1
00:00:00,000 --> 00:00:03,090
Urban air temperatures
using land cover type

2
00:00:03,090 --> 00:00:04,800
and satellite observations

3
00:00:04,800 --> 00:00:06,465
of surface temperatures.

4
00:00:06,465 --> 00:00:07,185
Here.

5
00:00:07,185 --> 00:00:08,595
Garrett, your
tendon starts.

6
00:00:08,595 --> 00:00:09,910
Now.

7
00:00:15,980 --> 00:00:19,755
Everybody. Thanks.
Sure. I have weeks.

8
00:00:19,755 --> 00:00:22,770
After all local
talks yesterday,

9
00:00:22,770 --> 00:00:24,720
I kind of wanted to
just show ten minutes.

10
00:00:24,720 --> 00:00:26,805
That's birds have
my brush here.

11
00:00:26,805 --> 00:00:28,050
But I guess, you know,

12
00:00:28,050 --> 00:00:30,825
someone has to get
depressing climb to talk.

13
00:00:30,825 --> 00:00:33,000
So let's dive right in.

14
00:00:33,000 --> 00:00:35,010
So I think the
through line on

15
00:00:35,010 --> 00:00:36,780
this talk is
really dealing

16
00:00:36,780 --> 00:00:37,800
with the influence of

17
00:00:37,800 --> 00:00:40,625
land cover, air
temperatures.

18
00:00:40,625 --> 00:00:41,840
And maybe you've thought

19
00:00:41,840 --> 00:00:42,920
about this, you know,

20
00:00:42,920 --> 00:00:44,330
if you've ever
put your foot on

21
00:00:44,330 --> 00:00:46,789
paid summer and
immediately

22
00:00:46,789 --> 00:00:48,665
jumped to the grass?

23
00:00:48,665 --> 00:00:51,140
Maybe you thought
less about how

24
00:00:51,140 --> 00:00:52,700
the same companies
are actually

25
00:00:52,700 --> 00:00:55,280
impacting temperatures
above it?

26
00:00:55,280 --> 00:00:57,694
It's just give
you some context.

27
00:00:57,694 --> 00:00:59,180
My wheelchair don't have

28
00:00:59,180 --> 00:01:00,860
to go too far
back in history.

29
00:01:00,860 --> 00:01:02,000
Just look at the heatwave

30
00:01:02,000 --> 00:01:03,230
of the summer where we

31
00:01:03,230 --> 00:01:05,720
saw temperatures
the Northwest,

32
00:01:05,720 --> 00:01:07,010
about 35 degrees

33
00:01:07,010 --> 00:01:09,245
Fahrenheit higher
than normal.

34
00:01:09,245 --> 00:01:10,670
And he said several

35
00:01:10,670 --> 00:01:12,994
adverse human
environmental impacts.

36
00:01:12,994 --> 00:01:15,860
For example, ER
visits due to Hebrew

37
00:01:15,860 --> 00:01:17,780
up about 70 times

38
00:01:17,780 --> 00:01:19,490
what they normally
would be.

39
00:01:19,490 --> 00:01:22,070
Unfortunately, to

40
00:01:22,070 --> 00:01:24,920
many deaths and
the Washington,

41
00:01:24,920 --> 00:01:26,765
Oregon area over 200.

42
00:01:26,765 --> 00:01:28,190
If you look
around per length

43
00:01:28,190 --> 00:01:31,415
specifically where
the deaths occurred,

44
00:01:31,415 --> 00:01:33,260
we overlay this
with a heatmap

45
00:01:33,260 --> 00:01:35,105
where red is the
hotter areas,

46
00:01:35,105 --> 00:01:37,925
see a lot of hot areas
and where it happened.

47
00:01:37,925 --> 00:01:39,530
And in fact,

48
00:01:39,530 --> 00:01:41,510
about 50 percent
of those who

49
00:01:41,510 --> 00:01:44,525
perished had
no central AC.

50
00:01:44,525 --> 00:01:46,430
So you know what's
going on here?

51
00:01:46,430 --> 00:01:48,395
Well, the obvious one
is climate change.

52
00:01:48,395 --> 00:01:50,720
Since 2005, the,

53
00:01:50,720 --> 00:01:53,690
there's the ten hottest
years on record.

54
00:01:53,690 --> 00:01:56,090
And the second part
is that cities

55
00:01:56,090 --> 00:01:58,160
tend to be hotter than
the surrounding areas.

56
00:01:58,160 --> 00:02:00,875
In fact, up to about
10 degrees Fahrenheit.

57
00:02:00,875 --> 00:02:02,630
And the bad news is this

58
00:02:02,630 --> 00:02:04,850
is probably going
to get worse due

59
00:02:04,850 --> 00:02:08,870
to one urban intensification
where by 2050,

60
00:02:08,870 --> 00:02:11,030
70 percent of the
world population will

61
00:02:11,030 --> 00:02:13,220
be in cities closer
to 90 percent.

62
00:02:13,220 --> 00:02:16,655
And US into that
we're going to see

63
00:02:16,655 --> 00:02:18,680
increased frequency
and length

64
00:02:18,680 --> 00:02:22,010
of heat waves in the US.

65
00:02:22,010 --> 00:02:25,535
So all this to say
is that it's very

66
00:02:25,535 --> 00:02:28,595
important for us to
measure air temperature.

67
00:02:28,595 --> 00:02:30,350
That's what humans
and the environment

68
00:02:30,350 --> 00:02:31,850
are experiencing.

69
00:02:31,850 --> 00:02:33,545
In order to mitigate it.

70
00:02:33,545 --> 00:02:34,970
The difficulty
comes in and

71
00:02:34,970 --> 00:02:36,140
city specifically,

72
00:02:36,140 --> 00:02:38,330
because over a
short distance,

73
00:02:38,330 --> 00:02:40,970
you can have a pretty
significant change

74
00:02:40,970 --> 00:02:42,470
and what's on the ground.

75
00:02:42,470 --> 00:02:43,970
So it's fairly difficult

76
00:02:43,970 --> 00:02:46,550
to capture with networks.

77
00:02:46,550 --> 00:02:48,515
It's heartbeat
enough out there.

78
00:02:48,515 --> 00:02:51,500
Whereas we could also
use remote sensing or

79
00:02:51,500 --> 00:02:54,724
satellite data to get
spatial coverage.

80
00:02:54,724 --> 00:02:56,870
Issue then is
that this data

81
00:02:56,870 --> 00:02:58,325
measures what's
on the ground

82
00:02:58,325 --> 00:03:00,065
for the surface,

83
00:03:00,065 --> 00:03:02,270
which is little
less relevant to

84
00:03:02,270 --> 00:03:04,550
human experience
in particular.

85
00:03:04,550 --> 00:03:07,205
And so kind of the key
there is, you know,

86
00:03:07,205 --> 00:03:09,575
how do we get

87
00:03:09,575 --> 00:03:10,850
the spatial coverage

88
00:03:10,850 --> 00:03:12,080
of the surface
temperature,

89
00:03:12,080 --> 00:03:13,430
but more of the relevance

90
00:03:13,430 --> 00:03:14,360
and the air temperatures

91
00:03:14,360 --> 00:03:17,345
are synergistic way to
combine these data.

92
00:03:17,345 --> 00:03:19,115
In the reason why we're

93
00:03:19,115 --> 00:03:20,630
interested in
that, for example,

94
00:03:20,630 --> 00:03:21,980
we'll be looking
at heat and

95
00:03:21,980 --> 00:03:23,660
equities in the city.

96
00:03:23,660 --> 00:03:25,685
So we can see
here For Poilane,

97
00:03:25,685 --> 00:03:27,350
wealthier areas tend to be

98
00:03:27,350 --> 00:03:29,360
cooler or more greener.

99
00:03:29,360 --> 00:03:32,060
He generally it
tends to impact

100
00:03:32,060 --> 00:03:33,560
minoritized populations

101
00:03:33,560 --> 00:03:35,765
more severely as well.

102
00:03:35,765 --> 00:03:38,735
So all that to
say, you know,

103
00:03:38,735 --> 00:03:40,370
is there a way
we can take this

104
00:03:40,370 --> 00:03:43,205
spatially good coverage of

105
00:03:43,205 --> 00:03:44,270
a surface temperatures and

106
00:03:44,270 --> 00:03:45,499
critic air temperatures

107
00:03:45,499 --> 00:03:46,640
to see how it's affecting

108
00:03:46,640 --> 00:03:48,590
people and the
environment.

109
00:03:48,590 --> 00:03:51,065
And so generally
the idea is,

110
00:03:51,065 --> 00:03:53,000
you want to do this
simply because you only

111
00:03:53,000 --> 00:03:53,240
have

112
00:03:53,240 --> 00:03:56,010
surface temperature even
though temperature.

113
00:03:59,020 --> 00:04:03,410
Well, sorry about that
brief intermission.

114
00:04:03,410 --> 00:04:06,230
They're ISO so boy,

115
00:04:06,230 --> 00:04:08,255
a blue screen my computer.

116
00:04:08,255 --> 00:04:11,780
So getting back to talking

117
00:04:11,780 --> 00:04:15,410
about these degree rudder

118
00:04:15,410 --> 00:04:16,610
hoping to do is

119
00:04:16,610 --> 00:04:19,595
predict temperature
for my career.

120
00:04:19,595 --> 00:04:21,470
Summary of the first
couple points is

121
00:04:21,470 --> 00:04:24,020
that key can
be bad, right?

122
00:04:24,020 --> 00:04:26,570
So China go quickly here.

123
00:04:26,570 --> 00:04:29,270
Generally, we can
predict air temperature,

124
00:04:29,270 --> 00:04:31,610
surface temperature
with our old friend,

125
00:04:31,610 --> 00:04:33,935
Y equals MX plus B.

126
00:04:33,935 --> 00:04:37,130
But the dad joke
version of best is,

127
00:04:37,130 --> 00:04:39,785
why does MX plus B

128
00:04:39,785 --> 00:04:40,970
meaning that we really

129
00:04:40,970 --> 00:04:42,920
don't understand
these coefficients.

130
00:04:42,920 --> 00:04:44,570
They don't have a
physical meaning,

131
00:04:44,570 --> 00:04:46,640
a relationship
there empirical.

132
00:04:46,640 --> 00:04:48,290
Furthermore,
when they build

133
00:04:48,290 --> 00:04:50,285
these relationships
in the past,

134
00:04:50,285 --> 00:04:52,370
it's ignoring
separately and

135
00:04:52,370 --> 00:04:54,980
collision-free all have
different impacts.

136
00:04:54,980 --> 00:04:57,485
This alternatively,
what we can do

137
00:04:57,485 --> 00:05:00,620
is use kind of an energy
balanced approach.

138
00:05:00,620 --> 00:05:02,060
And this is just to
show you that I'm

139
00:05:02,060 --> 00:05:03,920
not completely
making this up.

140
00:05:03,920 --> 00:05:06,065
But we can gets

141
00:05:06,065 --> 00:05:07,910
air temperature based

142
00:05:07,910 --> 00:05:09,395
only on surface
temperature.

143
00:05:09,395 --> 00:05:10,970
And then we can
benchmark lab versus

144
00:05:10,970 --> 00:05:13,145
just a traditional
linear model.

145
00:05:13,145 --> 00:05:15,200
Hey, I real quick,

146
00:05:15,200 --> 00:05:16,790
we can see your notes
if you want to do

147
00:05:16,790 --> 00:05:19,625
the presentation view
instead of a shoot.

148
00:05:19,625 --> 00:05:20,360
Okay.

149
00:05:20,360 --> 00:05:23,250
Giving away all
my secrets.

150
00:05:23,380 --> 00:05:27,440
We're looking ahead
until Okay, Let me see.

151
00:05:27,440 --> 00:05:30,960
No sloppiness,
unfortunately.

152
00:05:32,950 --> 00:05:35,495
Next to that
display settings,

153
00:05:35,495 --> 00:05:38,750
Ramadan, Tom, her up time.

154
00:05:38,750 --> 00:05:42,770
Yeah, It's the gifts
you display settings,

155
00:05:42,770 --> 00:05:46,775
ESRD, the Zoom Wireless
was blocking it.

156
00:05:46,775 --> 00:05:48,965
Okay. Sorry about that.

157
00:05:48,965 --> 00:05:50,540
I'm just trying to hit

158
00:05:50,540 --> 00:05:52,190
every single
technical difficulty

159
00:05:52,190 --> 00:05:53,480
at once here.

160
00:05:53,480 --> 00:05:56,870
Yeah, so in order to
do this prediction,

161
00:05:56,870 --> 00:05:59,240
we need extensive data.

162
00:05:59,240 --> 00:06:00,890
And so luckily one of

163
00:06:00,890 --> 00:06:02,615
our collaborators you
exchange this has

164
00:06:02,615 --> 00:06:06,575
air temperature data
over multiple cities.

165
00:06:06,575 --> 00:06:08,884
And with this data,

166
00:06:08,884 --> 00:06:10,400
we can use surface
temperature

167
00:06:10,400 --> 00:06:11,555
data from satellites,

168
00:06:11,555 --> 00:06:12,710
air temperature data from

169
00:06:12,710 --> 00:06:15,080
the back and
landcover to break

170
00:06:15,080 --> 00:06:17,630
out individual
predictive relationships

171
00:06:17,630 --> 00:06:20,250
to predict this
air temperature.

172
00:06:20,980 --> 00:06:24,095
So just jumping into
cellular assaults.

173
00:06:24,095 --> 00:06:26,105
If you look at the
air temperatures

174
00:06:26,105 --> 00:06:28,714
across these different
land cover types,

175
00:06:28,714 --> 00:06:30,050
we can see if we're taking

176
00:06:30,050 --> 00:06:31,130
kinda bloodstream and

177
00:06:31,130 --> 00:06:33,710
coins self-developed
and forests,

178
00:06:33,710 --> 00:06:35,870
that there are
significant differences

179
00:06:35,870 --> 00:06:36,980
it's suggesting

180
00:06:36,980 --> 00:06:38,795
or predicting
air temperature

181
00:06:38,795 --> 00:06:41,075
by land cover makes sense.

182
00:06:41,075 --> 00:06:43,130
If you look at
the results of

183
00:06:43,130 --> 00:06:45,274
that model or
that prediction.

184
00:06:45,274 --> 00:06:48,020
Now, we see a
pretty low error

185
00:06:48,020 --> 00:06:49,340
in our predictions.

186
00:06:49,340 --> 00:06:50,945
And similar to
the benchmark,

187
00:06:50,945 --> 00:06:52,340
which is good
benchmark been

188
00:06:52,340 --> 00:06:54,110
just a simple
linear model.

189
00:06:54,110 --> 00:06:55,580
It's typically better than

190
00:06:55,580 --> 00:06:58,010
most studies where they

191
00:06:58,010 --> 00:06:59,330
have a lower volume data.

192
00:06:59,330 --> 00:07:00,170
That's an advantage,

193
00:07:00,170 --> 00:07:01,444
but maybe more
importantly,

194
00:07:01,444 --> 00:07:03,830
we predict air temperature
by land cover.

195
00:07:03,830 --> 00:07:05,390
It tends to do better
than if you just

196
00:07:05,390 --> 00:07:07,805
did kind of approach.

197
00:07:07,805 --> 00:07:10,460
Then finally were
able to show is

198
00:07:10,460 --> 00:07:13,535
just if we go back
into its coefficients,

199
00:07:13,535 --> 00:07:16,370
we want to see some
separability alone,

200
00:07:16,370 --> 00:07:18,260
kind of the end points of

201
00:07:18,260 --> 00:07:19,580
the land cover classes for

202
00:07:19,580 --> 00:07:21,155
this benchmark
linear model,

203
00:07:21,155 --> 00:07:23,420
there's really not much.

204
00:07:23,420 --> 00:07:25,400
But when we look at
our physically-based

205
00:07:25,400 --> 00:07:26,540
or energy balance model,

206
00:07:26,540 --> 00:07:29,660
we tendencies ability
there, which is good.

207
00:07:29,660 --> 00:07:32,060
It means it's more
generalizable anymore.

208
00:07:32,060 --> 00:07:34,100
So we also see that when

209
00:07:34,100 --> 00:07:36,260
natural land covers
have higher slopes,

210
00:07:36,260 --> 00:07:37,700
which means they
kind of have more

211
00:07:37,700 --> 00:07:39,425
of a buffer to
temperature.

212
00:07:39,425 --> 00:07:42,140
Again, which is what
we'd expect to see.

213
00:07:42,140 --> 00:07:44,465
So what does
all this mean?

214
00:07:44,465 --> 00:07:46,340
Basically, in areas where

215
00:07:46,340 --> 00:07:47,480
we don't have attempts

216
00:07:47,480 --> 00:07:50,270
to use these relationships
to predicted,

217
00:07:50,270 --> 00:07:51,530
but also we're
understanding

218
00:07:51,530 --> 00:07:52,820
what's physically
happening to

219
00:07:52,820 --> 00:07:54,950
some extent based
on the fact that

220
00:07:54,950 --> 00:07:57,230
they're built on these,

221
00:07:57,230 --> 00:07:58,475
these are physically based

222
00:07:58,475 --> 00:08:00,185
energy balance
relationships.

223
00:08:00,185 --> 00:08:01,970
And we can use that also

224
00:08:01,970 --> 00:08:05,510
to look at CDC or
the high areas are.

225
00:08:05,510 --> 00:08:07,220
But it is important
when we look to

226
00:08:07,220 --> 00:08:09,890
mitigate some of the
effects of urban heat.

227
00:08:09,890 --> 00:08:11,915
We in the future
to address

228
00:08:11,915 --> 00:08:13,760
inequities in

229
00:08:13,760 --> 00:08:15,785
environmental impacts
that we face.

230
00:08:15,785 --> 00:08:19,160
So thanks and thanks
for bearing with me for

231
00:08:19,160 --> 00:08:20,900
all those problems and
Latin like engineering

232
00:08:20,900 --> 00:08:22,970
decided to shut down.
Yes, I can have.

233
00:08:22,970 --> 00:08:25,260
I appreciate
your attention.

234
00:08:25,750 --> 00:08:28,350
Thank you very much.