1
00:00:04,100 --> 00:00:06,333
- I'd like to introduce 
to you Rick Anthes 
who is a former student 
here going way back

2
00:00:06,433 --> 00:00:10,500
into 1962 when he was 
an undergraduate student

3
00:00:10,600 --> 00:00:12,400
from out of state, I believe.

4
00:00:12,500 --> 00:00:15,633
Is that right? 
You came from Virginia.

5
00:00:15,733 --> 00:00:19,966
And then he went on to a 
masters degree and a PhD.

6
00:00:20,066 --> 00:00:23,466
His talk today is gonna be 
more of a general topic

7
00:00:23,566 --> 00:00:26,233
just about philosophy and 
modeling and what we can do

8
00:00:26,333 --> 00:00:28,200
with weather prediction overall,

9
00:00:28,300 --> 00:00:31,066
called "Demons and Butterflies," 
so here's Rick.

10
00:00:31,166 --> 00:00:34,166
(applause)

11
00:00:38,933 --> 00:00:41,666
- So without further ado, and 
you'll see why in a minute,

12
00:00:41,766 --> 00:00:46,433
I'm going to get started, 
and with a prologue.

13
00:00:47,433 --> 00:00:50,300
Actually more than half 
of the slides are not even

14
00:00:50,400 --> 00:00:51,833
directly related to my talk.

15
00:00:51,933 --> 00:00:53,666
This is kind of interesting.

16
00:00:53,766 --> 00:00:58,300
The prologue is very long, 
and the title of the prologue

17
00:00:58,400 --> 00:01:04,533
is "Sandy, Harvey, Irma and 
Maria: A Photographic Essay."

18
00:01:04,633 --> 00:01:07,966
And I hope this is relevant 
from a societal point of view

19
00:01:08,066 --> 00:01:12,600
about what the real talk 
is going to be about.

20
00:01:12,700 --> 00:01:15,600
So hurricanes, you've 
seen many pictures.

21
00:01:15,700 --> 00:01:18,800
They're absolutely 
beautiful from space.

22
00:01:18,900 --> 00:01:21,533
They're just majestic, 
artistic almost.

23
00:01:21,633 --> 00:01:24,000
They're just amazing to see.

24
00:01:24,100 --> 00:01:26,533
And so many people, this 
is one of the reasons

25
00:01:26,633 --> 00:01:28,000
I got interested in hurricanes

26
00:01:28,100 --> 00:01:31,033
because from a distance 
they're just wonderful.

27
00:01:33,100 --> 00:01:35,300
But when you actually 
experience one,

28
00:01:35,400 --> 00:01:39,100
it's anything but beautiful. 
Hell on Earth.

29
00:01:40,166 --> 00:01:41,933
And I'll start out with Sandy.

30
00:01:42,033 --> 00:01:46,200
This is a split image of New 
York City on a typical night

31
00:01:47,300 --> 00:01:51,800
at top, and the 
night after Sandy hit

32
00:01:51,900 --> 00:01:54,333
in the lower half where almost

33
00:01:54,433 --> 00:01:57,300
all the city is without power.

34
00:01:58,333 --> 00:02:00,066
So now I won't say anything

35
00:02:00,166 --> 00:02:02,433
for the next hundred 
slides or so.

36
00:02:06,766 --> 00:02:09,600
(somber music)

37
00:02:09,700 --> 00:05:39,600
♪

38
00:05:39,700 --> 00:05:43,100
And then, I'm trying to 
inject a tiny bit of humor

39
00:05:43,200 --> 00:05:44,500
at the end of this.

40
00:05:44,600 --> 00:05:46,700
There's Hurricane Nate 
where the chief damage

41
00:05:46,800 --> 00:05:50,100
was blowing a bunch of 
pumpkins out of a field.

42
00:05:51,200 --> 00:05:54,833
But anyway these are picture 
stories that I don't think

43
00:05:54,933 --> 00:05:57,033
I get from the news.

44
00:05:59,166 --> 00:06:02,466
You hear of 27 deaths and 
a billion dollars of damage

45
00:06:02,566 --> 00:06:06,133
and this kinda thing and 
it doesn't until you see

46
00:06:06,233 --> 00:06:08,733
the variety of people who 
are affected by these things

47
00:06:08,833 --> 00:06:11,566
and the amount of time it 
takes to rebuild their lives.

48
00:06:14,466 --> 00:06:16,500
So what about Hurricane Sandy?

49
00:06:16,600 --> 00:06:19,966
And I was inspired to do 
this talk right after Sandy

50
00:06:20,066 --> 00:06:21,833
which was 2012.

51
00:06:21,933 --> 00:06:25,100
It formed in the western 
Caribbean late in the year,

52
00:06:25,200 --> 00:06:29,266
October 22nd, almost 
this time of year.

53
00:06:29,366 --> 00:06:33,700
It made landfall in New Jersey 
on October 29th, very late.

54
00:06:33,800 --> 00:06:36,566
The largest Atlantic 
hurricane on record.

55
00:06:36,666 --> 00:06:38,133
And these are statistics.

56
00:06:38,233 --> 00:06:39,766
They just don't tell 
the story you saw.

57
00:06:39,866 --> 00:06:44,900
53 people killed, 
$32 billion in damage

58
00:06:45,000 --> 00:06:47,566
which is about 1/3 of 
the government sequester

59
00:06:47,666 --> 00:06:49,400
at that point.

60
00:06:49,500 --> 00:06:54,700
So how good were the forecasts 
of Hurricane Sandy?

61
00:06:54,800 --> 00:06:58,100
They were superb, 
they were excellent,

62
00:06:58,200 --> 00:07:02,366
they were unbelievable, 
and it wasn't by chance.

63
00:07:03,566 --> 00:07:05,600
How many more lives 
would have been lost

64
00:07:05,700 --> 00:07:07,366
without these 
excellent forecasts?

65
00:07:07,466 --> 00:07:08,833
You saw the damage 
that was done,

66
00:07:08,933 --> 00:07:12,800
and it's a remarkable thing 
that only 53 people were killed.

67
00:07:12,900 --> 00:07:16,600
Probably thousands would have 
been killed if that storm

68
00:07:16,700 --> 00:07:18,766
had come in unannounced.

69
00:07:20,200 --> 00:07:22,933
And imagine if people 
weren't prepared for it

70
00:07:23,033 --> 00:07:25,066
what the loss of life would be.

71
00:07:25,166 --> 00:07:29,233
Never before had a hurricane 
approached the East Coast

72
00:07:29,333 --> 00:07:32,166
from the east in late October.

73
00:07:33,333 --> 00:07:37,233
Came in from the east. 
Never before.

74
00:07:37,333 --> 00:07:41,400
Here is a set of tracks, 
the historical tracks,

75
00:07:41,500 --> 00:07:44,466
of hurricanes that came 
within 200 nautical miles

76
00:07:44,566 --> 00:07:47,600
of New York City

77
00:07:47,700 --> 00:07:50,400
from 1851 to 2011,

78
00:07:50,500 --> 00:07:54,366
the entire record up until 2012.

79
00:07:54,466 --> 00:07:56,500
You see every one 
of those tracks,

80
00:07:56,600 --> 00:08:00,666
by the time the storm hit 
past the Virginia capes

81
00:08:00,766 --> 00:08:03,366
was heading off to the 
north and northeast.

82
00:08:03,466 --> 00:08:09,466
The Hurricane Sandy 
came up the coast

83
00:08:09,566 --> 00:08:14,400
and, if you were a forecaster 
who had no satellite

84
00:08:14,500 --> 00:08:19,100
observations or models 
and you saw that track,

85
00:08:19,200 --> 00:08:22,500
everybody would have forecast 
continuing moving out to sea.

86
00:08:22,600 --> 00:08:25,266
Instead it did something that 
no storm had ever done before.

87
00:08:25,366 --> 00:08:27,266
It turned to the 
left and came in

88
00:08:27,366 --> 00:08:30,433
and made landfall from the east.

89
00:08:30,533 --> 00:08:34,600
Never ever before happening, 
and yet it was well predicted.

90
00:08:34,700 --> 00:08:36,200
How can that be?

91
00:08:39,500 --> 00:08:41,133
Here's a forecast

92
00:08:41,233 --> 00:08:43,533
from the European Centre for 
Medium-Range Weather Forecasting

93
00:08:43,633 --> 00:08:46,000
9 1/2 days before landfall.

94
00:08:46,100 --> 00:08:48,600
And on the left you can 
see that their outlook

95
00:08:48,700 --> 00:08:50,666
nearly 10 days 
before landfall

96
00:08:50,766 --> 00:08:58,500
had this grayish area 
off the Atlantic Coast

97
00:08:58,600 --> 00:09:02,200
where already they 
were forecasting,

98
00:09:02,300 --> 00:09:04,900
some of their models were 
forecasting major events,

99
00:09:05,000 --> 00:09:07,933
significant probability 
of the severe windstorm

100
00:09:08,033 --> 00:09:10,066
affecting the Northeastern 
United States.

101
00:09:10,166 --> 00:09:12,866
Then by the time 
three days later,

102
00:09:12,966 --> 00:09:16,033
6 1/2 days before landfall, 
the various models

103
00:09:16,133 --> 00:09:19,466
were forecasting those 
tracks that you see

104
00:09:19,566 --> 00:09:21,433
in the middle panel.

105
00:09:21,533 --> 00:09:24,933
And already almost a 
week before landfall,

106
00:09:25,033 --> 00:09:27,400
the models were 
forecasting this left turn,

107
00:09:27,500 --> 00:09:29,500
which again had never 
happened before.

108
00:09:29,600 --> 00:09:32,866
So this is not an empirical 
model based on past data.

109
00:09:32,966 --> 00:09:36,100
It is a model based 
on laws of physics

110
00:09:36,200 --> 00:09:38,100
and mathematics 
and observations.

111
00:09:38,200 --> 00:09:41,966
And then the observed track 
is shown on the right panel.

112
00:09:42,066 --> 00:09:43,433
So extremely well forecast.

113
00:09:43,533 --> 00:09:47,066
People had many days of warnings

114
00:09:47,166 --> 00:09:49,766
and were well prepared.

115
00:09:54,100 --> 00:09:56,366
A lot of people say, "Well, we 
got through that disaster.

116
00:09:56,466 --> 00:09:59,066
"It'll never happen 
again in my lifetime

117
00:09:59,166 --> 00:10:01,066
"and it can't 
happen to me again."

118
00:10:01,166 --> 00:10:03,066
Well, this is this year.

119
00:10:03,166 --> 00:10:07,033
Jump forward nine years to 2017,

120
00:10:07,133 --> 00:10:09,066
or five years I guess it is.

121
00:10:09,166 --> 00:10:12,333
Through September there 
have been 15 separate

122
00:10:12,433 --> 00:10:15,866
$1 billion weather 
and climate disasters

123
00:10:15,966 --> 00:10:17,666
just through September.

124
00:10:17,766 --> 00:10:19,766
So we're on a track 
for a record year.

125
00:10:19,866 --> 00:10:21,400
And you see they're 
all over the country.

126
00:10:21,500 --> 00:10:23,133
There's hurricanes, 
there's tornado outbreaks,

127
00:10:23,233 --> 00:10:27,133
there's fires, all 
kinds of things.

128
00:10:27,233 --> 00:10:31,933
And it's just no question 
this is not just anecdotal.

129
00:10:32,033 --> 00:10:35,366
There's no question that 
the frequency of severe

130
00:10:35,466 --> 00:10:40,033
weather and climate 
events is getting higher.

131
00:10:40,133 --> 00:10:41,666
And why is that?

132
00:10:41,766 --> 00:10:44,800
Well, this is not a 
talk on global warming,

133
00:10:44,900 --> 00:10:48,266
but, "It is global 
warming, stupid."

134
00:10:48,366 --> 00:10:51,133
And that's all I'll 
say about that.

135
00:10:51,233 --> 00:10:53,700
So the lecture is 
"Demons and Butterflies,"

136
00:10:53,800 --> 00:10:55,366
but I'm trying to 
set the context

137
00:10:55,466 --> 00:10:58,566
for something that's important.

138
00:10:59,566 --> 00:11:00,900
What is this?

139
00:11:01,000 --> 00:11:03,166
It's a fortuneteller. 
It's a wizard.

140
00:11:03,266 --> 00:11:05,733
Somebody that tells the future.

141
00:11:08,733 --> 00:11:11,333
Amazing people would 
believe this guy

142
00:11:11,433 --> 00:11:13,166
before they believe science.

143
00:11:15,166 --> 00:11:18,266
But foretelling the 
future has always been

144
00:11:18,366 --> 00:11:20,433
a fascination of humanity,

145
00:11:20,533 --> 00:11:22,433
and prophets over the ages

146
00:11:22,533 --> 00:11:24,766
have been worshiped 
and vilified.

147
00:11:24,866 --> 00:11:27,300
It's said that-- You've already 
heard this joke I'm sure.

148
00:11:27,400 --> 00:11:29,100
A meteorologist says 
weather forecasters

149
00:11:29,200 --> 00:11:32,133
are the second oldest 
profession in the world.

150
00:11:32,233 --> 00:11:34,833
People want to know what's 
gonna happen in the future.

151
00:11:35,933 --> 00:11:38,666
So foretelling the future 
has always been a fascination

152
00:11:38,766 --> 00:11:40,533
whether it's forecasting 
the stock market

153
00:11:40,633 --> 00:11:44,066
or forecasting a football game 
results or whatever it is,

154
00:11:44,166 --> 00:11:47,100
people love to talk about 
forecasting the future.

155
00:11:48,666 --> 00:11:50,933
And you see it in these 
common expressions.

156
00:11:51,033 --> 00:11:54,166
I should have known. 
I should have seen it coming.

157
00:11:56,233 --> 00:11:58,566
In retrospect, it was obvious.

158
00:12:00,366 --> 00:12:01,766
20-20 hindsight.

159
00:12:04,533 --> 00:12:07,533
The signs were there 
for all to see.

160
00:12:08,633 --> 00:12:10,366
Sixth sense.

161
00:12:12,633 --> 00:12:14,233
Premonition.

162
00:12:16,566 --> 00:12:19,300
And in today already 
walks tomorrow.

163
00:12:20,900 --> 00:12:23,133
The present is big 
with the future.

164
00:12:24,333 --> 00:12:28,133
Good detectives such as Sherlock 
Holmes and Hercule Poirot

165
00:12:28,233 --> 00:12:30,700
deduce what has happened and 
sometimes what will happen

166
00:12:30,800 --> 00:12:32,666
from a few observations.

167
00:12:36,533 --> 00:12:39,700
Foretelling the future can 
be based on past behavior,

168
00:12:39,800 --> 00:12:43,333
empiricism, or the natural 
laws of mathematics,

169
00:12:43,433 --> 00:12:45,233
physics, and chemistry.

170
00:12:45,333 --> 00:12:48,633
But all predictions, 
one way or another,

171
00:12:48,733 --> 00:12:51,300
are based on observations.

172
00:12:51,400 --> 00:12:55,100
Whether you're a fortuneteller

173
00:12:55,200 --> 00:12:59,266
or a mathematical 
modeler of hurricanes,

174
00:12:59,366 --> 00:13:02,900
you're using observations 
one way or another.

175
00:13:04,066 --> 00:13:07,333
Well, the philosophy 
of forecasts

176
00:13:07,433 --> 00:13:10,233
goes back many years

177
00:13:10,333 --> 00:13:12,566
and Gottfried Leibniz,

178
00:13:12,666 --> 00:13:17,033
famous for Leibniz's rule in 
mathematics, if you know that.

179
00:13:17,133 --> 00:13:19,566
I think we all learned that 
in our calculus courses.

180
00:13:19,666 --> 00:13:22,866
1646 to 1716.

181
00:13:22,966 --> 00:13:24,900
A very interesting quote.

182
00:13:25,833 --> 00:13:28,333
"Everything proceeds 
mathematically.

183
00:13:28,433 --> 00:13:31,366
"If someone could have 
sufficient insight

184
00:13:31,466 --> 00:13:34,533
"into the inner parts of 
things, and in addition

185
00:13:34,633 --> 00:13:37,700
"had remembrance and 
intelligence enough to consider

186
00:13:37,800 --> 00:13:41,800
"all of the circumstances 
and take them into account,

187
00:13:41,900 --> 00:13:45,666
"he would be a prophet and see 
the future in the present

188
00:13:45,766 --> 00:13:47,566
"as in a mirror."

189
00:13:47,666 --> 00:13:50,533
So read that carefully.

190
00:13:50,633 --> 00:13:53,766
You have to have insight, 
remembrance, intelligence,

191
00:13:53,866 --> 00:13:55,533
and to consider everything.

192
00:13:55,633 --> 00:13:59,700
This is foreseeing models in 
a way, very complex systems.

193
00:13:59,800 --> 00:14:01,733
If you could 
understand everything,

194
00:14:01,833 --> 00:14:03,300
you could predict everything.

195
00:14:04,933 --> 00:14:08,800
And even more direct,

196
00:14:08,900 --> 00:14:11,233
the Marquis de Laplace.

197
00:14:11,333 --> 00:14:14,166
You know about 
de Laplacians, right,

198
00:14:14,266 --> 00:14:17,166
Laplace in mathematics. 
He was a mathematician.

199
00:14:17,266 --> 00:14:21,166
He dreamed of an intelligent 
being, an intellect,

200
00:14:21,266 --> 00:14:24,700
which was later dubbed, 
I guess by his colleagues,

201
00:14:24,800 --> 00:14:28,666
"Laplace's Demon," 
who knew the positions.

202
00:14:29,800 --> 00:14:33,066
He dreamed of an intelligent 
being who knew the positions

203
00:14:33,166 --> 00:14:37,333
and velocities of every 
single atom in the universe.

204
00:14:38,333 --> 00:14:41,700
And using Newton's 
equations of motion,

205
00:14:41,800 --> 00:14:45,266
he could predict the motion 
of each one of these atoms,

206
00:14:45,366 --> 00:14:48,100
all the molecules that 
the atoms are part of.

207
00:14:48,200 --> 00:14:52,933
They didn't know about smaller, 
subatomic atoms at that time,

208
00:14:53,033 --> 00:14:56,266
but predict the future 
of the entire universe.

209
00:14:57,566 --> 00:15:01,200
And this long quote 
at the end is actually

210
00:15:01,300 --> 00:15:05,166
very prescient in 
terms of the theory

211
00:15:05,266 --> 00:15:07,833
behind developing numerical 
weather prediction models.

212
00:15:07,933 --> 00:15:09,866
We may regard the present 
state of the universe

213
00:15:09,966 --> 00:15:13,366
as the effect of its past 
and the cause of its future,

214
00:15:13,466 --> 00:15:16,700
an intellect, which at any 
given moment knew all the forces

215
00:15:16,800 --> 00:15:19,366
that animate nature and 
the mutual positions

216
00:15:19,466 --> 00:15:21,266
of the beings that compose it.

217
00:15:21,366 --> 00:15:24,733
If this intellect, 
think supercomputer,

218
00:15:24,833 --> 00:15:27,733
were vast enough to submit 
the data to analysis,

219
00:15:27,833 --> 00:15:29,966
could condense into a 
single formula to move

220
00:15:30,066 --> 00:15:32,266
one of the greatest 
bodies of the universe

221
00:15:32,366 --> 00:15:34,200
and that of the lightest atom,

222
00:15:34,300 --> 00:15:37,033
for such an intellect, 
nothing could be uncertain,

223
00:15:37,133 --> 00:15:39,466
and the future just like 
the past would be present

224
00:15:39,566 --> 00:15:41,066
before its eyes.

225
00:15:42,066 --> 00:15:45,433
The condition of every 
one of us in the room,

226
00:15:45,533 --> 00:15:51,166
every molecule, 
every wave out there,

227
00:15:51,266 --> 00:15:53,800
and all around the world, 
including the molecules

228
00:15:53,900 --> 00:15:59,400
in life itself, if you knew 
exactly where they were today,

229
00:15:59,500 --> 00:16:01,600
according to Laplace, 
you could predict

230
00:16:01,700 --> 00:16:03,266
everything in the future,

231
00:16:03,366 --> 00:16:05,100
how humans would behave, 
when they would die,

232
00:16:05,200 --> 00:16:07,200
how many children 
they would have,

233
00:16:07,300 --> 00:16:08,666
how many children the 
children would have,

234
00:16:08,766 --> 00:16:10,333
and so on and so on.

235
00:16:10,433 --> 00:16:13,400
A perfectly deterministic 
system if you knew

236
00:16:13,500 --> 00:16:16,366
where everything was and 
you knew all of the laws

237
00:16:16,466 --> 00:16:18,200
that we follow.

238
00:16:20,166 --> 00:16:21,666
That was Laplace's view.

239
00:16:22,666 --> 00:16:26,533
And then Niels Bohr, 
the famous physicist,

240
00:16:26,633 --> 00:16:30,200
had a much simpler 
statement which sounds to me

241
00:16:30,300 --> 00:16:34,233
like Yogi Berra, more like 
Yogi Berra than Niels Bohr.

242
00:16:35,466 --> 00:16:37,600
"Prediction is difficult, 
especially the future."

243
00:16:37,700 --> 00:16:40,333
True, I think we can 
all agree with that

244
00:16:40,433 --> 00:16:44,366
even though we may question 
Laplace's Demon a little bit.

245
00:16:46,733 --> 00:16:49,700
Well, Bjerknes, Vilhelm 
Bjerknes, getting into our field

246
00:16:49,800 --> 00:16:53,533
in 1904, the father of 
the Norwegian school

247
00:16:53,633 --> 00:16:58,033
of weather prediction, 
said the following.

248
00:16:58,133 --> 00:17:00,366
"If it is true, as any 
scientist believes,

249
00:17:00,466 --> 00:17:02,733
"that subsequent states 
of the atmosphere develop

250
00:17:02,833 --> 00:17:05,266
"from preceding ones 
according to physical laws,

251
00:17:05,366 --> 00:17:07,833
"one will agree that 
the necessary and 
sufficient conditions

252
00:17:07,933 --> 00:17:11,033
"for a rational solution of 
the problem of meteorological

253
00:17:11,133 --> 00:17:12,933
"prediction are the following.

254
00:17:13,033 --> 00:17:15,133
"Number one, one has to know 
with sufficient accuracy

255
00:17:15,233 --> 00:17:17,666
"the state of the 
atmosphere at a given time."

256
00:17:17,766 --> 00:17:19,366
Those are the observations, and,

257
00:17:19,466 --> 00:17:22,700
"One has to know with 
sufficient accuracy the laws

258
00:17:22,800 --> 00:17:24,766
"according to which one 
state of the atmosphere

259
00:17:24,866 --> 00:17:26,266
"develops from another."

260
00:17:26,366 --> 00:17:29,766
That's the mathematics and 
physics of how motion reacts

261
00:17:29,866 --> 00:17:34,933
to forces at a given time.

262
00:17:35,033 --> 00:17:37,400
And this is definitely the basis

263
00:17:37,500 --> 00:17:39,333
for numerical 
weather prediction.

264
00:17:40,966 --> 00:17:45,233
Well, then some 50 
years later, along comes

265
00:17:45,333 --> 00:17:48,633
brilliant mathematician, 
actually a meteorologist,

266
00:17:48,733 --> 00:17:51,766
who became brilliant in 
the field of mathematics,

267
00:17:51,866 --> 00:17:54,833
one of the few ones that 
ever did this, Ed Lorenz

268
00:17:54,933 --> 00:17:59,000
from MIT, got into chaos theory

269
00:17:59,100 --> 00:18:03,900
and is alleged to have 
said, at least interpreted

270
00:18:04,000 --> 00:18:06,366
to have said, "Does the 
flap of a butterfly's wing

271
00:18:06,466 --> 00:18:10,366
"in Brazil set off 
a tornado in Texas?"

272
00:18:10,466 --> 00:18:14,533
This is now in popular 
mythology, in popular speeches,

273
00:18:14,633 --> 00:18:17,500
is the butterfly effect, 
and it's the idea

274
00:18:17,600 --> 00:18:20,333
that you can never measure 
everything to a sufficient

275
00:18:20,433 --> 00:18:24,066
accuracy to make a 
good prediction, a 
perfect prediction.

276
00:18:24,166 --> 00:18:26,466
There's always gonna be 
a butterfly somewhere

277
00:18:26,566 --> 00:18:28,800
that you don't know, 
you can't follow,

278
00:18:28,900 --> 00:18:31,500
and the butterfly flaps 
its wings and that sets off

279
00:18:31,600 --> 00:18:34,900
a cascade of events 
that lead to something

280
00:18:35,000 --> 00:18:38,933
as severe as a tornado in Texas 
or a hurricane in New Jersey.

281
00:18:43,000 --> 00:18:46,900
In the '70s, Greg was talking 
about why I was developing

282
00:18:47,000 --> 00:18:50,500
numerical models and he had 
the so-called Mesoscale,

283
00:18:50,600 --> 00:18:52,900
and a lot of the larger 
scale dynamists said,

284
00:18:53,000 --> 00:18:56,400
"that you're wasting your time 
because the smaller scales

285
00:18:56,500 --> 00:18:58,600
"of motion are never 
gonna be predictable,

286
00:18:58,700 --> 00:19:01,566
"and why are you trying 
to do Mesoscale models?"

287
00:19:01,666 --> 00:19:03,833
So I was trying to think 
of a rebuttal to this

288
00:19:03,933 --> 00:19:07,633
and I came up with the idea 
that in many synoptic situations

289
00:19:07,733 --> 00:19:10,066
large scale situations, 
the small scales are forced

290
00:19:10,166 --> 00:19:11,533
by the larger scales.

291
00:19:11,633 --> 00:19:14,900
So if you know the large-scale 
waves, they produce fronts

292
00:19:15,000 --> 00:19:17,366
in more or less the right 
place, smaller scale events.

293
00:19:17,466 --> 00:19:20,800
They produce areas favorable 
for convection and so forth,

294
00:19:20,900 --> 00:19:24,066
so that if you know the large 
scale initial conditions

295
00:19:24,166 --> 00:19:27,033
and you can predict them, 
they will lead to

296
00:19:27,133 --> 00:19:32,166
small-scale phenomena, 
create small-scale phenomena,

297
00:19:32,266 --> 00:19:35,466
even before the small- 
scale phenomena exist.

298
00:19:35,566 --> 00:19:38,333
And so you actually 
see that today.

299
00:19:38,433 --> 00:19:40,966
People are forecasting 
tornado outbreaks

300
00:19:41,066 --> 00:19:45,133
three or four days before 
tornadoes actually occur,

301
00:19:45,233 --> 00:19:47,666
even start to occur, 
because the large scale

302
00:19:47,766 --> 00:19:49,400
has predicted well 
and it predicts

303
00:19:49,500 --> 00:19:52,566
the environment of tornadoes.

304
00:19:52,666 --> 00:19:56,733
So, that was my argument 
for the fact that there

305
00:19:56,833 --> 00:20:00,233
was predictability in the 
smaller scales of motion,

306
00:20:00,333 --> 00:20:02,000
which according to 
predictability theories

307
00:20:02,100 --> 00:20:05,500
should be less predictable 
than for the very large scales.

308
00:20:05,600 --> 00:20:09,666
Anyway, if we look at some 
real data in this case.

309
00:20:09,766 --> 00:20:13,900
These are forecast accuracies 
of the European Centre

310
00:20:14,000 --> 00:20:17,733
for Medium-Range Weather 
Forecasting, ECMWF.

311
00:20:17,833 --> 00:20:21,066
Since 1981, has had the best

312
00:20:21,166 --> 00:20:23,566
global prediction 
model in the world.

313
00:20:23,666 --> 00:20:25,766
And the United States has 
tried desperately to catch up

314
00:20:25,866 --> 00:20:29,466
to this model, but has 
always lagged behind it

315
00:20:29,566 --> 00:20:31,900
by about a half a day's 
worth of forecast.

316
00:20:32,000 --> 00:20:35,566
What this shows is the-- 
You don't need to understand

317
00:20:35,666 --> 00:20:39,000
what the skill scores are, 
but 100 at the top

318
00:20:39,100 --> 00:20:42,800
would be a perfect forecast, 
and 30 would be like

319
00:20:42,900 --> 00:20:46,133
a correlation coefficient of 
30, wouldn't be much value,

320
00:20:46,233 --> 00:20:48,333
but still some 
value over guessing.

321
00:20:48,433 --> 00:20:52,233
And the colors are different 
links to the forecast.

322
00:20:52,333 --> 00:20:58,000
So the blue envelope is 
I guess day three forecast.

323
00:20:58,100 --> 00:21:01,233
So the day three forecast 
tend to be very accurate

324
00:21:01,333 --> 00:21:04,033
and they've been increasing 
in accuracy with time

325
00:21:04,133 --> 00:21:08,200
going from about 87% in 
the Northern Hemisphere

326
00:21:08,300 --> 00:21:13,500
in 1981, to nearly 96% 
in recent years.

327
00:21:13,600 --> 00:21:15,166
And they've leveled off.

328
00:21:15,266 --> 00:21:18,933
They aren't getting much 
better, because apparently

329
00:21:19,033 --> 00:21:21,300
the 3-day forecast 
is close to as good

330
00:21:21,400 --> 00:21:22,800
as it's ever gonna get.

331
00:21:22,900 --> 00:21:25,400
The lower part of that 
green envelope at the top

332
00:21:25,500 --> 00:21:27,666
is the Southern 
Hemisphere forecast.

333
00:21:27,766 --> 00:21:30,600
And you can see that it 
was much worse in 1981

334
00:21:30,700 --> 00:21:33,000
than the Northern Hemisphere.

335
00:21:33,100 --> 00:21:35,200
And why is that? It's because 
the Southern Hemisphere

336
00:21:35,300 --> 00:21:36,833
doesn't have as much data.

337
00:21:36,933 --> 00:21:39,800
There are a lot more oceans 
and they don't have

338
00:21:39,900 --> 00:21:41,566
as many balloons and so forth.

339
00:21:41,666 --> 00:21:46,566
But that gap is closed, until 
today it's almost nonexistent

340
00:21:46,666 --> 00:21:48,300
and that's because 
of satellites.

341
00:21:48,400 --> 00:21:49,833
Satellites are global.

342
00:21:49,933 --> 00:21:52,700
They measure globally, and 
so the Southern Hemisphere

343
00:21:52,800 --> 00:21:56,166
gets just as good observations 
as the Northern Hemisphere now.

344
00:21:56,266 --> 00:21:59,100
And this is a dramatic 
testament to the power

345
00:21:59,200 --> 00:22:02,566
of global satellites of 
which Wisconsin, of course,

346
00:22:02,666 --> 00:22:08,266
is the leader in the world 
in satellite meteorology.

347
00:22:08,366 --> 00:22:11,166
The red curves are like 
five-day forecasts

348
00:22:11,266 --> 00:22:13,200
and the green curves are 
seven days or whatever.

349
00:22:13,300 --> 00:22:15,966
It doesn't matter, but the 
forecasts are getting better

350
00:22:16,066 --> 00:22:17,566
at all-time scales.

351
00:22:17,666 --> 00:22:20,633
And the gap between 
the Northern Hemisphere

352
00:22:20,733 --> 00:22:23,333
at the top of each of the bands 
and the Southern Hemisphere

353
00:22:23,433 --> 00:22:26,533
is diminishing because 
of global observations.

354
00:22:26,633 --> 00:22:32,066
Absolute truth, positive 
truth of the impact of

355
00:22:32,166 --> 00:22:36,233
global models and 
satellite observations.

356
00:22:36,333 --> 00:22:40,733
So another one of the reasons 
I'm bringing up the resolution

357
00:22:40,833 --> 00:22:43,766
of these global 
models one at a time

358
00:22:43,866 --> 00:22:48,266
roughly at the same time 
scale as the bottom,

359
00:22:48,366 --> 00:22:51,800
so in 1981 the European Centre 
model had a horizontal

360
00:22:51,900 --> 00:22:55,400
spacing between data points 
of about 200 kilometers,

361
00:22:55,500 --> 00:22:56,666
100 some miles.

362
00:22:56,766 --> 00:22:59,533
And what you're seeing 
here is the resolution

363
00:22:59,633 --> 00:23:03,033
getting finer and finer, 
the pixel size getting 
smaller and smaller.

364
00:23:03,133 --> 00:23:05,633
The model is resolving 
finer and finer scales.

365
00:23:05,733 --> 00:23:07,766
And that's the picture 
of a hurricane.

366
00:23:07,866 --> 00:23:13,933
When it gets down to 16 
kilometers and the last one is

367
00:23:14,033 --> 00:23:18,433
10 kilometers, you can 
actually see Hurricane Katrina.

368
00:23:18,533 --> 00:23:20,333
But that's what Hurricane 
Katrina looks like

369
00:23:20,433 --> 00:23:22,633
at 250 kilometers 
on the far left.

370
00:23:22,733 --> 00:23:24,666
It's just a blur, a smudge.

371
00:23:24,766 --> 00:23:28,133
So you can't even resolve 
hurricanes back in 1981

372
00:23:28,233 --> 00:23:31,566
with these models, 
but currently you can,

373
00:23:31,666 --> 00:23:35,733
and that's a testament to 
the value of computer power

374
00:23:35,833 --> 00:23:39,433
and good models in addition 
to the good observations.

375
00:23:39,533 --> 00:23:42,000
So the predictions are 
getting better all the time.

376
00:23:43,333 --> 00:23:47,400
The next slide shows a 
different way of judging

377
00:23:47,500 --> 00:23:49,200
how good models are.

378
00:23:49,300 --> 00:23:54,833
This goes back to 2008

379
00:23:54,933 --> 00:23:57,533
technology of 
geostationary satellites.

380
00:23:57,633 --> 00:24:00,433
It's the Meteosat observations.

381
00:24:00,533 --> 00:24:04,100
And if I didn't have 
these things labeled,

382
00:24:04,200 --> 00:24:08,600
even in 2008, you 
have trouble telling

383
00:24:08,700 --> 00:24:12,833
which is the model and 
which is the satellite,

384
00:24:12,933 --> 00:24:15,433
if you didn't have 
these labeled, right?

385
00:24:15,533 --> 00:24:18,366
And you could probably tell if 
you stared at it long enough,

386
00:24:18,466 --> 00:24:21,333
if you were an expert, 
if you were either an 
expert in the models

387
00:24:21,433 --> 00:24:23,833
or in the geostationary 
satellite imagery,

388
00:24:23,933 --> 00:24:27,700
but just looking at it, the 
casual person is gonna say

389
00:24:27,800 --> 00:24:30,500
that model on the 
right is damn good,

390
00:24:30,600 --> 00:24:32,566
even without looking at numbers.

391
00:24:33,566 --> 00:24:36,366
And so you know the model 
is doing something right.

392
00:24:36,466 --> 00:24:41,833
And we were looking 
at one of the

393
00:24:41,933 --> 00:24:45,466
tornado models this 
morning or this afternoon.

394
00:24:45,566 --> 00:24:48,000
In the tornado model, 
the clouds were so accurate,

395
00:24:48,100 --> 00:24:49,633
you just know that 
they're right,

396
00:24:49,733 --> 00:24:51,000
even without a lot of numbers.

397
00:24:51,100 --> 00:24:53,433
But this is actually one 
way of verifying models

398
00:24:53,533 --> 00:24:56,733
is to look at the 
pattern recognition.

399
00:24:56,833 --> 00:24:59,766
Humans are very good at 
seeing whether something 
looks good or not,

400
00:24:59,866 --> 00:25:02,333
whether it's right or not, 
and you can see it there.

401
00:25:02,433 --> 00:25:04,033
And that was quite 
a few years ago.

402
00:25:04,133 --> 00:25:10,800
The models in the geostationary 
satellite like GOES are,

403
00:25:10,900 --> 00:25:13,333
what do you call 
them now, GOES-16?

404
00:25:13,433 --> 00:25:14,800
(mumbled response)

405
00:25:14,900 --> 00:25:19,166
Yeah, is a much 
better resolution

406
00:25:19,266 --> 00:25:21,966
on the satellites and much 
better resolution on the models.

407
00:25:22,066 --> 00:25:24,733
So they still look 
very good together.

408
00:25:25,766 --> 00:25:27,366
But numbers are important.

409
00:25:27,466 --> 00:25:31,533
Here's a record of 
official hurricane

410
00:25:31,633 --> 00:25:35,100
track errors over 
time from 1970,

411
00:25:35,200 --> 00:25:37,933
before models and 
before satellites

412
00:25:38,033 --> 00:25:41,033
out to 2016, the 
latest data I had

413
00:25:41,133 --> 00:25:42,933
from the National 
Hurricane Center.

414
00:25:43,033 --> 00:25:47,100
And the track errors 
are in nautical miles,

415
00:25:47,200 --> 00:25:52,166
which are very close to miles, 
from zero to 700.

416
00:25:52,266 --> 00:25:54,300
In this case, a 
low number is good

417
00:25:54,400 --> 00:25:56,666
because the track error, the 
error, the position error

418
00:25:56,766 --> 00:25:58,966
of the storm is smaller.

419
00:25:59,066 --> 00:26:00,933
And so you can see that 
the different forecasts,

420
00:26:01,033 --> 00:26:04,933
the red curve is 24 hours, 
the green is 48 hours

421
00:26:05,033 --> 00:26:09,600
and so on up to 120 hours, 
the dark blue at the top.

422
00:26:09,700 --> 00:26:12,866
With some year-to-year 
variation, all of these

423
00:26:12,966 --> 00:26:16,433
official forecasts 
are getting better.

424
00:26:17,433 --> 00:26:19,700
And they're starting 
to maybe converge.

425
00:26:19,800 --> 00:26:21,533
Of course, you can't 
get any better than zero,

426
00:26:21,633 --> 00:26:24,966
so at least the one-day, the 
two-day, and the three-day

427
00:26:25,066 --> 00:26:27,266
forecasts are getting 
pretty darn good

428
00:26:27,366 --> 00:26:32,733
to less than a 50-mile error 
in position of the storm.

429
00:26:34,133 --> 00:26:37,866
So anecdotally the 
forecasts are getting better,

430
00:26:37,966 --> 00:26:40,266
like Hurricane Sandy.

431
00:26:40,366 --> 00:26:41,900
Statistically they're 
getting better.

432
00:26:42,000 --> 00:26:45,066
We know why, it's 
computer, it's models,

433
00:26:45,166 --> 00:26:47,233
it's satellite 
observations of all kinds.

434
00:26:47,333 --> 00:26:49,800
It's better physics, it's 
scientists working on this.

435
00:26:49,900 --> 00:26:51,600
This is not an accident.

436
00:26:52,600 --> 00:26:56,933
It's not because of 
political philosophy or

437
00:26:57,033 --> 00:26:58,800
the people are better 
or anything like that.

438
00:26:58,900 --> 00:27:05,833
It's pure science, physics, 
measurements, education.

439
00:27:05,933 --> 00:27:09,733
This is all something that 
we did, we as a community,

440
00:27:09,833 --> 00:27:13,500
as a university community, 
as government centers

441
00:27:13,600 --> 00:27:16,200
in Europe and the United 
States, we did this.

442
00:27:16,300 --> 00:27:19,933
This is not forgone, 
this is not an accident.

443
00:27:20,033 --> 00:27:22,966
It's the results of 
mathematicians and physicists

444
00:27:23,066 --> 00:27:25,366
and chemists and 
computer scientists

445
00:27:25,466 --> 00:27:28,166
and educated people and 
supportive graduate students

446
00:27:28,266 --> 00:27:31,533
and government grants 
doing all of this.

447
00:27:31,633 --> 00:27:33,333
And it's saving 
thousands of lives,

448
00:27:33,433 --> 00:27:35,466
maybe hundreds of 
thousands of lives,

449
00:27:35,566 --> 00:27:37,100
just in this one little area,

450
00:27:37,200 --> 00:27:38,833
this little tiny area 
of weather prediction,

451
00:27:38,933 --> 00:27:41,100
it's science and education.

452
00:27:41,200 --> 00:27:44,300
It's not philosophy, 
it's not praying to God.

453
00:27:44,400 --> 00:27:46,100
It's doing something about it.

454
00:27:46,200 --> 00:27:49,466
It's helping God by doing 
something for ourselves.

455
00:27:50,466 --> 00:27:53,100
And yet you have people 
that want to cut funding

456
00:27:53,200 --> 00:27:54,700
in these areas.

457
00:27:54,800 --> 00:27:58,566
The total NESDIS budget is 
about $2.1 billion a year,

458
00:27:58,666 --> 00:28:00,200
the satellite budget 
in the United States.

459
00:28:00,300 --> 00:28:04,233
$2.1 billion, it sounds 
like a lot, right?

460
00:28:04,333 --> 00:28:07,300
We've had 15 $1 billion 
disasters already

461
00:28:07,400 --> 00:28:09,266
in nine months of this year.

462
00:28:13,600 --> 00:28:16,066
And they, they want to cut it.

463
00:28:16,166 --> 00:28:18,733
I'm trying to design the 
satellite system for 2030

464
00:28:18,833 --> 00:28:20,566
and beyond, when I'll be dead.

465
00:28:20,666 --> 00:28:23,266
I still think it's important 
even though I'll be dead.

466
00:28:23,366 --> 00:28:25,900
I have children, I probably 
won't have grandchildren,

467
00:28:26,000 --> 00:28:27,666
but many of you 
have grandchildren.

468
00:28:27,766 --> 00:28:29,433
They're gonna have children.

469
00:28:30,800 --> 00:28:34,100
And we need to be 
preparing for the science

470
00:28:34,200 --> 00:28:36,100
and the forecasts 
of 2030 and beyond.

471
00:28:36,200 --> 00:28:37,566
That's what I'm trying to do.

472
00:28:37,666 --> 00:28:39,233
Let's get back to what 
might be possible.

473
00:28:39,333 --> 00:28:41,600
Let's get back to what 
some fun stuff is.

474
00:28:42,600 --> 00:28:44,833
All right, well this 
is an interesting thing

475
00:28:44,933 --> 00:28:48,933
that's actually fairly old 
technology in the modeling area

476
00:28:49,033 --> 00:28:51,333
and visualization 
compared to some of the 
things you can do now.

477
00:28:51,433 --> 00:28:54,966
But what it is, is a five-day 
forecast using a massive model,

478
00:28:55,066 --> 00:28:57,333
a very high resolution 
advanced model

479
00:28:57,433 --> 00:29:00,100
of Cyclone Nargis 
which is a major storm

480
00:29:00,200 --> 00:29:03,566
that developed in 
the Indian Ocean.

481
00:29:03,666 --> 00:29:06,800
And this shows the five-day 
forecast in this model

482
00:29:06,900 --> 00:29:08,666
of this genesis of this storm.

483
00:29:08,766 --> 00:29:11,833
It's kind of an interesting, 
beautiful depiction

484
00:29:11,933 --> 00:29:15,033
but these are basically the 
wind flow at different levels.

485
00:29:15,133 --> 00:29:18,233
The greenish colors 
are low level flow,

486
00:29:18,333 --> 00:29:21,466
and the reddish colors 
are the upper level jets,

487
00:29:21,566 --> 00:29:23,533
and so you see a 
kinda low level flow.

488
00:29:23,633 --> 00:29:26,300
You're looking at the 
Indian Ocean there.

489
00:29:26,400 --> 00:29:31,266
And you can see with 
time this vortex develops

490
00:29:31,366 --> 00:29:32,733
in the Indian Ocean.

491
00:29:32,833 --> 00:29:34,466
And there you can see it.

492
00:29:34,566 --> 00:29:37,966
That's Cyclone Nargis in 2008.

493
00:29:38,066 --> 00:29:40,366
You can see this developing

494
00:29:40,466 --> 00:29:43,700
with no hint of anything of that 
scale in the initial conditions.

495
00:29:43,800 --> 00:29:45,566
The large scale just did it.

496
00:29:46,600 --> 00:29:50,333
It was predictability 
of that tropical cyclone

497
00:29:50,433 --> 00:29:53,966
five days in advance 
by this global model

498
00:29:54,066 --> 00:29:58,500
in the right place and pretty 
much at the right time.

499
00:29:58,600 --> 00:30:01,700
Not exactly in the right place 
or exactly at the right time.

500
00:30:01,800 --> 00:30:06,233
And this is becoming or 
could become a routine.

501
00:30:06,333 --> 00:30:09,400
You saw this in Hurricane Sandy 
which is a real data case.

502
00:30:09,500 --> 00:30:11,266
So there's the cyclone 
well developed.

503
00:30:11,366 --> 00:30:14,366
You can see the low-level 
inflow and the outflow,

504
00:30:14,466 --> 00:30:18,833
global model and 
initialized with real data.

505
00:30:19,866 --> 00:30:22,500
So again, we know 
what we're doing.

506
00:30:23,533 --> 00:30:26,366
Okay, here's a 
climate model for,

507
00:30:26,466 --> 00:30:28,933
I say quote "September 2000"

508
00:30:29,033 --> 00:30:32,400
which might be September 
2500, 50 years from now,

509
00:30:32,500 --> 00:30:34,300
for an entire month.

510
00:30:34,400 --> 00:30:37,433
And again, does this 
pass the reality test?

511
00:30:37,533 --> 00:30:39,900
Do you see hurricanes 
forming in the Atlantic

512
00:30:40,000 --> 00:30:43,000
and moving toward Florida 
or toward New Orleans?

513
00:30:43,100 --> 00:30:46,166
Does this look realistic? 
Well, look at that.

514
00:30:46,266 --> 00:30:48,200
That looks exactly 
like Hurricane Katrina.

515
00:30:48,300 --> 00:30:51,300
There's more forming 
in the Atlantic.

516
00:30:51,400 --> 00:30:54,800
These large-scale models, 
with the righty physics

517
00:30:54,900 --> 00:30:59,133
and right ocean 
interactions, can produce

518
00:30:59,233 --> 00:31:03,166
hurricanes in the right 
place and time ontologically

519
00:31:03,266 --> 00:31:05,233
at least at the right seasons.

520
00:31:06,733 --> 00:31:10,133
Okay, so I am gonna 
start wrapping this up.

521
00:31:11,500 --> 00:31:15,733
Anyway, the summary is 
getting away from Laplace's

522
00:31:15,833 --> 00:31:18,500
abstraction and being able 
to predict everything

523
00:31:18,600 --> 00:31:22,166
all the time at all scales 
and every human's behavior

524
00:31:22,266 --> 00:31:24,366
and all their children's 
behavior and all that.

525
00:31:24,466 --> 00:31:28,200
There is evidence that you 
can have greatly improved

526
00:31:28,300 --> 00:31:31,433
forecasts of such severe 
weather as tropical cyclones.

527
00:31:31,533 --> 00:31:33,933
I hope you remember the 
prologue, but these are really

528
00:31:34,033 --> 00:31:37,566
high impact events, 
days in advance.

529
00:31:37,666 --> 00:31:40,833
And then the boring line

530
00:31:40,933 --> 00:31:43,500
but you've got to keep 
pounding this home

531
00:31:43,600 --> 00:31:48,200
to the politicians 
that this is not by accident.

532
00:31:48,300 --> 00:31:49,833
We need high resolution models,

533
00:31:49,933 --> 00:31:51,566
probably at four 
kilometers or better.

534
00:31:51,666 --> 00:31:54,000
We need improved physics, 
it means understanding.

535
00:31:54,100 --> 00:31:57,333
We need more PhD students 
working on these problems.

536
00:31:57,433 --> 00:31:59,900
And we need interactive 
ocean-atmosphere models.

537
00:32:00,000 --> 00:32:02,466
We need improved and initial 
conditions in the atmosphere

538
00:32:02,566 --> 00:32:05,966
temperature, water vapor, and 
winds-satellite observations

539
00:32:06,066 --> 00:32:07,966
are absolutely essential here.

540
00:32:08,066 --> 00:32:10,933
We need better data 
assimilation techniques.

541
00:32:11,033 --> 00:32:15,033
That's ways of using these 
strange forms of data

542
00:32:15,133 --> 00:32:18,200
from the satellites and 
we need faster computers.

543
00:32:18,300 --> 00:32:22,533
So again this is one slice 
and one aspect of society

544
00:32:22,633 --> 00:32:24,700
but we know how to do it,

545
00:32:24,800 --> 00:32:28,933
and we just don't have the will 
to do it, it seems sometimes.

546
00:32:30,066 --> 00:32:33,933
So back to the answer, 
the big picture.

547
00:32:34,033 --> 00:32:38,200
Who wins, the butterfly or 
the demon, Laplace's Demon?

548
00:32:39,400 --> 00:32:41,066
Well, there is a 
different between

549
00:32:41,166 --> 00:32:43,166
what is theoretically possible,

550
00:32:43,266 --> 00:32:44,933
that's what's called 
predictability,

551
00:32:45,033 --> 00:32:46,733
and what can ever 
actually be done,

552
00:32:46,833 --> 00:32:49,266
and that's actually predictions.

553
00:32:49,366 --> 00:32:52,033
The demon may be 
theoretically possible,

554
00:32:52,133 --> 00:32:54,866
and that's the question 
I think for philosophers.

555
00:32:55,966 --> 00:32:58,233
Not for us, because 
we'll never be there

556
00:32:58,333 --> 00:32:59,733
in a practical point of view.

557
00:32:59,833 --> 00:33:03,100
But the butterflies are 
ultimately going to win.

558
00:33:03,200 --> 00:33:07,266
And I see there's no reason 
not to help the demon a little,

559
00:33:07,366 --> 00:33:11,300
in such beneficial activities 
as weather prediction.

560
00:33:11,400 --> 00:33:15,533
And that's the end of my talk.

561
00:33:15,633 --> 00:33:17,066
Thank you very much.

562
00:33:17,166 --> 00:33:19,300
I'd be happy to take 
any questions or outrageous

563
00:33:19,400 --> 00:33:24,400
statements or challenges or 
denial or whatever

564
00:33:24,500 --> 00:33:26,233
alternative points of view.

565
00:33:26,333 --> 00:33:27,633
Thank you.

566
00:33:27,733 --> 00:33:30,400
(applause)