2 00:00:06,420 --> 00:00:14,300 The third way of measuring temperatures is by satellite. 4 00:00:14,300 --> 00:00:18,940 And at first blush, this would seem like it is the ideal way of making measurements 5 00:00:18,940 --> 00:00:21,080 because it's out in space and nobody's touching 6 00:00:21,080 --> 00:00:23,650 it and it's covering the whole world. 7 00:00:23,650 --> 00:00:29,120 It's actually not the best way to get a time series measurement of temperature 8 00:00:29,120 --> 00:00:35,500 because orbits of the satellite can decay, and over the time span 9 00:00:35,500 --> 00:00:38,770 that we're talking about there were several satellites that flew, 10 00:00:38,770 --> 00:00:41,420 so they all had to be calibrated against each other 11 00:00:42,540 --> 00:00:45,710 in order to come up with a time series. 12 00:00:45,710 --> 00:00:48,390 The satellite is really good at 13 00:00:48,390 --> 00:00:53,250 making a map, an instantaneous snapshot of surface temperatures. 14 00:00:53,250 --> 00:00:57,100 But for a time series it's a bit more work. 15 00:00:57,100 --> 00:01:01,240 And for a long time the satellites didn't show 16 00:01:01,240 --> 00:01:06,630 the same warming as the land and the sea surface temperature records were showing. 17 00:01:06,630 --> 00:01:12,190 This seemed like it could be a major problem. 18 00:01:12,190 --> 00:01:17,380 It turns out that the way that satellites work 19 00:01:17,380 --> 00:01:23,090 is they measure microwave radiation that comes from molecular oxygen. 20 00:01:23,090 --> 00:01:26,520 And just like greenhouse black body 21 00:01:26,520 --> 00:01:31,870 radiation, this wavelength of light 22 00:01:31,870 --> 00:01:34,800 is more intense when the oxygen is warmer. 23 00:01:34,800 --> 00:01:38,040 The difficulty is that the satellite is 24 00:01:38,040 --> 00:01:40,490 sensitive not just to the temperature right at 25 00:01:40,490 --> 00:01:43,600 the ground, but it's sensitive to temperatures 26 00:01:43,600 --> 00:01:47,410 throughout a range of altitudes in the atmosphere. 27 00:01:47,410 --> 00:01:52,550 And so if there's cooling higher up, that can pull 28 00:01:52,550 --> 00:01:55,560 the temperature the satellite sees down. 29 00:01:55,560 --> 00:01:57,760 You don't get the full signal 30 00:01:57,760 --> 00:02:00,660 from the ground, which is what we're interested in. 31 00:02:00,660 --> 00:02:05,260 After the latest round of corrections, 32 00:02:05,260 --> 00:02:08,950 you can see in this plot that the satellite temperatures, although very 33 00:02:08,950 --> 00:02:13,950 spiky, are also consistent with the land and sea surface temperature records. 34 00:02:13,950 --> 00:02:17,660 And that they show this warming in the last few decades that will 35 00:02:17,660 --> 00:02:21,780 turn out to be the smoking gun for a human impact on earth's climate.