1 00:00:00,033 --> 00:00:02,202 If you've ever looked at the weather forecast, 2 00:00:02,202 --> 00:00:04,338 you might be familiar with a percentage. 3 00:00:04,338 --> 00:00:06,540 It's the probability of precipitation. 4 00:00:06,540 --> 00:00:10,344 For example, sixty percent rain means that when similar 5 00:00:10,344 --> 00:00:13,580 forecasts have occurred in the past, only sixty percent 6 00:00:13,580 --> 00:00:14,815 actually rained. 7 00:00:14,815 --> 00:00:18,385 Uncertainties can sometimes leave us unprepared for storms 8 00:00:18,385 --> 00:00:19,286 and floods. 9 00:00:19,286 --> 00:00:23,423 But predictions improve as we monitor storms in greater detail 10 00:00:23,423 --> 00:00:26,193 and now NASA's newest precipitation satellite 11 00:00:26,193 --> 00:00:30,564 GPM is measuring the individual water drops that make up a storm. 12 00:00:30,564 --> 00:00:32,866 GPM's the first time we've been able to get this type 13 00:00:32,866 --> 00:00:34,801 of information on a global scale. 14 00:00:34,801 --> 00:00:39,339 To understand why this matters, it helps to know how weather predictions are made. 15 00:00:39,339 --> 00:00:42,943 First, computer programs known as weather models turn the 16 00:00:42,943 --> 00:00:46,246 processes in the atmosphere into math equations, 17 00:00:46,246 --> 00:00:48,282 governed by the laws of physics. 18 00:00:48,282 --> 00:00:51,985 These equations receive measurements of the land, sea, and air 19 00:00:51,985 --> 00:00:55,155 to get a view of the atmosphere's current state 20 00:00:55,155 --> 00:00:57,858 then predicts how it will change over time. 21 00:00:57,858 --> 00:01:00,727 To ensure predictions are correct, equations need 22 00:01:00,727 --> 00:01:03,163 to represent the atmosphere accurately. 23 00:01:03,163 --> 00:01:06,400 But one thing that's uncertain in the equations is how 24 00:01:06,400 --> 00:01:09,770 precipitation is structured within storms. 25 00:01:09,770 --> 00:01:13,340 A storm is made up of water drops of different sizes but without 26 00:01:13,340 --> 00:01:16,510 knowing these measurements many weather models operate under an 27 00:01:16,510 --> 00:01:19,046 assumption. What they do instead is have some assumption 28 00:01:19,046 --> 00:01:22,516 about how that's distributed. So as an example, for every one 29 00:01:22,516 --> 00:01:25,919 hundred small drops, there would be ten medium sized drops or 30 00:01:25,919 --> 00:01:28,755 just one large drop. Assumptions are made because 31 00:01:28,755 --> 00:01:32,326 researchers have only studied drops in isolated areas. 32 00:01:32,326 --> 00:01:35,796 Now for the first time, GPM is measuring the size and 33 00:01:35,796 --> 00:01:37,698 distribution of drops around the world. 34 00:01:37,698 --> 00:01:41,735 If we go into a storm we'll see varying drops sizes labeled 35 00:01:41,735 --> 00:01:42,402 different colors. 36 00:01:42,402 --> 00:01:47,007 Near the top, there are many small drops around .5,1 and 37 00:01:47,007 --> 00:01:51,745 2 millimeters. As they fall to the middle of the storm, some drops 38 00:01:51,745 --> 00:01:55,048 bump into others causing them to grow in size. 39 00:01:55,048 --> 00:02:00,887 Right at the bottom, drops can grow around four, five and six millimeters. 40 00:02:00,887 --> 00:02:05,292 This 3-D mosaic of water drops is called drop size distribution. 41 00:02:05,292 --> 00:02:08,628 It shows a high concentration of small drops colored in blues and 42 00:02:08,628 --> 00:02:11,698 greens near the top and lower concentrations of big 43 00:02:11,698 --> 00:02:14,534 drops colored in reds and yellows near the bottom. 44 00:02:14,534 --> 00:02:17,771 A storm with a higher ratio of red and yellow will contain more 45 00:02:17,771 --> 00:02:21,942 water than a higher ratio of blue and green. Without knowing 46 00:02:21,942 --> 00:02:26,146 the relationship or ratio of those large drops to the smaller or medium-sized drops. 47 00:02:26,146 --> 00:02:28,048 We can have a big error in how much 48 00:02:28,048 --> 00:02:31,018 rain we know fell and that can have some big implications for 49 00:02:31,018 --> 00:02:35,656 knowing long term accumulations which can help for flash flood prediction. 50 00:02:35,656 --> 00:02:38,425 Not only does it give a more accurate measurement of rainfall 51 00:02:38,425 --> 00:02:42,529 but drop sizes also give insight into the winds within a storm. 52 00:02:42,529 --> 00:02:45,699 Thunderstorms have a lot of wind associated with them and we all 53 00:02:45,699 --> 00:02:46,366 know this. 54 00:02:46,366 --> 00:02:48,769 But the strength of that wind actually depends on the size of 55 00:02:48,769 --> 00:02:51,571 the drops that are falling from that storm in some ways. 56 00:02:51,571 --> 00:02:54,908 Because a storm with small drops will have more evaporation, 57 00:02:54,908 --> 00:02:57,711 which cools the air more, which creates stronger winds. 58 00:02:57,711 --> 00:03:00,280 We've never been able to see how water droplet sizes vary 59 00:03:00,280 --> 00:03:01,882 globally until now. 60 00:03:01,882 --> 00:03:04,151 So what causes them to vary in different places? 61 00:03:04,151 --> 00:03:08,255 One factor is the temperature of the environment drops grow in. 62 00:03:08,255 --> 00:03:10,724 In the mid-latitudes a lot of those raindrops actually 63 00:03:10,724 --> 00:03:14,795 originated as snowflakes or even hailstones and snowflakes can 64 00:03:14,795 --> 00:03:16,663 grow a lot larger than cloud droplets can. 65 00:03:16,663 --> 00:03:19,332 So you have these big snowflakes that then melt into big 66 00:03:19,332 --> 00:03:20,267 raindrops. 67 00:03:20,267 --> 00:03:23,637 In contrast, over the oceans and in the tropics, they tend to be 68 00:03:23,637 --> 00:03:27,040 smaller and the reason is because smaller raindrops tend 69 00:03:27,040 --> 00:03:30,877 to originate from clouds that don't have any ice in them. 70 00:03:30,877 --> 00:03:33,613 It's worth noting, however, these measurements are only a 71 00:03:33,613 --> 00:03:37,117 small part of the equation. The drop size distribution is one of 72 00:03:37,117 --> 00:03:40,754 many factors that determines how big a storm will grow, how long 73 00:03:40,754 --> 00:03:43,690 it will last and how much rain it will ultimately produce. 74 00:03:43,690 --> 00:03:46,560 As GPM improves our understanding of precipitation 75 00:03:46,560 --> 00:03:49,963 from space, that information will be vital in improving 76 00:03:49,963 --> 00:03:52,966 weather models and thus helping us better predict and prepare 77 00:03:52,966 --> 00:03:53,600 for our weather.