --> 00:18:14,909 the electrons that get freed from the 447 00:18:19,490 --> 00:18:17,580 detector get trapped by ionized 448 00:18:21,320 --> 00:18:19,500 impurities in the detector and you don't 449 00:18:23,240 --> 00:18:21,330 really reach full sensitivity until you 450 00:18:25,730 --> 00:18:23,250 fill all those charged traps and that 451 00:18:27,230 --> 00:18:25,740 makes for this ramping up baseline so we 452 00:18:29,659 --> 00:18:27,240 have to we have to deal with this we 453 00:18:31,310 --> 00:18:29,669 have to subtract this from our data the 454 00:18:33,169 --> 00:18:31,320 other effect is there's a telescope 455 00:18:36,169 --> 00:18:33,179 oscillation and we're taking spectrum 456 00:18:39,710 --> 00:18:36,179 the the Spitzer telescope is is 457 00:18:42,950 --> 00:18:39,720 wonderful but it has a point 05 arc 458 00:18:45,019 --> 00:18:42,960 second oscillation in position with it 459 00:18:47,240 --> 00:18:45,029 with a period of about one hour and 460 00:18:49,310 --> 00:18:47,250 because the telescope is oscillating 461 00:18:51,350 --> 00:18:49,320 that means the slit that the star is 462 00:18:53,560 --> 00:18:51,360 moving on the slit and the amount of 463 00:18:56,450 --> 00:18:53,570 light that comes through our slit is 464 00:18:59,570 --> 00:18:56,460 oscillating by several percent do to 465 00:19:01,759 --> 00:18:59,580 that telescope oscillation well you can 466 00:19:03,769 --> 00:19:01,769 imagine if we didn't subtract this we're 467 00:19:05,720 --> 00:19:03,779 looking for the spectra of a planet 468 00:19:06,860 --> 00:19:05,730 which is a few tenths of a percent this 469 00:19:08,180 --> 00:19:06,870 is an order of magnitude this 470 00:19:10,899 --> 00:19:08,190 oscillation is an order of magnitude 471 00:19:14,960 --> 00:19:10,909 larger than the planet signal 472 00:19:16,879 --> 00:19:14,970 fortunately that the where the telescope 473 00:19:20,210 --> 00:19:16,889 is pointed is not a function of 474 00:19:22,639 --> 00:19:20,220 wavelength it's so the phase of this 475 00:19:25,039 --> 00:19:22,649 oscillation is absolutely independent of 476 00:19:25,830 --> 00:19:25,049 wavelength its amplitude is slightly 477 00:19:28,950 --> 00:19:25,840 dependent 478 00:19:32,460 --> 00:19:28,960 wavelength but we can nevertheless 479 00:19:36,269 --> 00:19:32,470 subtract it if you that we do see the 480 00:19:39,419 --> 00:19:36,279 planet at all I've simply summed all of 481 00:19:40,860 --> 00:19:39,429 the wavelengths over time and you can 482 00:19:43,049 --> 00:19:40,870 see here is that here is the total 483 00:19:44,610 --> 00:19:43,059 intensity in our spectra as a function 484 00:19:47,519 --> 00:19:44,620 and this is both eclipses averaged 485 00:19:49,980 --> 00:19:47,529 together and you can see the ramp see 486 00:19:52,350 --> 00:19:49,990 how it rises up and then approaches some 487 00:19:54,480 --> 00:19:52,360 asymptotic value this these dotted lines 488 00:19:56,430 --> 00:19:54,490 are where ingress and egress of the 489 00:19:58,980 --> 00:19:56,440 planets eclipses occur you can see there 490 00:20:00,659 --> 00:19:58,990 is a decrease at this time and that's 491 00:20:03,630 --> 00:20:00,669 the eclipse of the planet you can also 492 00:20:05,490 --> 00:20:03,640 see the residual oscillation since we've 493 00:20:07,500 --> 00:20:05,500 averaged two independent eclipses the 494 00:20:09,060 --> 00:20:07,510 oscillations not a big because the 495 00:20:10,740 --> 00:20:09,070 oscillation had a different phase 496 00:20:12,510 --> 00:20:10,750 because those in those eclipses were 497 00:20:14,850 --> 00:20:12,520 measured a week apart the telescope 498 00:20:16,680 --> 00:20:14,860 repointed to completely reset the phase 499 00:20:19,470 --> 00:20:16,690 of the oscillation so it tends to 500 00:20:21,180 --> 00:20:19,480 average out here but we'd what we do is 501 00:20:23,250 --> 00:20:21,190 we independently subtract it in each 502 00:20:24,779 --> 00:20:23,260 eclipse but I average these together 503 00:20:29,060 --> 00:20:24,789 because I have to show you first of all 504 00:20:32,159 --> 00:20:29,070 that we do see the eclipse of the planet 505 00:20:35,639 --> 00:20:32,169 now here's how we actually get the 506 00:20:36,960 --> 00:20:35,649 spectra it's a multi-step process the 507 00:20:39,149 --> 00:20:36,970 first thing we have to do is we have 508 00:20:41,220 --> 00:20:39,159 images of the spectra we have to 509 00:20:43,320 --> 00:20:41,230 actually get spectra that means we have 510 00:20:45,269 --> 00:20:43,330 to remove all those bad pixels we have 511 00:20:47,610 --> 00:20:45,279 to subtract the background and we have 512 00:20:51,240 --> 00:20:47,620 to sum over a special window and we 513 00:20:52,500 --> 00:20:51,250 actually spent about three months doing 514 00:20:54,840 --> 00:20:52,510 this and we did it three different ways 515 00:20:56,820 --> 00:20:54,850 we had to sort of custom ways to do it 516 00:20:58,500 --> 00:20:56,830 where we would look at the actual time 517 00:21:00,930 --> 00:20:58,510 dependence of the intensity in each 518 00:21:03,630 --> 00:21:00,940 pixel and we would throw away bad pixels 519 00:21:06,019 --> 00:21:03,640 we literally every pixel in the detector 520 00:21:09,090 --> 00:21:06,029 had to become our personal friend and 521 00:21:11,010 --> 00:21:09,100 and we had to two members of our team do 522 00:21:13,200 --> 00:21:11,020 this with their own custom procedures 523 00:21:15,450 --> 00:21:13,210 then there are their standard software 524 00:21:17,669 --> 00:21:15,460 to do it by the Spitzer method and we 525 00:21:18,990 --> 00:21:17,679 did that too and we then we took all 526 00:21:21,269 --> 00:21:19,000 those results and put them through our 527 00:21:24,240 --> 00:21:21,279 entire analysis to get the spectrum of 528 00:21:26,850 --> 00:21:24,250 the planet and we can't just pick the 529 00:21:29,070 --> 00:21:26,860 method we like the best we have to have 530 00:21:31,049 --> 00:21:29,080 some objective way of doing it and the 531 00:21:33,029 --> 00:21:31,059 objective method that we decided to pick 532 00:21:36,050 --> 00:21:33,039 the best method who is we would compare 533 00:21:38,630 --> 00:21:36,060 the planets spectrum on the two eclipses 534 00:21:40,220 --> 00:21:38,640 and we would do a chi-squared of the 535 00:21:41,870 --> 00:21:40,230 difference between the two eclipses from 536 00:21:44,180 --> 00:21:41,880 the planet spectrum and the method that 537 00:21:47,030 --> 00:21:44,190 produced the minimum chi-squared is a 538 00:21:49,250 --> 00:21:47,040 method we used it turned out to be one 539 00:21:51,850 --> 00:21:49,260 of our custom extraction methods which 540 00:21:53,630 --> 00:21:51,860 isn't surprising because the 541 00:21:56,180 --> 00:21:53,640 characteristics of the detector do 542 00:21:58,550 --> 00:21:56,190 change with time so if we use something 543 00:22:01,330 --> 00:21:58,560 care to customize for our own data 544 00:22:05,030 --> 00:22:01,340 that's going to produce the best results 545 00:22:07,280 --> 00:22:05,040 then what we do is we really don't we 546 00:22:10,630 --> 00:22:07,290 really don't subtract spectra we really 547 00:22:13,790 --> 00:22:10,640 treat these spectra as a series of 548 00:22:16,160 --> 00:22:13,800 photometry measurements in other words 549 00:22:18,920 --> 00:22:16,170 we normalize the intensities at each 550 00:22:21,590 --> 00:22:18,930 wavelength to into an average value of 551 00:22:24,770 --> 00:22:21,600 unity another way of saying that is we 552 00:22:27,740 --> 00:22:24,780 divide every spectrum by the average 553 00:22:29,390 --> 00:22:27,750 spectrum of the star and that normalizes 554 00:22:31,840 --> 00:22:29,400 the intensity of each wavelength to 555 00:22:34,400 --> 00:22:31,850 unity and it puts our final results in 556 00:22:38,050 --> 00:22:34,410 in what's called contrast units the 557 00:22:42,980 --> 00:22:38,060 ratio of the planet to the star we then 558 00:22:45,290 --> 00:22:42,990 give up the average eclipse and in other 559 00:22:49,220 --> 00:22:45,300 words we subtract the average time 560 00:22:53,330 --> 00:22:49,230 series and when we do that this average 561 00:22:55,220 --> 00:22:53,340 eclipse goes away we subtract it out of 562 00:22:57,290 --> 00:22:55,230 the data what we're left with are the 563 00:23:00,770 --> 00:22:57,300 differences in that eclipse from one 564 00:23:03,290 --> 00:23:00,780 wavelength to the next so if the planets 565 00:23:05,540 --> 00:23:03,300 spectrum were constant over our 566 00:23:08,630 --> 00:23:05,550 wavelength range we would see nothing 567 00:23:10,610 --> 00:23:08,640 but because we do see the average 568 00:23:12,230 --> 00:23:10,620 eclipse we would know that the point 569 00:23:13,700 --> 00:23:12,240 that we are detecting the flux from the 570 00:23:17,030 --> 00:23:13,710 planet but that it's constant with 571 00:23:18,740 --> 00:23:17,040 wavelength now the advantage of 572 00:23:21,650 --> 00:23:18,750 subtracting that average time series is 573 00:23:24,200 --> 00:23:21,660 it removes most of the systematics it 574 00:23:26,480 --> 00:23:24,210 removes the intensity ramp apart from a 575 00:23:28,070 --> 00:23:26,490 linear term and it and it very 576 00:23:33,260 --> 00:23:28,080 effectively removes the telescope 577 00:23:35,840 --> 00:23:33,270 oscillation we then to each time series 578 00:23:37,040 --> 00:23:35,850 we fit in eclipse curve and we do allow 579 00:23:39,260 --> 00:23:37,050 that eclipse curve to have a linear 580 00:23:42,620 --> 00:23:39,270 baseline a sloping up or down at each 581 00:23:44,360 --> 00:23:42,630 wavelength and we derive the planet 582 00:23:46,460 --> 00:23:44,370 spectrum from the amplitude of those 583 00:23:47,560 --> 00:23:46,470 eclipse curve we those are differential 584 00:23:49,810 --> 00:23:47,570 eclipses 585 00:23:51,660 --> 00:23:49,820 let's say that the planet is brighter at 586 00:23:54,310 --> 00:23:51,670 one wavelength than it is at another 587 00:23:56,200 --> 00:23:54,320 then when we subtract the average time 588 00:23:58,570 --> 00:23:56,210 series we will have a little bit of 589 00:24:00,010 --> 00:23:58,580 residual extra eclipse at that 590 00:24:04,480 --> 00:24:00,020 wavelength with the planet is brighter 591 00:24:06,490 --> 00:24:04,490 and thus will and and our Eclipse our 592 00:24:09,250 --> 00:24:06,500 method of fitting the Eclipse curve will 593 00:24:11,650 --> 00:24:09,260 detect that little extra Eclipse and we 594 00:24:14,590 --> 00:24:11,660 would take that extra Eclipse depth and 595 00:24:16,660 --> 00:24:14,600 translate it into an extra depth for an 596 00:24:18,340 --> 00:24:16,670 extra brightness for the planet we then 597 00:24:21,730 --> 00:24:18,350 go through and reject any wavelengths 598 00:24:23,680 --> 00:24:21,740 that have for fits if if this fit to the 599 00:24:26,920 --> 00:24:23,690 Eclipse curve doesn't show the correct 600 00:24:28,840 --> 00:24:26,930 ingress time the correct egress time the 601 00:24:30,490 --> 00:24:28,850 correct central phase we throw it away 602 00:24:33,730 --> 00:24:30,500 and about five percent of the 603 00:24:35,260 --> 00:24:33,740 wavelengths have residual you know ramp 604 00:24:37,600 --> 00:24:35,270 effect so the systematics are not 605 00:24:39,600 --> 00:24:37,610 effectively removed or those pixels are 606 00:24:43,230 --> 00:24:39,610 very poorly behaved we throw those away 607 00:24:47,710 --> 00:24:43,240 we then compared to independent eclipses 608 00:24:50,140 --> 00:24:47,720 and this is an example of a differential 609 00:24:54,790 --> 00:24:50,150 eclipse this is our the point where we 610 00:24:56,050 --> 00:24:54,800 find the brightest peak in the planet 611 00:24:59,890 --> 00:24:56,060 spectrum and this is a differential 612 00:25:02,740 --> 00:24:59,900 eclipse it's centered at phase 0.5 where 613 00:25:06,610 --> 00:25:02,750 it's supposed to whisk this dash line is 614 00:25:09,130 --> 00:25:06,620 is a theoretical eclipse curve that has 615 00:25:10,900 --> 00:25:09,140 the correct ingress time and egress time 616 00:25:13,000 --> 00:25:10,910 in the correct central phase we don't 617 00:25:14,650 --> 00:25:13,010 ship this back and forth all we do is we 618 00:25:19,630 --> 00:25:14,660 fit its amplitude we stretch it up and 619 00:25:23,350 --> 00:25:19,640 down here are the aggregate results for 620 00:25:26,830 --> 00:25:23,360 two eclipses of this planet 0 is right 621 00:25:28,300 --> 00:25:26,840 here so you can see we're detecting at 622 00:25:30,790 --> 00:25:28,310 the longest wavelength we basically 623 00:25:33,250 --> 00:25:30,800 don't we barely detect the planets 624 00:25:36,310 --> 00:25:33,260 continuum we know we can't really say 625 00:25:39,250 --> 00:25:36,320 anything about any structure there are 626 00:25:41,710 --> 00:25:39,260 there are two features in this spectrum 627 00:25:43,360 --> 00:25:41,720 that we regard as real in the sense that 628 00:25:45,700 --> 00:25:43,370 although there are only three or four 629 00:25:47,620 --> 00:25:45,710 Sigma detection and we would like not to 630 00:25:49,870 --> 00:25:47,630 be in the three or four Sigma domain but 631 00:25:52,030 --> 00:25:49,880 that's the amount of photons we have but 632 00:25:54,310 --> 00:25:52,040 these features have to have passed at 633 00:25:56,110 --> 00:25:54,320 least all of our tests and actually 634 00:25:57,580 --> 00:25:56,120 there's another of there's another 635 00:25:58,899 --> 00:25:57,590 aspect that I'll talk about that it's 636 00:26:01,659 --> 00:25:58,909 even more robust 637 00:26:03,460 --> 00:26:01,669 but the two peaks near about nine and a 638 00:26:05,499 --> 00:26:03,470 half microns we see this kind of broad 639 00:26:07,930 --> 00:26:05,509 peak in the spectrum and you notice that 640 00:26:09,489 --> 00:26:07,940 the blue points and the red points both 641 00:26:13,239 --> 00:26:09,499 show it these are two independent 642 00:26:14,710 --> 00:26:13,249 eclipses this is the stretching 643 00:26:17,440 --> 00:26:14,720 resonance between silicon and oxygen 644 00:26:20,049 --> 00:26:17,450 there are many many Astrophysical 645 00:26:22,690 --> 00:26:20,059 sources not extrasolar planets but no 646 00:26:24,669 --> 00:26:22,700 proto planetary objects and disks and 647 00:26:28,749 --> 00:26:24,679 support which shall silicate emission at 648 00:26:31,389 --> 00:26:28,759 this wavelength and so this is a very 649 00:26:33,849 --> 00:26:31,399 familiar feature to astronomers we think 650 00:26:36,330 --> 00:26:33,859 that this is indicative of silicate 651 00:26:38,859 --> 00:26:36,340 clouds in the atmosphere of the planet 652 00:26:42,430 --> 00:26:38,869 there is another sharper feature in the 653 00:26:43,989 --> 00:26:42,440 spectrum that that passes our tests to 654 00:26:47,049 --> 00:26:43,999 be real in fact that differential 655 00:26:48,729 --> 00:26:47,059 eclipse that I showed you this eclipse 656 00:26:53,950 --> 00:26:48,739 is made at the at the peak of that 657 00:26:55,899 --> 00:26:53,960 narrow feature and initially we this is 658 00:26:57,430 --> 00:26:55,909 at seven point seven eight microns and 659 00:26:59,589 --> 00:26:57,440 initially we thought this was the q 660 00:27:01,119 --> 00:26:59,599 branch of methane that was showing an 661 00:27:03,999 --> 00:27:01,129 emission because the planet had a hot 662 00:27:06,369 --> 00:27:04,009 stratosphere and we've since been able 663 00:27:07,960 --> 00:27:06,379 to determine that its wavelength is not 664 00:27:09,519 --> 00:27:07,970 coincident with the q branch the q 665 00:27:11,799 --> 00:27:09,529 branch is at like seven point six five 666 00:27:13,389 --> 00:27:11,809 microns and it shifts the longer 667 00:27:16,330 --> 00:27:13,399 wavelengths at hotter temperatures but 668 00:27:17,919 --> 00:27:16,340 not enough and we very carefully checked 669 00:27:19,989 --> 00:27:17,929 the wavelength calibration and we talked 670 00:27:22,269 --> 00:27:19,999 to the folks at the Spitzer Science 671 00:27:24,609 --> 00:27:22,279 Center and they tell us that there is 672 00:27:26,440 --> 00:27:24,619 that the wavelength calibration of 673 00:27:28,419 --> 00:27:26,450 Spitzer it's out of the question for it 674 00:27:31,869 --> 00:27:28,429 to be off by it would have to be off by 675 00:27:34,690 --> 00:27:31,879 like two pixels so this teacher can't be 676 00:27:37,269 --> 00:27:34,700 messing about the only thing that we 677 00:27:38,950 --> 00:27:37,279 can't eliminate we don't know what it is 678 00:27:41,649 --> 00:27:38,960 but we can't eliminate that it's due to 679 00:27:44,169 --> 00:27:41,659 some feature due to the there's a 680 00:27:47,820 --> 00:27:44,179 carbon-carbon stretching residence near 681 00:27:50,560 --> 00:27:47,830 near about 7.8 microns and in most 682 00:27:54,339 --> 00:27:50,570 Astrophysical sources this is seen in 683 00:27:59,409 --> 00:27:54,349 polycyclic aromatic hydrocarbons pah the 684 00:28:01,029 --> 00:27:59,419 pah has another emission feature at 11.3 685 00:28:03,729 --> 00:28:01,039 microns which we don't see anything 686 00:28:06,580 --> 00:28:03,739 there but the reason we can't eliminate 687 00:28:08,499 --> 00:28:06,590 this as being pah is that is that that 688 00:28:09,640 --> 00:28:08,509 other resonance weakens in high 689 00:28:12,400 --> 00:28:09,650 radiation environment 690 00:28:14,830 --> 00:28:12,410 it's and we're in point oh five AU from 691 00:28:17,620 --> 00:28:14,840 a relatively hot star certainly a high 692 00:28:20,050 --> 00:28:17,630 radiation environment so it's possible 693 00:28:22,420 --> 00:28:20,060 that we're only seeing this CC stretch 694 00:28:24,940 --> 00:28:22,430 there is another carbon-carbon stretch 695 00:28:27,070 --> 00:28:24,950 at 6.2 microns and we have proposed that 696 00:28:29,530 --> 00:28:27,080 Spitzer observed at shorter wavelengths 697 00:28:31,630 --> 00:28:29,540 so we won't really conclude anything 698 00:28:33,010 --> 00:28:31,640 about this until we can look at that but 699 00:28:35,740 --> 00:28:33,020 that other at the wavelength of that 700 00:28:36,790 --> 00:28:35,750 other resonance now this lower panel 701 00:28:38,910 --> 00:28:36,800 shows you the average of these two 702 00:28:43,480 --> 00:28:38,920 spectra the red line is just a blackbody 703 00:28:45,520 --> 00:28:43,490 the blue line is a fancy model our team 704 00:28:48,460 --> 00:28:45,530 fear of Sara Seager this is one of her 705 00:28:50,440 --> 00:28:48,470 hot models models by Adam Burroughs 706 00:28:52,330 --> 00:28:50,450 showed the same general behavior what 707 00:28:54,070 --> 00:28:52,340 what these hot extrasolar planets are 708 00:28:55,840 --> 00:28:54,080 supposed to have their supposed to be a 709 00:28:59,440 --> 00:28:55,850 fall off at Short wavelengths due to 710 00:29:01,570 --> 00:28:59,450 water absorption here the blue line is 711 00:29:03,340 --> 00:29:01,580 supposed to decrease downwards you can 712 00:29:05,800 --> 00:29:03,350 see that the points certainly don't do 713 00:29:07,360 --> 00:29:05,810 that if anything they tend to trim the 714 00:29:11,080 --> 00:29:07,370 other direction and then jump up in this 715 00:29:16,540 --> 00:29:11,090 silicate residence another aspect of 716 00:29:22,660 --> 00:29:16,550 this we bend it in these broad bins and 717 00:29:24,760 --> 00:29:22,670 you can see this is the really coarse 718 00:29:28,270 --> 00:29:24,770 bidding of the spectrum if we in the two 719 00:29:30,640 --> 00:29:28,280 points here are if we if we include or 720 00:29:33,610 --> 00:29:30,650 not include this narrow feature in the 721 00:29:36,280 --> 00:29:33,620 bin we use this to convince ourselves 722 00:29:38,320 --> 00:29:36,290 that this peak due to the near the SiO 723 00:29:40,120 --> 00:29:38,330 stretching resonance is real in the data 724 00:29:42,250 --> 00:29:40,130 actually we only claim that this 725 00:29:45,280 --> 00:29:42,260 difference is is real this difference 726 00:29:48,070 --> 00:29:45,290 here is a little it that's a little not 727 00:29:49,630 --> 00:29:48,080 so significant so it's in a sense it's 728 00:29:50,890 --> 00:29:49,640 only half significant it's only 729 00:29:55,110 --> 00:29:50,900 significant if you come from the 730 00:30:00,160 --> 00:29:58,180 to show you a similar result this is the 731 00:30:03,640 --> 00:30:00,170 grill mayor a towel result their their 732 00:30:06,070 --> 00:30:03,650 spectrum of HD 189733 B they have better 733 00:30:07,990 --> 00:30:06,080 signal to noise because their planet has 734 00:30:10,720 --> 00:30:08,000 a higher contrast to the star the star 735 00:30:12,490 --> 00:30:10,730 is smaller the planet too stark contrast 736 00:30:15,700 --> 00:30:12,500 ratio is better by about a factor of two 737 00:30:18,220 --> 00:30:15,710 or three here's their data points and a 738 00:30:22,210 --> 00:30:18,230 similar model I believe this is an atom 739 00:30:24,280 --> 00:30:22,220 burrows model again water absorption 740 00:30:25,900 --> 00:30:24,290 would cause this model to trend 741 00:30:28,090 --> 00:30:25,910 downwards you can see the data don't do 742 00:30:30,340 --> 00:30:28,100 that they're flat we know that the 743 00:30:32,470 --> 00:30:30,350 specter of these planets will turn down 744 00:30:34,600 --> 00:30:32,480 at some point below beyond this because 745 00:30:37,030 --> 00:30:34,610 the Spitzer photometry at four and a 746 00:30:39,670 --> 00:30:37,040 half microns show a much lower contrast 747 00:30:41,860 --> 00:30:39,680 so somewhere this this spectrum falls 748 00:30:48,130 --> 00:30:41,870 off the cliff towards lower values but 749 00:30:51,820 --> 00:30:48,140 not in this wavelength range so why are 750 00:30:55,090 --> 00:30:51,830 we seeing no water absorption the sort 751 00:30:57,580 --> 00:30:55,100 of the most naive possibility would be 752 00:31:00,190 --> 00:30:57,590 that the planets have no water and 753 00:31:03,130 --> 00:31:00,200 that's we considered that for about 754 00:31:07,060 --> 00:31:03,140 three milliseconds and the reason that 755 00:31:08,710 --> 00:31:07,070 we have to reject that is that it's so 756 00:31:10,450 --> 00:31:08,720 easy to make water it's hard to avoid 757 00:31:13,120 --> 00:31:10,460 wat wat oxygen is the third most 758 00:31:15,340 --> 00:31:13,130 abundant element in the universe and 759 00:31:16,870 --> 00:31:15,350 hydrogen is certainly very molecular 760 00:31:19,120 --> 00:31:16,880 hydrogen is very abundant in these 761 00:31:21,220 --> 00:31:19,130 objects it seems virtually physically 762 00:31:22,870 --> 00:31:21,230 impossible to make these planets for 763 00:31:25,330 --> 00:31:22,880 that significant amounts of water vapor 764 00:31:27,520 --> 00:31:25,340 in their atmosphere so they have to have 765 00:31:29,800 --> 00:31:27,530 some water so the only really two 766 00:31:32,230 --> 00:31:29,810 explanations are that perhaps their 767 00:31:34,200 --> 00:31:32,240 masked by high clouds that would be 768 00:31:37,380 --> 00:31:34,210 consistent with the fact that we see is 769 00:31:40,259 --> 00:31:37,390 we see a silicate emission feature 770 00:31:43,269 --> 00:31:40,269 notice that 771 00:31:46,090 --> 00:31:43,279 real maridel don't see really a silicate 772 00:31:49,239 --> 00:31:46,100 emission feature but the star that this 773 00:31:51,039 --> 00:31:49,249 planet orbits is a que drawer and if 774 00:31:53,440 --> 00:31:51,049 there are features in emission we would 775 00:31:55,659 --> 00:31:53,450 expect them to be excited more when you 776 00:31:57,879 --> 00:31:55,669 orbit a hotter star our planet was a 777 00:32:00,369 --> 00:31:57,889 hundred star so that's consistent so 778 00:32:02,200 --> 00:32:00,379 maybe there are high clouds and maybe 779 00:32:04,149 --> 00:32:02,210 their silicate clouds that are masking 780 00:32:06,099 --> 00:32:04,159 the water absorption below it there's 781 00:32:07,499 --> 00:32:06,109 other evidence for high clouds and the 782 00:32:09,999 --> 00:32:07,509 atmospheres of these hot Jupiters 783 00:32:11,590 --> 00:32:10,009 another possibility is that there's some 784 00:32:14,769 --> 00:32:11,600 perturbation to the temperature gradient 785 00:32:16,989 --> 00:32:14,779 in fact since we see features in 786 00:32:18,969 --> 00:32:16,999 emission if there are thermal features 787 00:32:22,180 --> 00:32:18,979 that would require a reverse temperature 788 00:32:23,889 --> 00:32:22,190 gradient the stratosphere in effect it 789 00:32:27,009 --> 00:32:23,899 takes less than that to mask the high 790 00:32:29,979 --> 00:32:27,019 water this is a this is a result by 791 00:32:32,649 --> 00:32:29,989 Jonathan Kourtney it's a model in which 792 00:32:35,229 --> 00:32:32,659 they incorporate the effects of dynamics 793 00:32:36,759 --> 00:32:35,239 that is strong circulation that is 794 00:32:39,969 --> 00:32:36,769 believed to occur in the hot Jupiter 795 00:32:42,129 --> 00:32:39,979 atmospheres and they show you what that 796 00:32:43,509 --> 00:32:42,139 it fit with that strong circulation the 797 00:32:46,869 --> 00:32:43,519 effect that it has on the emerging 798 00:32:48,399 --> 00:32:46,879 spectrum when the when you observe here 799 00:32:50,440 --> 00:32:48,409 you are looking at the planet here 800 00:32:52,629 --> 00:32:50,450 here's the star when the planet is 801 00:32:54,549 --> 00:32:52,639 behind the star at secondary eclipse or 802 00:32:57,999 --> 00:32:54,559 is opposite from you then you're looking 803 00:33:00,339 --> 00:32:58,009 at the dayside of the planet and the 804 00:33:03,099 --> 00:33:00,349 what Courtney I'll find is that the 805 00:33:05,499 --> 00:33:03,109 temperature profile on the dayside the 806 00:33:07,479 --> 00:33:05,509 effect of the strong circulation is to 807 00:33:10,060 --> 00:33:07,489 is to make the temperature profile much 808 00:33:11,889 --> 00:33:10,070 more isothermal so the spectra on the 809 00:33:15,009 --> 00:33:11,899 dayside are much more like a black body 810 00:33:17,680 --> 00:33:15,019 in this spectral range and this dip 811 00:33:19,629 --> 00:33:17,690 downwards due to water at seven microns 812 00:33:22,029 --> 00:33:19,639 is not nearly so prominent at secondary 813 00:33:24,099 --> 00:33:22,039 eclipse so in other words when we can 814 00:33:26,229 --> 00:33:24,109 measure the spectra that secondary 815 00:33:28,450 --> 00:33:26,239 eclipse that's when the spectra are 816 00:33:30,070 --> 00:33:28,460 least interesting in this respect and 817 00:33:33,489 --> 00:33:30,080 that's unfortunate but that would 818 00:33:35,229 --> 00:33:33,499 explain the and my colleague Sara Seager 819 00:33:37,690 --> 00:33:35,239 also has other ideas on how the 820 00:33:39,519 --> 00:33:37,700 temperature gradient could be perturbed 821 00:33:41,589 --> 00:33:39,529 she thinks that there might be affected 822 00:33:43,509 --> 00:33:41,599 deep stratosphere a temperature reversal 823 00:33:45,999 --> 00:33:43,519 that extends to fairly high pressures 824 00:33:48,129 --> 00:33:46,009 now I can't really articulate the 825 00:33:50,109 --> 00:33:48,139 physics underlying these models in 826 00:33:50,409 --> 00:33:50,119 particular I would like Jonathan if he's 827 00:33:57,820 --> 00:33:50,419 on 828 00:33:59,379 --> 00:33:57,830 profile isothermal but maybe he's not 829 00:34:03,129 --> 00:33:59,389 online maybe one of his collaborators 830 00:34:04,749 --> 00:34:03,139 can explain that to me well what I will 831 00:34:07,299 --> 00:34:04,759 do is I want to try to extend these a 832 00:34:09,099 --> 00:34:07,309 little bit because what we're really 833 00:34:13,359 --> 00:34:09,109 interested in is the spectroscopy of 834 00:34:15,250 --> 00:34:13,369 habitable planets and we can't where we 835 00:34:17,530 --> 00:34:15,260 probably can't do that with Spitzer 836 00:34:19,240 --> 00:34:17,540 there are several sort of super 837 00:34:21,399 --> 00:34:19,250 earth-mass planets that are known to 838 00:34:25,720 --> 00:34:21,409 orbit close to M Dwarfs for example 839 00:34:27,789 --> 00:34:25,730 Gliese 876 D of is a 7.5 earth-mass 840 00:34:31,930 --> 00:34:27,799 planet that orbits close to the M dwarf 841 00:34:34,569 --> 00:34:31,940 lewis at 876 and this system is five 842 00:34:37,149 --> 00:34:34,579 parsecs from Earth and in fact Sara 843 00:34:39,220 --> 00:34:37,159 Seager and I have a room have a 844 00:34:42,490 --> 00:34:39,230 photometric program to try to measure 845 00:34:44,470 --> 00:34:42,500 the phase modulation of the infrared 846 00:34:47,079 --> 00:34:44,480 phase modulation to try to detect the 847 00:34:49,750 --> 00:34:47,089 planet at all photometrically but we 848 00:34:53,980 --> 00:34:49,760 haven't dared to try to measure its 849 00:34:56,020 --> 00:34:53,990 spectrum in order to measure the specter 850 00:34:58,690 --> 00:34:56,030 of these planets we have to go to what I 851 00:35:01,240 --> 00:34:58,700 regard as the as NASA's most important 852 00:35:03,670 --> 00:35:01,250 planet measuring mission even though 853 00:35:05,319 --> 00:35:03,680 it's not build that way and this is this 854 00:35:08,319 --> 00:35:05,329 is a model of the James Webb Space 855 00:35:10,829 --> 00:35:08,329 Telescope if Spitzer can measure the 856 00:35:14,500 --> 00:35:10,839 specter of two bright giant planets 857 00:35:17,170 --> 00:35:14,510 Spitzer's 85 centimeters in aperture 858 00:35:19,559 --> 00:35:17,180 this james webb space telescope is six 859 00:35:22,839 --> 00:35:19,569 point six point five meters in diameter 860 00:35:26,200 --> 00:35:22,849 25 square meter collecting area runs 861 00:35:28,539 --> 00:35:26,210 from point six to 25 microns the this 862 00:35:31,539 --> 00:35:28,549 this telescope will have fantastic 863 00:35:34,780 --> 00:35:31,549 capabilities if it is even half a stable 864 00:35:36,490 --> 00:35:34,790 and it's in you know will be it is not 865 00:35:38,109 --> 00:35:36,500 earth orbiting it's a little groggy in 866 00:35:40,270 --> 00:35:38,119 point where it should be in a thermally 867 00:35:43,299 --> 00:35:40,280 stable environment you can see this 868 00:35:47,319 --> 00:35:43,309 model here this is set up at Goddard and 869 00:35:49,420 --> 00:35:47,329 this these layers of material here are 870 00:35:52,510 --> 00:35:49,430 the sunshield what I like most are these 871 00:35:54,069 --> 00:35:52,520 potted plants that go below it I think 872 00:35:57,160 --> 00:35:54,079 these are the in case there are any more 873 00:35:59,240 --> 00:35:57,170 budget cuts these are the D scope option 874 00:36:04,520 --> 00:35:59,250 NASA will watch these in lieu of the 875 00:36:07,490 --> 00:36:04,530 shield now we've been doing calculations 876 00:36:10,370 --> 00:36:07,500 of what jwst would see if we did our 877 00:36:13,940 --> 00:36:10,380 Spitzer investigation with jwst for a 878 00:36:15,590 --> 00:36:13,950 nearby bright M dwarf we have this IDL 879 00:36:17,720 --> 00:36:15,600 code that will calculate we'll go 880 00:36:20,180 --> 00:36:17,730 through our analysis and calculate the 881 00:36:23,540 --> 00:36:20,190 signal to noise in the planet spectrum 882 00:36:26,090 --> 00:36:23,550 at a given spectral resolving power for 883 00:36:28,460 --> 00:36:26,100 a postulated stellar temperature we've 884 00:36:31,820 --> 00:36:28,470 input here on M dwarf which is point 885 00:36:35,840 --> 00:36:31,830 three stellar radii 3,500 Kelvin we 886 00:36:38,270 --> 00:36:35,850 placed a planet at point O five AU the 887 00:36:40,130 --> 00:36:38,280 nice thing about M Dwarfs is that if the 888 00:36:42,200 --> 00:36:40,140 planet transits if it's in clothes in it 889 00:36:43,850 --> 00:36:42,210 transits that has a hike transit 890 00:36:46,700 --> 00:36:43,860 probability that's also where the 891 00:36:49,190 --> 00:36:46,710 habitable zone is so planets transiting 892 00:36:51,890 --> 00:36:49,200 M Dwarfs may be in the habitable zone 893 00:36:54,890 --> 00:36:51,900 this planet we've made it a super 894 00:36:56,960 --> 00:36:54,900 earth-mass planet to earth radii its 895 00:37:00,110 --> 00:36:56,970 temperature follows from these 896 00:37:02,120 --> 00:37:00,120 parameters and an assumed albedo which 897 00:37:04,550 --> 00:37:02,130 in this case we assume point three its 898 00:37:07,400 --> 00:37:04,560 temperatures 379 Kelvin much cooler than 899 00:37:08,840 --> 00:37:07,410 the planet that I just then the giant 900 00:37:11,960 --> 00:37:08,850 planets I've just been talking about we 901 00:37:14,540 --> 00:37:11,970 placed it 10 parsecs from Earth this is 902 00:37:17,090 --> 00:37:14,550 the signal to noise that jwst would hat 903 00:37:19,730 --> 00:37:17,100 would would obtain in 200 hours of 904 00:37:21,860 --> 00:37:19,740 observation and you may scoff and you 905 00:37:25,130 --> 00:37:21,870 may think we'll we will never get 200 906 00:37:27,560 --> 00:37:25,140 hours on jwst oh yes we will if we can 907 00:37:31,520 --> 00:37:27,570 measure we can measure the spectrum of a 908 00:37:33,560 --> 00:37:31,530 habitable planet orbiting an m-dwarf 200 909 00:37:35,510 --> 00:37:33,570 hours is not out of line compared to 910 00:37:37,580 --> 00:37:35,520 like things like the Hubble Deep Field 911 00:37:40,070 --> 00:37:37,590 and the are what our extra galactic 912 00:37:42,140 --> 00:37:40,080 colleagues do so we think that in fact 913 00:37:44,510 --> 00:37:42,150 that the way to measure spectra of 914 00:37:47,360 --> 00:37:44,520 habitable at least some kinds of 915 00:37:49,460 --> 00:37:47,370 habitable earth-like planets is to look 916 00:37:54,110 --> 00:37:49,470 for planets transiting m dwarfs and to 917 00:37:58,100 --> 00:37:54,120 go after them with jwst so i'm going to 918 00:38:00,290 --> 00:37:58,110 end by concluding there are two seven to 919 00:38:03,020 --> 00:38:00,300 14 micron spectra of hot Jupiters they 920 00:38:05,450 --> 00:38:03,030 show evan evidence both planets show 921 00:38:07,430 --> 00:38:05,460 measured by two independent groups show 922 00:38:08,750 --> 00:38:07,440 evidence for masking of their predicted 923 00:38:10,520 --> 00:38:08,760 water absorption 924 00:38:12,020 --> 00:38:10,530 that there's evidence for silicate 925 00:38:14,540 --> 00:38:12,030 clouds on at least one of those planets 926 00:38:15,890 --> 00:38:14,550 in either case we need some kind of 927 00:38:17,510 --> 00:38:15,900 perturbations to the temperature 928 00:38:20,480 --> 00:38:17,520 gradient as compared to simple 929 00:38:23,840 --> 00:38:20,490 equilibrium models and finally we're 930 00:38:26,480 --> 00:38:23,850 looking forward to jwst taking this to 931 00:38:32,510 --> 00:38:26,490 the spectra of hot Earth's and I will 932 00:38:35,350 --> 00:38:32,520 entertain questions and I'm sure there 933 00:38:37,400 --> 00:38:35,360 are lots of questions for the 934 00:38:41,150 --> 00:38:37,410 sequestering of oxygen to make hear him 935 00:38:46,120 --> 00:38:41,160 I'll repeat the question boy there's one 936 00:38:49,100 --> 00:38:46,130 of Hoover's grab the huggable for the 937 00:38:50,960 --> 00:38:49,110 sequestering in the legend let me just 938 00:38:52,730 --> 00:38:50,970 ask people around the net if they would 939 00:38:54,170 --> 00:38:52,740 raise their hands on webex and will 940 00:38:55,900 --> 00:38:54,180 calling them from here but let's go 941 00:39:01,340 --> 00:38:55,910 ahead with the questions it's got it 942 00:39:03,440 --> 00:39:01,350 okay okay I recall that toured for brown 943 00:39:07,600 --> 00:39:03,450 dwarfs in the late L and early tee 944 00:39:11,240 --> 00:39:07,610 sequence there's a dichotomy of how you 945 00:39:12,860 --> 00:39:11,250 sequester how you the taxonomy of those 946 00:39:14,810 --> 00:39:12,870 things whether how cloudy there are and 947 00:39:16,700 --> 00:39:14,820 so in fact that reinforces the idea of a 948 00:39:19,880 --> 00:39:16,710 cloud of something so loquacious 949 00:39:22,100 --> 00:39:19,890 material over a great hot Jupiter but 950 00:39:23,810 --> 00:39:22,110 also there's limbs evidence for carbon 951 00:39:25,370 --> 00:39:23,820 monoxide as well I wondered what the 952 00:39:28,600 --> 00:39:25,380 arguments are against putting all the 953 00:39:30,830 --> 00:39:28,610 oxygen into co as opposed to water a 954 00:39:32,900 --> 00:39:30,840 Glenn's question is why can't you just 955 00:39:34,670 --> 00:39:32,910 put all the oxygen in the co formation 956 00:39:36,560 --> 00:39:34,680 and I think the sort of simple-minded 957 00:39:38,750 --> 00:39:36,570 answer is that cosmically there's more 958 00:39:40,250 --> 00:39:38,760 oxygen in their carbon so if you do that 959 00:39:42,980 --> 00:39:40,260 you have oxygen left over and that goes 960 00:39:45,800 --> 00:39:42,990 into water you can there are suggestions 961 00:39:47,090 --> 00:39:45,810 that there are carbon planets if there 962 00:39:49,490 --> 00:39:47,100 are carbon planets where there's more 963 00:39:51,410 --> 00:39:49,500 carbon than oxygen than carbon will then 964 00:39:53,150 --> 00:39:51,420 all the oxygen or most of it will go 965 00:39:54,950 --> 00:39:53,160 into co formation and you won't have 966 00:39:59,300 --> 00:39:54,960 water advanced but I think carbon 967 00:40:04,430 --> 00:39:59,310 planets are considered exotic so that's 968 00:40:11,130 --> 00:40:07,950 have a question here hecka go to bright 969 00:40:14,099 --> 00:40:11,140 you actually show is the nice high road 970 00:40:15,510 --> 00:40:14,109 drive the spectra I show the question is 971 00:40:17,130 --> 00:40:15,520 are the spectra show for the night side 972 00:40:19,440 --> 00:40:17,140 the spectra were measured during 973 00:40:20,940 --> 00:40:19,450 secondary Eclipse secondary Eclipse 974 00:40:22,770 --> 00:40:20,950 we're looking at the dayside of the 975 00:40:24,539 --> 00:40:22,780 planet the planet on the opposite side 976 00:40:27,480 --> 00:40:24,549 of the star will start shining directly 977 00:40:29,099 --> 00:40:27,490 out of it I mean I mean you're planning 978 00:40:31,289 --> 00:40:29,109 design is behind so if you don't get to 979 00:40:34,559 --> 00:40:31,299 cope and putting trevino onions are 980 00:40:36,420 --> 00:40:34,569 being behind this time then you get when 981 00:40:38,370 --> 00:40:36,430 you get this is truck from here when 982 00:40:44,220 --> 00:40:38,380 it's in front of you minus the one in 983 00:40:46,279 --> 00:40:44,230 fact we're seeing we're seeing that the 984 00:40:49,020 --> 00:40:46,289 internet in Eclipse and out of eclipse 985 00:40:51,089 --> 00:40:49,030 phase difference is very small compared 986 00:40:52,980 --> 00:40:51,099 to the orbit of the planet so we're 987 00:40:54,990 --> 00:40:52,990 we're essentially looking at the same 988 00:40:58,470 --> 00:40:55,000 subtracting the same phase out of 989 00:40:59,730 --> 00:40:58,480 eclipses in Eclipse so we're the 990 00:41:01,920 --> 00:40:59,740 different that difference is just the 991 00:41:05,160 --> 00:41:01,930 thermal readmission of the photons from 992 00:41:06,569 --> 00:41:05,170 the planet I can try that what you're 993 00:41:10,170 --> 00:41:06,579 really doing it each wavelength 994 00:41:12,329 --> 00:41:10,180 immeasurably contrast between out of the 995 00:41:15,690 --> 00:41:12,339 coach genetic codes it's that contrast 996 00:41:17,940 --> 00:41:15,700 we can show the spectral signature at 997 00:41:19,559 --> 00:41:17,950 that waiver that's correct yes we're 998 00:41:21,779 --> 00:41:19,569 measuring the white button the excess 999 00:41:24,269 --> 00:41:21,789 intensity at 7.8 minor enhancements 1000 00:41:25,829 --> 00:41:24,279 we're reconstructing here we're 1001 00:41:27,990 --> 00:41:25,839 reconstructing the spectrum of the 1002 00:41:31,400 --> 00:41:28,000 planet from the wavelength dependence of 1003 00:41:33,670 --> 00:41:31,410 the secondary Eclipse amplitude correct 1004 00:41:37,160 --> 00:41:33,680 my question 1005 00:41:39,470 --> 00:41:37,170 okay Goddard still asking just a Drake 1006 00:41:44,210 --> 00:41:39,480 there's 14 planets that have been 1007 00:41:46,310 --> 00:41:44,220 detected in an eclipse they're 14 bright 1008 00:41:48,110 --> 00:41:46,320 extrasolar transiting planets and 1009 00:41:50,030 --> 00:41:48,120 clients and fainter ones what are the 1010 00:41:54,200 --> 00:41:50,040 chances of getting structure from the 1011 00:41:58,010 --> 00:41:54,210 other 12 the the two that I've shown you 1012 00:42:00,140 --> 00:41:58,020 are but by far the two brightest so I at 1013 00:42:01,610 --> 00:42:00,150 least have not proposed to use fitzer on 1014 00:42:04,070 --> 00:42:01,620 the fainter runs maybe someone else has 1015 00:42:06,050 --> 00:42:04,080 I think you would really be photon stars 1016 00:42:08,990 --> 00:42:06,060 in that case and you would require 1017 00:42:12,320 --> 00:42:09,000 multiple multiple eclipses to obtain 1018 00:42:18,980 --> 00:42:12,330 spectra jwst on the other hand will 1019 00:42:21,380 --> 00:42:18,990 really do these planets really well come 1020 00:42:23,150 --> 00:42:21,390 up to the table and ask ask the question 1021 00:42:27,260 --> 00:42:23,160 actually and then maybe we shout at the 1022 00:42:29,870 --> 00:42:27,270 other notes you've shown the silicon 1023 00:42:31,880 --> 00:42:29,880 emission from the planetary potential a 1024 00:42:33,830 --> 00:42:31,890 haze how do you know that those 1025 00:42:36,290 --> 00:42:33,840 silicates aren't in the disk of the 1026 00:42:37,790 --> 00:42:36,300 system ah that's a good question he 1027 00:42:39,410 --> 00:42:37,800 asked how do we know that the silicate 1028 00:42:42,590 --> 00:42:39,420 is really associated with the planets 1029 00:42:46,640 --> 00:42:42,600 and not associated with it with a disk 1030 00:42:48,530 --> 00:42:46,650 around the star well because what we see 1031 00:42:50,600 --> 00:42:48,540 is in eclipse minus out of it clicks 1032 00:42:52,610 --> 00:42:50,610 that's a very short that's a very short 1033 00:42:55,310 --> 00:42:52,620 Kemp relatively short temporal 1034 00:42:57,590 --> 00:42:55,320 modulation only really a small fraction 1035 00:43:00,170 --> 00:42:57,600 of any disk could be if there's a lump 1036 00:43:02,870 --> 00:43:00,180 in the disk it would have to be very 1037 00:43:04,700 --> 00:43:02,880 closely associated to the planet we 1038 00:43:06,590 --> 00:43:04,710 couldn't rule out the possibility that 1039 00:43:08,810 --> 00:43:06,600 there's some extended dust envelope 1040 00:43:10,310 --> 00:43:08,820 around the planet that's emitting but 1041 00:43:12,500 --> 00:43:10,320 people but you don't see that during 1042 00:43:14,060 --> 00:43:12,510 transfer the radius of this planet 1043 00:43:16,370 --> 00:43:14,070 during transit at infrared wavelengths 1044 00:43:18,110 --> 00:43:16,380 is the same as in the visible so if 1045 00:43:20,950 --> 00:43:18,120 there's if there's some dust around the 1046 00:43:23,060 --> 00:43:20,960 planet it's very low optical depth and 1047 00:43:26,089 --> 00:43:23,070 it would have to get very close to the 1048 00:43:31,579 --> 00:43:28,339 Drake I'd like to ask a question and 1049 00:43:33,319 --> 00:43:31,589 then we have a question at Colorado okay 1050 00:43:37,489 --> 00:43:33,329 question I'd like to ask comes back to 1051 00:43:39,229 --> 00:43:37,499 the discussion you included about the 1052 00:43:40,729 --> 00:43:39,239 effects of the temperature gradient 1053 00:43:42,979 --> 00:43:40,739 since you're looking in the thermal 1054 00:43:44,359 --> 00:43:42,989 infrared of course the visibility of 1055 00:43:46,579 --> 00:43:44,369 spectral features is going to depend 1056 00:43:48,410 --> 00:43:46,589 upon a vertical temperature contrast in 1057 00:43:50,450 --> 00:43:48,420 the atmosphere you mentioned the 1058 00:43:52,219 --> 00:43:50,460 possible effects of circulation could 1059 00:43:55,370 --> 00:43:52,229 you just talk a little bit more about 1060 00:43:57,890 --> 00:43:55,380 what you expect the vertical temperature 1061 00:44:01,489 --> 00:43:57,900 profiles in these atmospheres to be and 1062 00:44:03,289 --> 00:44:01,499 how variations in that or uncertainty in 1063 00:44:05,299 --> 00:44:03,299 your understanding of that affects your 1064 00:44:07,069 --> 00:44:05,309 interpretation of the spectra 1065 00:44:11,660 --> 00:44:07,079 particularly whether or not you see a 1066 00:44:14,749 --> 00:44:11,670 feature well you do expect when these 1067 00:44:17,089 --> 00:44:14,759 planets were initially discovered my 1068 00:44:18,979 --> 00:44:17,099 first reaction was that they probably 1069 00:44:21,680 --> 00:44:18,989 have very hot stratosphere scuzz there 1070 00:44:23,630 --> 00:44:21,690 so in so close to this star we know 1071 00:44:27,940 --> 00:44:23,640 jupiter has a stratosphere because it's 1072 00:44:31,279 --> 00:44:27,950 it no it has methane methane absorbs 1073 00:44:33,200 --> 00:44:31,289 solar radiation heats the atmosphere if 1074 00:44:34,430 --> 00:44:33,210 you put a planet in that close to the 1075 00:44:36,440 --> 00:44:34,440 star I would think it would have a very 1076 00:44:39,620 --> 00:44:36,450 strong stratosphere but when people that 1077 00:44:42,229 --> 00:44:39,630 shows that shows how poor my intuition 1078 00:44:44,839 --> 00:44:42,239 is because when people actually put the 1079 00:44:47,089 --> 00:44:44,849 opacities of water vapor and the various 1080 00:44:49,430 --> 00:44:47,099 molecules into their codes they found 1081 00:44:51,979 --> 00:44:49,440 that indeed the the temperature in the 1082 00:44:53,479 --> 00:44:51,989 outer atmosphere was raised slightly but 1083 00:44:55,759 --> 00:44:53,489 not enough to actually invert the 1084 00:44:57,650 --> 00:44:55,769 temperature so the temperature profiles 1085 00:45:00,499 --> 00:44:57,660 of these planets are not all that 1086 00:45:03,339 --> 00:45:00,509 different from brown dwarf atmospheres 1087 00:45:05,749 --> 00:45:03,349 in the sense that their their their 1088 00:45:08,450 --> 00:45:05,759 highest temperatures are down deep and 1089 00:45:11,150 --> 00:45:08,460 the external boundary temperatures are 1090 00:45:12,920 --> 00:45:11,160 on the order of a hundred Kelvin versus 1091 00:45:16,370 --> 00:45:12,930 fifteen hundred Kelvin deeper deeper 1092 00:45:18,349 --> 00:45:16,380 down now maybe the observations are 1093 00:45:21,380 --> 00:45:18,359 showing maybe the Spitzer observations 1094 00:45:23,630 --> 00:45:21,390 are showing us that didn't we have more 1095 00:45:26,420 --> 00:45:23,640 to learn there but at least that's the 1096 00:45:28,130 --> 00:45:26,430 current understanding i guess the second 1097 00:45:30,739 --> 00:45:28,140 part of my question was is that 1098 00:45:33,400 --> 00:45:30,749 affecting the visibility of features so 1099 00:45:35,630 --> 00:45:33,410 for example when you don't see a feature 1100 00:45:37,100 --> 00:45:35,640 could it be that the 1101 00:45:38,600 --> 00:45:37,110 that you're looking for is there but the 1102 00:45:41,120 --> 00:45:38,610 temperature contrast to enable you to 1103 00:45:46,430 --> 00:45:41,130 see it isn't yes yes that's that's 1104 00:45:51,830 --> 00:45:46,440 that's the predominant view yes okay I'm 1105 00:45:57,680 --> 00:45:51,840 University of Colorado hydrate god this 1106 00:46:00,800 --> 00:45:57,690 is Tom ears nice talk um I wanted to ask 1107 00:46:02,990 --> 00:46:00,810 a question actually two questions on the 1108 00:46:05,750 --> 00:46:03,000 first one is what are the prospects of 1109 00:46:09,290 --> 00:46:05,760 performing this type of experiment with 1110 00:46:11,540 --> 00:46:09,300 a large telescope from the ground in the 1111 00:46:13,610 --> 00:46:11,550 same name for red and what are the 1112 00:46:16,400 --> 00:46:13,620 prospects for doing this in the 1113 00:46:18,770 --> 00:46:16,410 ultraviolet from space and then the 1114 00:46:21,710 --> 00:46:18,780 second question is that you described an 1115 00:46:22,910 --> 00:46:21,720 alternative analysis of your own data by 1116 00:46:24,980 --> 00:46:22,920 another group and I wonder if you could 1117 00:46:28,340 --> 00:46:24,990 just mention what their results were now 1118 00:46:30,260 --> 00:46:28,350 they differ from yours um okay let's see 1119 00:46:31,670 --> 00:46:30,270 let's take the first part of that 1120 00:46:34,040 --> 00:46:31,680 question first can we do this from the 1121 00:46:35,780 --> 00:46:34,050 ground the ground-based observers are 1122 00:46:37,640 --> 00:46:35,790 knocking on the door of Planet 1123 00:46:40,460 --> 00:46:37,650 protection there is a tentative like 1124 00:46:42,550 --> 00:46:40,470 three sigma detection of one of the oval 1125 00:46:46,580 --> 00:46:42,560 transiting planets at two microns 1126 00:46:48,320 --> 00:46:46,590 bye-bye snellen that we think that the 1127 00:46:50,930 --> 00:46:48,330 way to do this from the ground is to is 1128 00:46:53,240 --> 00:46:50,940 to find a system with a bright nearby 1129 00:46:54,830 --> 00:46:53,250 comparison star so we can take out the 1130 00:46:57,110 --> 00:46:54,840 effects of the Earth's atmosphere and 1131 00:46:58,850 --> 00:46:57,120 that would add an additional 1132 00:47:01,370 --> 00:46:58,860 complication to the analysis that we 1133 00:47:04,430 --> 00:47:01,380 think it's possibly and there are 1134 00:47:05,600 --> 00:47:04,440 proposals being submitted to try that 1135 00:47:08,420 --> 00:47:05,610 and we think it's eventually that will 1136 00:47:10,780 --> 00:47:08,430 be possible of the advantage of course 1137 00:47:13,370 --> 00:47:10,790 doing it from the ground is that is that 1138 00:47:17,000 --> 00:47:13,380 Spitzer will run out of cryogen in the 1139 00:47:18,440 --> 00:47:17,010 spring of 2009 and and also Spitzer has 1140 00:47:19,820 --> 00:47:18,450 limited wavelength coverage there are 1141 00:47:21,620 --> 00:47:19,830 important features in the spectra 1142 00:47:24,140 --> 00:47:21,630 between two to five microns where 1143 00:47:25,520 --> 00:47:24,150 Spitzer has no wavelength we were not 1144 00:47:28,160 --> 00:47:25,530 accessible to those wavelengths been 1145 00:47:29,840 --> 00:47:28,170 Spitzer then the second part of the 1146 00:47:32,480 --> 00:47:29,850 question was what about the Swain 1147 00:47:35,290 --> 00:47:32,490 analysis and so I have to go back to my 1148 00:47:37,370 --> 00:47:35,300 talk you make my talk come up again 1149 00:47:39,110 --> 00:47:37,380 Martha we need the presenter ballbag 1150 00:47:41,480 --> 00:47:39,120 yeah you should be able to control right 1151 00:47:51,330 --> 00:47:41,490 now this one 1152 00:47:53,640 --> 00:47:51,340 okay okay how do I make it oh in 1153 00:48:03,800 --> 00:47:53,650 tribute go down to the bottom vertical 1154 00:48:03,810 --> 00:48:18,440 ok 1155 00:48:24,020 --> 00:48:21,380 whereas our analysis uses these 1156 00:48:26,240 --> 00:48:24,030 differential eclipses what what marks 1157 00:48:27,740 --> 00:48:26,250 Wayne's analysis well I guess his papers 1158 00:48:29,900 --> 00:48:27,750 still under review what he does is he 1159 00:48:32,990 --> 00:48:29,910 actually subtracts the spectra out of 1160 00:48:35,480 --> 00:48:33,000 eclipse minus in eclipse and we do that 1161 00:48:38,000 --> 00:48:35,490 too only we do that as a check on the 1162 00:48:40,490 --> 00:48:38,010 analysis we have we we rely on these 1163 00:48:42,079 --> 00:48:40,500 differential eclipses and and I would 1164 00:48:45,349 --> 00:48:42,089 argue that you really have to rely on 1165 00:48:47,000 --> 00:48:45,359 the differential eclipses because if you 1166 00:48:50,180 --> 00:48:47,010 don't do that there can be effects in 1167 00:48:52,370 --> 00:48:50,190 the time series that that are 1168 00:48:54,440 --> 00:48:52,380 unaccounted for in your analysis and in 1169 00:48:56,810 --> 00:48:54,450 fact if you look at marks paper you find 1170 00:49:00,560 --> 00:48:56,820 that he has a term called G of T which 1171 00:49:02,930 --> 00:49:00,570 is the gain of the of the detector 1172 00:49:05,780 --> 00:49:02,940 system and he explicitly assumes that 1173 00:49:09,859 --> 00:49:05,790 that is wavelength independent let me go 1174 00:49:12,890 --> 00:49:09,869 back a little further this ramp this 1175 00:49:15,380 --> 00:49:12,900 ramp is like a game because it's due to 1176 00:49:17,540 --> 00:49:15,390 charge trapping in the detector when you 1177 00:49:20,450 --> 00:49:17,550 illuminate the detector with a faint 1178 00:49:22,430 --> 00:49:20,460 source that is the stellar spectrum at 1179 00:49:24,410 --> 00:49:22,440 long wavelengths it takes a lot longer 1180 00:49:27,620 --> 00:49:24,420 for it to ramp up to full sensitivity 1181 00:49:31,430 --> 00:49:27,630 than it does at at Short wavelengths 1182 00:49:32,990 --> 00:49:31,440 where there is a high intensity being 1183 00:49:34,790 --> 00:49:33,000 shown on the detector in other words the 1184 00:49:37,430 --> 00:49:34,800 gaining of the detector is effectively 1185 00:49:40,550 --> 00:49:37,440 wavelength dependent whereas marks 1186 00:49:43,880 --> 00:49:40,560 analysis explicitly assumes that it is 1187 00:49:45,319 --> 00:49:43,890 not and that's not correct and for that 1188 00:49:47,300 --> 00:49:45,329 reason I think he's probably not 1189 00:49:52,870 --> 00:49:47,310 deriving the spectrum of the planet at 1190 00:49:59,209 --> 00:49:55,759 second one hey there is a spectra from 1191 00:50:04,009 --> 00:49:59,219 space UV spectra from space have been 1192 00:50:09,289 --> 00:50:04,019 measured by Hubble okay Arizona have a 1193 00:50:13,269 --> 00:50:09,299 question if there were a ring around 1194 00:50:16,519 --> 00:50:13,279 this planet it would of course produce 1195 00:50:20,029 --> 00:50:16,529 likely emission lines or omission bands 1196 00:50:23,449 --> 00:50:20,039 whatever one calls it and one would help 1197 00:50:26,929 --> 00:50:23,459 distinguish that because the Eclipse 1198 00:50:30,069 --> 00:50:26,939 would be broader because of the diameter 1199 00:50:32,719 --> 00:50:30,079 of the ring system would be broader and 1200 00:50:36,109 --> 00:50:32,729 therefore I'd like to know what evidence 1201 00:50:38,929 --> 00:50:36,119 you have that in this omission that the 1202 00:50:42,079 --> 00:50:38,939 Eclipse does indeed have the threats 1203 00:50:46,069 --> 00:50:42,089 associated exactly with that you see of 1204 00:50:48,109 --> 00:50:46,079 the planet continuum well at one time at 1205 00:50:50,509 --> 00:50:48,119 one time rings around this planet were 1206 00:50:52,399 --> 00:50:50,519 my favorite theory because they're as 1207 00:50:54,139 --> 00:50:52,409 you know the planets radius is larger 1208 00:50:57,139 --> 00:50:54,149 than can be accounted for by simple 1209 00:50:58,999 --> 00:50:57,149 models one way to explain that would be 1210 00:51:00,409 --> 00:50:59,009 perhaps ring a ring system which was 1211 00:51:03,409 --> 00:51:00,419 increasing the effective area of the 1212 00:51:07,999 --> 00:51:03,419 planet that beautiful theory has been 1213 00:51:09,679 --> 00:51:08,009 contradicted by ugly facts the transit 1214 00:51:12,079 --> 00:51:09,689 of the planet was measured at Spitzer 1215 00:51:13,879 --> 00:51:12,089 way the transit the primary eclipse of 1216 00:51:16,069 --> 00:51:13,889 the planet was measured by Spitzer 1217 00:51:19,189 --> 00:51:16,079 wavelengths by Jeremy Richardson at 24 1218 00:51:21,620 --> 00:51:19,199 microns and he obtains an a radius for 1219 00:51:23,689 --> 00:51:21,630 the planet in very close agreement with 1220 00:51:25,879 --> 00:51:23,699 the visible results so if there are 1221 00:51:27,469 --> 00:51:25,889 rings they have almost the same optical 1222 00:51:31,519 --> 00:51:27,479 depth at two very very different 1223 00:51:34,339 --> 00:51:31,529 wavelengths which is may be possible but 1224 00:51:36,709 --> 00:51:34,349 but it does argue against it somewhat 1225 00:51:38,539 --> 00:51:36,719 then the other the the other argument 1226 00:51:40,370 --> 00:51:38,549 that is the width of the Eclipse is that 1227 00:51:41,479 --> 00:51:40,380 we have additional observations that we 1228 00:51:43,519 --> 00:51:41,489 haven't published yet and I haven't 1229 00:51:47,029 --> 00:51:43,529 shown you and I know that's a terrible 1230 00:51:49,189 --> 00:51:47,039 thing to say but what but add to 1231 00:51:50,779 --> 00:51:49,199 original discovery observations and if 1232 00:51:53,120 --> 00:51:50,789 the Eclipse were significantly broader 1233 00:51:55,159 --> 00:51:53,130 you know the aggressor egress Tom were 1234 00:51:57,559 --> 00:51:55,169 significantly different then we would 1235 00:51:59,280 --> 00:51:57,569 have seen that and also de Charbonneau 1236 00:52:01,050 --> 00:51:59,290 has additional data that he hasn't 1237 00:52:11,000 --> 00:52:01,060 published yet and he would have seen it 1238 00:52:17,700 --> 00:52:14,880 he's red light on them hello this is 1239 00:52:19,710 --> 00:52:17,710 paul davis at aims to operational 1240 00:52:23,310 --> 00:52:19,720 questions one you said you took half 1241 00:52:25,170 --> 00:52:23,320 your data before the secondary transept 1242 00:52:28,170 --> 00:52:25,180 and half the data during the secondary 1243 00:52:30,030 --> 00:52:28,180 transit wouldn't it be better in regard 1244 00:52:32,100 --> 00:52:30,040 to system editors to take a quarter 1245 00:52:33,630 --> 00:52:32,110 before a half during and a quarter after 1246 00:52:35,840 --> 00:52:33,640 yeah that's that's what we actually do 1247 00:52:40,230 --> 00:52:35,850 we started we take a quarter before 1248 00:52:42,000 --> 00:52:40,240 patent yes with some variation on that 1249 00:52:44,520 --> 00:52:42,010 depending on how quickly the telescope 1250 00:52:46,860 --> 00:52:44,530 people can the exact starting time and 1251 00:52:49,740 --> 00:52:46,870 give them a window okay so thank you the 1252 00:52:53,250 --> 00:52:49,750 other question is you mentioned the one 1253 00:52:55,320 --> 00:52:53,260 hour point 05 arcsecond oscillation and 1254 00:52:58,320 --> 00:52:55,330 you also mention doing a nod in the 1255 00:53:00,180 --> 00:52:58,330 middle of the transit measurement to get 1256 00:53:03,180 --> 00:53:00,190 a reference does the nod reset the 1257 00:53:05,100 --> 00:53:03,190 oscillation and can you use the knob to 1258 00:53:06,870 --> 00:53:05,110 control the oscillations or is it better 1259 00:53:09,480 --> 00:53:06,880 to deal with in calculation a lot 1260 00:53:12,530 --> 00:53:09,490 calculation alee the nod does the nod 1261 00:53:15,840 --> 00:53:12,540 does not seem to reset the oscillation 1262 00:53:17,610 --> 00:53:15,850 the having the telescope nod is both 1263 00:53:19,620 --> 00:53:17,620 good and bad i want to comment about the 1264 00:53:21,420 --> 00:53:19,630 nod that it's good in the sense that 1265 00:53:23,790 --> 00:53:21,430 when we see this like this differential 1266 00:53:27,000 --> 00:53:23,800 eclipse that i have shown you on the 1267 00:53:28,890 --> 00:53:27,010 screen up here that the that the ingress 1268 00:53:30,600 --> 00:53:28,900 of this eclipse is actually measured 1269 00:53:33,180 --> 00:53:30,610 with different pixels of pixels of the 1270 00:53:34,380 --> 00:53:33,190 detector than the egress because we're 1271 00:53:36,750 --> 00:53:34,390 always worried that you know there's 1272 00:53:39,480 --> 00:53:36,760 some rogue pixel who is causing a 1273 00:53:41,700 --> 00:53:39,490 spurious result so in so by having a 1274 00:53:44,550 --> 00:53:41,710 telescope nod there have to be sort of 1275 00:53:46,470 --> 00:53:44,560 two sets of rogue pixels and have to 1276 00:53:49,290 --> 00:53:46,480 conspire together to make an ingress and 1277 00:53:50,940 --> 00:53:49,300 egress at the right time and at that 1278 00:53:54,120 --> 00:53:50,950 point you just you don't worry about 1279 00:53:56,880 --> 00:53:54,130 things that bizarre the disadvantage of 1280 00:53:59,549 --> 00:53:56,890 the nod is that it may reset the ramp to 1281 00:53:59,549 --> 00:53:59,559 some extent 1282 00:54:05,489 --> 00:54:03,269 you can see right here there's a there's 1283 00:54:08,219 --> 00:54:05,499 a the charge trapping in the detector 1284 00:54:10,109 --> 00:54:08,229 when we move the spectrum when we move 1285 00:54:11,999 --> 00:54:10,119 the spectrum there's a little bit of a 1286 00:54:14,449 --> 00:54:12,009 reset of the ramp I would like not to 1287 00:54:17,370 --> 00:54:14,459 have that I've since concluded that for 1288 00:54:19,439 --> 00:54:17,380 true and also for reasons of comparing 1289 00:54:23,459 --> 00:54:19,449 to Karl grill mayer where they don't nod 1290 00:54:29,579 --> 00:54:23,469 that we propose to redo this to do it 1291 00:54:36,839 --> 00:54:29,589 without the nod that's that's a real 1292 00:54:39,390 --> 00:54:36,849 subtlety of the other day thank you okay 1293 00:54:40,799 --> 00:54:39,400 there are no more hands raised we can 1294 00:54:42,420 --> 00:54:40,809 send it back to Goddard for any more 1295 00:54:45,150 --> 00:54:42,430 questions at Goddard if there are any 1296 00:54:48,779 --> 00:54:45,160 there oh god it always has question oh I 1297 00:54:50,699 --> 00:54:48,789 know I know drink in the spectrum they 1298 00:54:53,069 --> 00:54:50,709 seem to be another feature at the longer 1299 00:54:54,779 --> 00:54:53,079 wavelength is that another feature at 1300 00:54:57,029 --> 00:54:54,789 the hunger way Ron Paul than I have 1301 00:54:58,709 --> 00:54:57,039 microns or somewhere around there well 1302 00:55:01,769 --> 00:54:58,719 some kind of scrubs but look at the 1303 00:55:03,719 --> 00:55:01,779 error bars here and look at the look at 1304 00:55:05,339 --> 00:55:03,729 the agreement or disagreement between 1305 00:55:08,130 --> 00:55:05,349 the two different eclipses which is what 1306 00:55:09,630 --> 00:55:08,140 we have to go by you can treat basically 1307 00:55:12,120 --> 00:55:09,640 just scatter at the long wavelengths and 1308 00:55:14,939 --> 00:55:12,130 in fact the spectrum goes the spectre go 1309 00:55:16,589 --> 00:55:14,949 out to 14 microns but the Spitzer people 1310 00:55:19,890 --> 00:55:16,599 say that they're scattered light beyond 1311 00:55:23,839 --> 00:55:19,900 13.2 so we don't even we don't even try 1312 00:55:26,489 --> 00:55:23,849 be on there it's called the teardrop 1313 00:55:32,339 --> 00:55:26,499 short low that's a short guy this is 1314 00:55:33,599 --> 00:55:32,349 this is short glo SL this is SL too we 1315 00:55:36,020 --> 00:55:33,609 don't have enough photons to get a high 1316 00:55:39,980 --> 00:55:37,880 have a question have you ever done here 1317 00:55:42,560 --> 00:55:39,990 Anunnaki effects for some of these 1318 00:55:46,940 --> 00:55:42,570 missions or have we considered non-lte 1319 00:55:49,310 --> 00:55:46,950 effects if if get this if if this 1320 00:55:52,010 --> 00:55:49,320 feature is is really due to a 1321 00:55:54,380 --> 00:55:52,020 carbon-carbon stretch which we think is 1322 00:55:55,790 --> 00:55:54,390 kind of speculative but if it were then 1323 00:56:00,470 --> 00:55:55,800 it would have to be due to fluorescence 1324 00:56:04,250 --> 00:56:00,480 which is non a non-thermal so yeah there 1325 00:56:08,720 --> 00:56:04,260 could be non thermal effects yeah okay 1326 00:56:11,600 --> 00:56:08,730 well when I look at the spectrum it 1327 00:56:14,300 --> 00:56:11,610 seems to me that both grill mares 1328 00:56:16,460 --> 00:56:14,310 respect that the image and yours could 1329 00:56:18,620 --> 00:56:16,470 be interpreted as having a double Pete 1330 00:56:21,820 --> 00:56:18,630 structure that was had a minimum near 1331 00:56:24,470 --> 00:56:21,830 7.6 microns and two beats one if I'm 1332 00:56:27,890 --> 00:56:24,480 firefight like wanna lower way life is 1333 00:56:31,670 --> 00:56:27,900 that supportable is that not correct hmm 1334 00:56:34,300 --> 00:56:31,680 I mean I think you could argue in favor 1335 00:56:37,460 --> 00:56:34,310 of a lot of different explanations but 1336 00:56:40,970 --> 00:56:37,470 but since there are proposals to 1337 00:56:42,530 --> 00:56:40,980 reabsorb oath systems and those and in 1338 00:56:44,930 --> 00:56:42,540 at these wavelengths and at other 1339 00:56:47,870 --> 00:56:44,940 wavelengths a lot much will be clarified 1340 00:56:49,580 --> 00:56:47,880 after this summer these these are 1341 00:56:55,220 --> 00:56:49,590 observable by Spitzer during the summers 1342 00:56:56,540 --> 00:56:55,230 open which I believe me what we proposed 1343 00:56:58,940 --> 00:56:56,550 to observe it at somewhat shorter 1344 00:57:01,100 --> 00:56:58,950 wavelengths 528 micron where we expect 1345 00:57:03,230 --> 00:57:01,110 the flux to decrease the contrast to 1346 00:57:05,300 --> 00:57:03,240 decrease strongly and Karl grill mayor 1347 00:57:08,840 --> 00:57:05,310 has observed it as proposed to observe 1348 00:57:11,210 --> 00:57:08,850 HD 189733 at many other wavelengths I 1349 00:57:16,130 --> 00:57:11,220 don't know the specifics where we do 1350 00:57:18,320 --> 00:57:16,140 next expect to see water where would we 1351 00:57:19,760 --> 00:57:18,330 next expect to see water actually I 1352 00:57:23,360 --> 00:57:19,770 think the best prospects for seeing 1353 00:57:26,210 --> 00:57:23,370 water are from the ground to detect the 1354 00:57:28,640 --> 00:57:26,220 there's a peek at 3.8 microns between 1355 00:57:30,020 --> 00:57:28,650 water and carbon dioxide absorption that 1356 00:57:31,850 --> 00:57:30,030 I think can be detected using 1357 00:57:35,030 --> 00:57:31,860 ground-based observations in these 1358 00:57:37,730 --> 00:57:35,040 planets so that's that's the way I think 1359 00:57:40,730 --> 00:57:37,740 is the real way to detect water Drake 1360 00:57:42,350 --> 00:57:40,740 could you just amplify for a second on 1361 00:57:45,319 --> 00:57:42,360 the comment you made just a minute ago 1362 00:57:48,349 --> 00:57:45,329 that if that featured around 1363 00:57:50,569 --> 00:57:48,359 just short of 8 microns is carbon-carbon 1364 00:57:55,849 --> 00:57:50,579 stretch it has to be fluorescence why is 1365 00:57:59,930 --> 00:57:55,859 that um because that's the way that's 1366 00:58:01,910 --> 00:57:59,940 the way a mission from PAH molecules is 1367 00:58:03,499 --> 00:58:01,920 seen in Astrophysical sources at these 1368 00:58:05,390 --> 00:58:03,509 wavelengths so I guess it doesn't have 1369 00:58:09,440 --> 00:58:05,400 to be but I would expect that it would 1370 00:58:10,969 --> 00:58:09,450 be in that case because you know 1371 00:58:13,670 --> 00:58:10,979 carbon-carbon stretch wouldn't would 1372 00:58:16,670 --> 00:58:13,680 indicate PAH and pah is seen in many 1373 00:58:18,229 --> 00:58:16,680 sources now by Spitzer and it's always 1374 00:58:21,140 --> 00:58:18,239 in a mission and it's new to 1375 00:58:22,609 --> 00:58:21,150 fluorescence certainly you're in you 1376 00:58:24,140 --> 00:58:22,619 know this planet is in close enough to a 1377 00:58:26,479 --> 00:58:24,150 hot enough star there's enough UV 1378 00:58:28,069 --> 00:58:26,489 because fluorescence and you're giving 1379 00:58:30,049 --> 00:58:28,079 up in dirty extremity region I mean 1380 00:58:32,749 --> 00:58:30,059 we're being speculative now but if there 1381 00:58:34,549 --> 00:58:32,759 are there are silicate clouds those 1382 00:58:37,009 --> 00:58:34,559 silicate grains could be associated with 1383 00:58:39,769 --> 00:58:37,019 pah molecules also and they could 1384 00:58:43,130 --> 00:58:39,779 fluoresce well when I think in this case 1385 00:58:44,870 --> 00:58:43,140 the fluorescence is not the same as kind 1386 00:58:47,239 --> 00:58:44,880 of fluorescence as if you had a diatomic 1387 00:58:49,640 --> 00:58:47,249 molecule which was pumped at a higher 1388 00:58:51,829 --> 00:58:49,650 statement re-radiated at different 1389 00:58:55,999 --> 00:58:51,839 wavelengths it's really more that you 1390 00:58:59,390 --> 00:58:56,009 might have a pH grain or macro molecule 1391 00:59:01,150 --> 00:58:59,400 that is flash heated by UV photon and 1392 00:59:04,819 --> 00:59:01,160 then you get vibrational excitation 1393 00:59:07,699 --> 00:59:04,829 which is seen in addition but the grain 1394 00:59:10,039 --> 00:59:07,709 itself is probably slightly warm welcome 1395 00:59:13,249 --> 00:59:10,049 by that UV photon that's that's exactly 1396 00:59:14,870 --> 00:59:13,259 correct yeah it's not quite the same 1397 00:59:16,400 --> 00:59:14,880 it's not question say but it's still 1398 00:59:19,009 --> 00:59:16,410 referred to as fluorescence because it's 1399 00:59:21,890 --> 00:59:19,019 a repercussion a roomie it's a 1400 00:59:29,400 --> 00:59:21,900 readmission of the UV radiation from the 1401 00:59:33,690 --> 00:59:31,620 well drink there aren't any hands raised 1402 00:59:38,570 --> 00:59:33,700 around the net and so I just wanted to 1403 00:59:52,020 --> 00:59:49,230 weekly and thank you again and look