1
00:00:08,530 --> 00:00:11,120
- [Woman] So our next
presenter is Chris Williams

2
00:00:11,120 --> 00:00:15,060
with the Graduate School of
Geography at Clark University.

3
00:00:15,060 --> 00:00:16,517
And he will be presenting on

4
00:00:16,517 --> 00:00:20,267
"Plots to Pixels to Policy
The Intrepid Journey

5
00:00:20,267 --> 00:00:23,070
"of a Wall-to-Wall
Forest Carbon Monitoring

6
00:00:23,070 --> 00:00:23,903
"and Assessment System."

7
00:00:23,903 --> 00:00:24,736
- Fantastic. All right.

8
00:00:24,736 --> 00:00:26,090
I changed it from Intrepid to Foolhardy

9
00:00:26,090 --> 00:00:28,840
because I knew that Chris
Woodall would all would be here

10
00:00:28,840 --> 00:00:31,950
and that I expect great hard questions

11
00:00:31,950 --> 00:00:35,653
because what we're trying to
do is a really hard thing.

12
00:00:36,760 --> 00:00:38,430
I want to, before I get too far into this,

13
00:00:38,430 --> 00:00:41,960
thank some of the funding sources.

14
00:00:41,960 --> 00:00:44,480
For example, NASA has
supported a lot of this work,

15
00:00:44,480 --> 00:00:46,960
so has the Nature Conservancy,
a bit of support from NSF.

16
00:00:46,960 --> 00:00:48,630
I'm a professor at Clark University

17
00:00:48,630 --> 00:00:50,800
in geography and environmental science.

18
00:00:50,800 --> 00:00:54,300
And I have a group, a team of
individuals working with me

19
00:00:54,300 --> 00:00:57,590
to try and cobble together

20
00:00:57,590 --> 00:00:59,810
a comprehensive wall-to-wall picture

21
00:00:59,810 --> 00:01:03,695
of forest carbon dynamics
across the conterminous U.S.,

22
00:01:03,695 --> 00:01:05,020
and that's what I'm going
to talk to you about today.

23
00:01:05,020 --> 00:01:08,769
So thanks to Natalia
Hasler, Huan Gu, Yu Zhou,

24
00:01:08,769 --> 00:01:11,233
and other members of my group
for their contributions.

25
00:01:15,828 --> 00:01:17,730
We all know that carbon
is under the microscope

26
00:01:17,730 --> 00:01:20,860
for good reason in forests
because we recognize

27
00:01:20,860 --> 00:01:24,460
that forest carbon plays
a really significant role

28
00:01:24,460 --> 00:01:28,160
in carbon emissions and carbon uptake

29
00:01:28,160 --> 00:01:31,890
across certainly the U.S.
and much of the world,

30
00:01:31,890 --> 00:01:34,990
forests being chief among
the natural working lands

31
00:01:34,990 --> 00:01:38,173
that contribute this kind
of ecosystem service.

32
00:01:39,100 --> 00:01:42,880
But in order for us to size up policies

33
00:01:42,880 --> 00:01:45,030
and assess the current state,

34
00:01:45,030 --> 00:01:48,470
we need great baseline
measurement capabilities,

35
00:01:48,470 --> 00:01:51,430
reporting capabilities,
opportunities to verify,

36
00:01:51,430 --> 00:01:54,160
and also just assessment
and understanding systems

37
00:01:54,160 --> 00:01:56,030
so that we can detect, monitor,

38
00:01:56,030 --> 00:01:58,750
and assess changes over time,

39
00:01:58,750 --> 00:02:02,310
and project where we think
things are going in the future.

40
00:02:02,310 --> 00:02:03,300
So what do we really need?

41
00:02:03,300 --> 00:02:05,690
We need to stand-level
estimates of carbon emissions

42
00:02:05,690 --> 00:02:08,250
and removals from the atmosphere;

43
00:02:08,250 --> 00:02:10,500
annually resolved, repeatable, everywhere,

44
00:02:10,500 --> 00:02:14,460
wall-to-wall, and ideally
attributed to drivers.

45
00:02:14,460 --> 00:02:16,960
Many of the methods on the
market today to do this

46
00:02:16,960 --> 00:02:20,120
have emphasized appropriately
live aboveground biomass,

47
00:02:20,120 --> 00:02:21,700
because it's something we can see remotely

48
00:02:21,700 --> 00:02:26,050
and detect spatially, but
really much of what we're after

49
00:02:26,050 --> 00:02:28,020
is carbon emissions and removals

50
00:02:28,020 --> 00:02:29,510
or exchange with the atmosphere,

51
00:02:29,510 --> 00:02:32,450
at least with respect to the policy world

52
00:02:32,450 --> 00:02:34,570
concerned with climate change.

53
00:02:34,570 --> 00:02:36,050
And so existing methods

54
00:02:36,050 --> 00:02:38,640
are often ambiguous about mechanisms.

55
00:02:38,640 --> 00:02:41,440
They often lack wall-to-wall coverage

56
00:02:41,440 --> 00:02:44,283
and may not have full
forecasting capability.

57
00:02:45,250 --> 00:02:47,760
In my group some time ago, we recognized

58
00:02:47,760 --> 00:02:51,310
a really significant opportunity
to take remote sensing

59
00:02:51,310 --> 00:02:53,410
of disturbances, forest disturbances,

60
00:02:53,410 --> 00:02:56,520
remote sensing of biomass at an epoch,

61
00:02:56,520 --> 00:02:57,790
or a moment in time,

62
00:02:57,790 --> 00:03:02,440
combine that with FIA
inventory data to assess

63
00:03:02,440 --> 00:03:04,790
that it can help us
understand stock accumulation

64
00:03:04,790 --> 00:03:08,260
and change over time in
forests of different types

65
00:03:08,260 --> 00:03:11,070
with also a carbon cycle
model that helps us extend

66
00:03:11,070 --> 00:03:15,210
beyond the the measured
quantities in the field

67
00:03:15,210 --> 00:03:20,210
in plots and also understand
full forest carbon dynamics.

68
00:03:20,750 --> 00:03:23,290
And so recognizing this opportunity,

69
00:03:23,290 --> 00:03:25,270
I developed what I'm calling now

70
00:03:25,270 --> 00:03:27,720
the National Forest
Carbon Monitoring System

71
00:03:27,720 --> 00:03:31,090
or Carbon Monitoring Reporting
and Assessment System,

72
00:03:31,090 --> 00:03:33,750
and it has a flow diagram
that looks like this.

73
00:03:33,750 --> 00:03:35,750
We're gonna take it one piece at a time.

74
00:03:37,350 --> 00:03:40,047
First we'll look at how
our team has been using

75
00:03:40,047 --> 00:03:43,120
FIA yield curves and plot-based data

76
00:03:43,120 --> 00:03:46,260
served through a validator to
generate above-ground biomass

77
00:03:46,260 --> 00:03:50,110
with age curves, and then
how we derive carbon stock

78
00:03:50,110 --> 00:03:53,993
and flux assessments at a
stand scale with those data.

79
00:03:55,290 --> 00:03:58,070
This is just one example
from, happens to be,

80
00:03:58,070 --> 00:03:59,460
from the Pacific Northwest,

81
00:03:59,460 --> 00:04:02,283
a Douglas fir
high-productivity stand forest.

82
00:04:03,130 --> 00:04:05,140
And overall, this isn't a singular forest.

83
00:04:05,140 --> 00:04:08,300
This is actually a
collection of forest stands

84
00:04:08,300 --> 00:04:12,200
where in-situ plot measurements
were made by the FIA,

85
00:04:12,200 --> 00:04:13,890
and they can be arrayed in this fashion

86
00:04:13,890 --> 00:04:16,670
on the lower left-hand curve where we show

87
00:04:16,670 --> 00:04:21,380
above-ground live wood with
stand age in age classes.

88
00:04:21,380 --> 00:04:23,740
And we are forcing our
cost of carbon cycle

89
00:04:23,740 --> 00:04:27,230
ecosystem process model
to match the yield curve

90
00:04:28,260 --> 00:04:31,870
carbon accumulation in a forest
of this type in this region,

91
00:04:31,870 --> 00:04:34,220
and of this site productivity class.

92
00:04:34,220 --> 00:04:37,510
When we do that, we then
derive curves, for example,

93
00:04:37,510 --> 00:04:41,050
of net ecosystem productivity
as a function of stand age.

94
00:04:41,050 --> 00:04:44,190
And this can be produced for a whole array

95
00:04:44,190 --> 00:04:46,240
of forest type groups

96
00:04:46,240 --> 00:04:49,440
and of high and low-productivity classes,

97
00:04:49,440 --> 00:04:51,840
which is a really important
additional variable.

98
00:04:53,510 --> 00:04:56,130
We've done that for all of
the dominant forest types

99
00:04:56,130 --> 00:04:58,950
and regions across the United States.

100
00:04:58,950 --> 00:05:02,770
We now couple that with
observations from remote sensing

101
00:05:02,770 --> 00:05:06,530
of biomass and of disturbance,
as well as ancillary data

102
00:05:06,530 --> 00:05:10,740
on forest type distribution
and site productivity class.

103
00:05:10,740 --> 00:05:13,170
And we do that so that
we can derive information

104
00:05:13,170 --> 00:05:15,890
about the stand age, disturbance legacy,

105
00:05:15,890 --> 00:05:20,760
and site conditions of
individual 30-meter pixels

106
00:05:21,620 --> 00:05:23,423
all across the conterminous U.S.

107
00:05:24,360 --> 00:05:27,060
So just to give you a quick
glimpse of how we do this,

108
00:05:27,060 --> 00:05:31,100
this is, we're currently
using the Kellndorfer NACP

109
00:05:31,100 --> 00:05:34,233
Aboveground Biomass
Carbon Baseline Dataset.

110
00:05:35,824 --> 00:05:38,920
And so, for example, this is
again our Pacific Northwest

111
00:05:38,920 --> 00:05:41,610
Douglas fir stand.

112
00:05:41,610 --> 00:05:44,050
If it was high-productivity,
and we had a measurement

113
00:05:44,050 --> 00:05:47,080
in an individual pixel of
about 10 kilograms of carbon

114
00:05:47,080 --> 00:05:51,000
per meter squared, we can
use our FIA derived curves

115
00:05:51,000 --> 00:05:55,210
to characterize what the stand
age distribution would be

116
00:05:55,210 --> 00:05:57,330
that we would assume is represented

117
00:05:58,220 --> 00:06:02,070
by this 10 kilogram of carbon
per meter squared condition.

118
00:06:02,070 --> 00:06:04,640
If it's a low-productivity stand,

119
00:06:04,640 --> 00:06:08,540
we can intersect that
observation with some uncertainty

120
00:06:08,540 --> 00:06:12,853
and derive a stand age band
shown with the green lines.

121
00:06:14,330 --> 00:06:16,320
So what we're doing here is we're taking

122
00:06:16,320 --> 00:06:18,090
aboveground biomass as a,

123
00:06:18,090 --> 00:06:20,330
to using it as a baseline initialization

124
00:06:20,330 --> 00:06:22,700
in the year of the observations,

125
00:06:22,700 --> 00:06:24,700
this was around the year 2000,

126
00:06:24,700 --> 00:06:28,130
and we're deriving a
biomass equivalent stand age

127
00:06:28,130 --> 00:06:31,510
that can then be a marker
of the current stocks,

128
00:06:31,510 --> 00:06:32,830
where they were previously

129
00:06:32,830 --> 00:06:35,800
if the stand stayed free of disturbance,

130
00:06:35,800 --> 00:06:38,334
and where it would be
going in the next, say,

131
00:06:38,334 --> 00:06:39,333
10, 20, 50 years.

132
00:06:40,700 --> 00:06:43,890
Also, we're relying on a Landsat-derived

133
00:06:43,890 --> 00:06:45,640
disturbance data products.

134
00:06:45,640 --> 00:06:48,630
These originated as the North
American Forest Dynamics

135
00:06:48,630 --> 00:06:52,790
Data Products 30-meter
mapping of disturbances.

136
00:06:52,790 --> 00:06:55,410
They initially were not attributed
to the disturbance type.

137
00:06:55,410 --> 00:06:58,380
But in my group, we
intersected those 30-meter

138
00:06:58,380 --> 00:07:01,210
Landsat-derived forest disturbance events

139
00:07:01,210 --> 00:07:05,290
with MTBS fire scars or
burn scar information,

140
00:07:05,290 --> 00:07:08,740
as well as aerial detection
survey data on bark beetles

141
00:07:08,740 --> 00:07:11,520
and other biotic agents
to try to attribute,

142
00:07:11,520 --> 00:07:15,640
in a cursory way, each of those
individual 30- meter scale

143
00:07:15,640 --> 00:07:19,340
year-specific disturbance
events to a driver

144
00:07:19,340 --> 00:07:20,963
and to map these out in space.

145
00:07:23,220 --> 00:07:28,150
Those two pieces, biomass and
disturbance, can be combined

146
00:07:28,150 --> 00:07:32,060
in a way to derive a
continuous map of stand age.

147
00:07:32,060 --> 00:07:34,530
And this is continuously mapped for us

148
00:07:34,530 --> 00:07:38,573
across the conterminous U.S.
from 1986 to 2010 at present.

149
00:07:40,037 --> 00:07:41,370
And once we have these two pieces,

150
00:07:41,370 --> 00:07:44,660
our yield curve derived carbon
sequestration, for example,

151
00:07:44,660 --> 00:07:46,480
or carbon stock dynamics,

152
00:07:46,480 --> 00:07:49,980
as well as information about
every pixel's forest type,

153
00:07:49,980 --> 00:07:53,400
site productivity class,
biomass, or disturbance

154
00:07:53,400 --> 00:07:54,870
from which we derive stand age,

155
00:07:54,870 --> 00:07:59,070
we can estimate net ecosystem
productivity, for example.

156
00:07:59,070 --> 00:08:01,240
And we're gonna zoom in on
New England plus New York

157
00:08:01,240 --> 00:08:02,830
in just a second.

158
00:08:02,830 --> 00:08:04,600
So that's the sort of data inputs

159
00:08:04,600 --> 00:08:06,380
and the intermediate variables,

160
00:08:06,380 --> 00:08:08,360
and now I'm going to try to transition

161
00:08:08,360 --> 00:08:10,130
to some applications that give you a sense

162
00:08:10,130 --> 00:08:11,470
of what we can do with this framework.

163
00:08:11,470 --> 00:08:14,300
We will also take a
look at some evaluation.

164
00:08:14,300 --> 00:08:16,380
Now we can, we have our annual maps

165
00:08:16,380 --> 00:08:17,850
of carbon stocks and fluxes

166
00:08:17,850 --> 00:08:20,310
at a 30-meter resolution nationally.

167
00:08:20,310 --> 00:08:23,380
And we're also looking at
applying them to understand

168
00:08:23,380 --> 00:08:25,580
regional carbon balance and change,

169
00:08:25,580 --> 00:08:28,270
to try to attribute
those changes to drivers,

170
00:08:28,270 --> 00:08:31,160
to estimate carbon stock
potential of landscapes,

171
00:08:31,160 --> 00:08:33,523
and also to assess carbon emissions risks.

172
00:08:35,590 --> 00:08:37,580
So zooming a little closer to home,

173
00:08:37,580 --> 00:08:40,820
this is a map of forest
carbon stocks in the year 2010

174
00:08:40,820 --> 00:08:42,530
according to our data product

175
00:08:42,530 --> 00:08:45,130
across New York plus New England States.

176
00:08:45,130 --> 00:08:48,730
And I just zoomed in on one
particular track to show

177
00:08:48,730 --> 00:08:50,210
not just aboveground biomass

178
00:08:50,210 --> 00:08:54,300
but the total live
biomass in this one box.

179
00:08:54,300 --> 00:08:57,720
We can report it at a pixel
level or over the region.

180
00:08:57,720 --> 00:09:01,410
This happens to hold 68
million metric tons of carbon.

181
00:09:01,410 --> 00:09:04,760
It's mapped here in kilograms
of carbon per meter squared.

182
00:09:04,760 --> 00:09:06,150
And our technique also extends

183
00:09:06,150 --> 00:09:08,240
beyond these biomass estimates

184
00:09:08,240 --> 00:09:12,273
to total ecosystem carbon stocks
as shown in this lower box.

185
00:09:13,270 --> 00:09:15,710
Now, another thing we
can do is, as I said,

186
00:09:15,710 --> 00:09:18,650
zoom in on a singular
pixel and ask not just

187
00:09:18,650 --> 00:09:21,620
what it held in terms of
its ecosystem carbon stocks

188
00:09:21,620 --> 00:09:23,460
in the year 2010,

189
00:09:23,460 --> 00:09:25,100
but based on its forest type,

190
00:09:25,100 --> 00:09:28,240
for example a maple-beech
birch forest type group,

191
00:09:28,240 --> 00:09:30,600
if the stand had about
six kilograms of carbon

192
00:09:30,600 --> 00:09:33,270
per meter squared in a
year, I don't know, 2000,

193
00:09:33,270 --> 00:09:35,313
how much might it hold in the future?

194
00:09:37,060 --> 00:09:41,340
So we now can take our
yield curve information,

195
00:09:41,340 --> 00:09:42,810
for example on the left-hand side

196
00:09:42,810 --> 00:09:46,920
for a Northeast maple-beech
birch of high productivity,

197
00:09:46,920 --> 00:09:50,650
the FIA data are sampled
with the red pluses.

198
00:09:50,650 --> 00:09:53,030
These curves are our
fits through those data

199
00:09:53,030 --> 00:09:57,573
with the curved turf fitting
routine that we have.

200
00:09:58,410 --> 00:10:01,020
And let's focus in on just this window

201
00:10:01,020 --> 00:10:03,773
of about 40- to 90-year-old stands.

202
00:10:04,767 --> 00:10:07,760
If it started at about
six kilograms of carbon

203
00:10:07,760 --> 00:10:11,550
per meter squared and
it was going to acquire

204
00:10:11,550 --> 00:10:14,260
about 10 after another 50 years,

205
00:10:14,260 --> 00:10:16,650
we can see the aboveground
biomass expected

206
00:10:16,650 --> 00:10:20,930
in this maple-beech birch
forest in the Northeast,

207
00:10:20,930 --> 00:10:22,820
but we also with our
framework are able to see

208
00:10:22,820 --> 00:10:25,590
what we expect in terms
of soil carbon dynamics,

209
00:10:25,590 --> 00:10:28,740
coarse woody debris dynamics,
below ground biomass dynamics,

210
00:10:28,740 --> 00:10:30,310
and so forth.

211
00:10:30,310 --> 00:10:33,060
Furthermore, over that 50-year-period

212
00:10:33,060 --> 00:10:34,990
from a stand age of 40 to 90,

213
00:10:34,990 --> 00:10:37,960
we expect about six kilograms of carbon

214
00:10:37,960 --> 00:10:41,220
to be stored in this
ecosystem if it remains

215
00:10:41,220 --> 00:10:44,770
free of disturbance from some initial year

216
00:10:44,770 --> 00:10:45,710
over that 50 years.

217
00:10:45,710 --> 00:10:47,520
And we see that most of it is in biomass

218
00:10:47,520 --> 00:10:49,630
but there's also significant accumulation

219
00:10:49,630 --> 00:10:51,710
in coarse woody debris.

220
00:10:51,710 --> 00:10:53,530
The framework also allows us to estimate

221
00:10:53,530 --> 00:10:56,690
year-specific fluxes
of carbon net exchange

222
00:10:56,690 --> 00:11:00,490
with the atmosphere in terms
of net ecosystem productivity.

223
00:11:00,490 --> 00:11:03,240
And that's what's shown
here with a slight decline

224
00:11:03,240 --> 00:11:07,460
over this 50-year period as
the yield curves would suggest

225
00:11:07,460 --> 00:11:10,023
due to a tendency towards saturation.

226
00:11:12,440 --> 00:11:15,500
We've done this for every
pixel in our landscape

227
00:11:15,500 --> 00:11:19,370
to just look at the 20-year
carbon sequestration opportunity

228
00:11:19,370 --> 00:11:23,100
that we expect at this
30-meter scale across New York

229
00:11:23,100 --> 00:11:24,663
and New England States,

230
00:11:25,920 --> 00:11:28,373
and that's shown in this map illustration.

231
00:11:30,460 --> 00:11:32,490
There are some areas where we see recent

232
00:11:32,490 --> 00:11:34,760
clearcutting having a
legacy of carbon emissions

233
00:11:34,760 --> 00:11:36,030
in those sites.

234
00:11:36,030 --> 00:11:38,450
That's why we see some negative values.

235
00:11:38,450 --> 00:11:40,010
And even over this 20-year period,

236
00:11:40,010 --> 00:11:42,330
that can result in a net emission

237
00:11:42,330 --> 00:11:46,210
for that 20-year period as
it tends towards regrowth

238
00:11:46,210 --> 00:11:49,520
and carbon sinks following
a clearing event.

239
00:11:49,520 --> 00:11:52,280
But much of the landscape
is shown to have moderate

240
00:11:52,280 --> 00:11:55,400
to high carbon uptake
sequestration potential,

241
00:11:55,400 --> 00:11:57,370
and this can be a really useful tool

242
00:11:57,370 --> 00:12:00,200
for policy makers and
state-level accounting

243
00:12:00,200 --> 00:12:03,203
and for assessing
opportunities going forward.

244
00:12:06,499 --> 00:12:07,950
So that's a nice prediction,

245
00:12:07,950 --> 00:12:09,990
but of course people would be wondering

246
00:12:09,990 --> 00:12:14,990
how effective we expect,
how how successful this is

247
00:12:15,000 --> 00:12:17,190
when we compare it to other data products.

248
00:12:17,190 --> 00:12:20,520
So this now is a time series
of aboveground biomass

249
00:12:20,520 --> 00:12:23,100
from 1986 to 2010.

250
00:12:23,100 --> 00:12:24,910
And we're gonna compare
this aboveground biomass

251
00:12:24,910 --> 00:12:28,630
to two data products,
actually three, excuse me.

252
00:12:28,630 --> 00:12:30,510
Our data product is in the red line,

253
00:12:30,510 --> 00:12:32,630
and this is for the Pacific Northwest,

254
00:12:32,630 --> 00:12:35,380
a place where we've
started to look in depth

255
00:12:35,380 --> 00:12:37,640
at what we've seen over the region.

256
00:12:37,640 --> 00:12:39,470
Starting on the left, if we focus in

257
00:12:39,470 --> 00:12:41,710
on recently undisturbed forests,

258
00:12:41,710 --> 00:12:45,430
we see that our estimate
is agreeing quite well

259
00:12:45,430 --> 00:12:48,270
with the data set that we
train it to agree with,

260
00:12:48,270 --> 00:12:50,820
the NBCD calendar for data set.

261
00:12:50,820 --> 00:12:53,380
So that's no surprise
but it is reassuring.

262
00:12:53,380 --> 00:12:56,550
And it also agrees well
with Steven Hagen's dataset

263
00:12:56,550 --> 00:12:57,870
with Sussan Saatchi and others,

264
00:12:57,870 --> 00:13:00,320
I think Grant Domke was
involved in this as well,

265
00:13:01,660 --> 00:13:03,180
over the Pacific Northwest region

266
00:13:03,180 --> 00:13:06,370
where we saw a tendency
towards increased carbon stocks

267
00:13:06,370 --> 00:13:08,860
particularly in the undisturbed forests.

268
00:13:08,860 --> 00:13:12,010
When we include the areas
where North American

269
00:13:12,010 --> 00:13:14,750
forest dynamics told us
disturbances had occurred,

270
00:13:14,750 --> 00:13:17,760
we find a little bit of a
different story where instead

271
00:13:17,760 --> 00:13:20,540
of a tendency for regional
accumulation of carbon,

272
00:13:20,540 --> 00:13:22,863
really there's a relatively flat,

273
00:13:23,710 --> 00:13:26,120
relatively constant aboveground biomass.

274
00:13:26,120 --> 00:13:28,940
And that also agrees well
with what we were seeing

275
00:13:28,940 --> 00:13:31,750
with our other data products.

276
00:13:31,750 --> 00:13:34,110
Furthermore, eMapR is a data product,

277
00:13:34,110 --> 00:13:37,370
also 30-meter biomass, annually-resolved,

278
00:13:37,370 --> 00:13:41,390
generated by Robert Kennedy
at Oregon State University

279
00:13:41,390 --> 00:13:45,030
with his LandTrendr and
some Google Earth engine

280
00:13:46,050 --> 00:13:48,110
neural network approaches.

281
00:13:48,110 --> 00:13:49,990
And he has shown, well,

282
00:13:49,990 --> 00:13:52,210
our products show very good agreement.

283
00:13:52,210 --> 00:13:55,990
They're both related to
Landsat-derived information

284
00:13:55,990 --> 00:13:58,900
but they're using that
in very different ways.

285
00:13:58,900 --> 00:14:02,400
Also on the right, I would
show good agreement overall

286
00:14:02,400 --> 00:14:04,470
between bio at a forest-type level

287
00:14:04,470 --> 00:14:05,960
across the Pacific Northwest

288
00:14:05,960 --> 00:14:08,410
between our data product in the red bands

289
00:14:08,410 --> 00:14:11,530
showing that year-to-year
variability range

290
00:14:11,530 --> 00:14:13,830
for this 25-year period and what we see

291
00:14:13,830 --> 00:14:17,440
in the NBCD and Hagen et al data sets.

292
00:14:17,440 --> 00:14:19,863
We see that across the
dominant forest types.

293
00:14:22,460 --> 00:14:23,720
Another thing I'd like to show you

294
00:14:23,720 --> 00:14:26,380
that we can do is to zoom
in on a local assessment.

295
00:14:26,380 --> 00:14:27,213
And this is showing,

296
00:14:27,213 --> 00:14:28,520
sorry for the busy graph,

297
00:14:28,520 --> 00:14:30,840
one particular postage stamp,

298
00:14:30,840 --> 00:14:34,857
and it's showing 1986, 94, 2002, and 2010,

299
00:14:34,857 --> 00:14:37,293
and this period of increased harvest.

300
00:14:38,330 --> 00:14:40,410
Another thing we can do is track

301
00:14:40,410 --> 00:14:42,920
regional scale monitoring and assessment.

302
00:14:42,920 --> 00:14:44,860
So we're looking at carbon stock dynamics,

303
00:14:44,860 --> 00:14:47,940
harvest removals,
emissions, fire emissions,

304
00:14:47,940 --> 00:14:50,910
and then net biome productivity
and ecosystem productivity

305
00:14:50,910 --> 00:14:54,833
on annual resolution across
the entire Pacific Northwest.

306
00:14:55,950 --> 00:14:58,060
I see that we have one minute.

307
00:14:58,060 --> 00:14:59,220
Another thing I'll just tell you

308
00:14:59,220 --> 00:15:01,810
is that we're doing state-level
reporting and monitoring

309
00:15:01,810 --> 00:15:03,380
and we're starting to zoom in

310
00:15:03,380 --> 00:15:06,740
on the carbon sequestration
opportunities, carbon stocks,

311
00:15:06,740 --> 00:15:09,080
in the '90s, 2000s, 2010s,

312
00:15:09,080 --> 00:15:12,460
and out to 2030 and '50.

313
00:15:12,460 --> 00:15:15,610
For example, here for New Hampshire.

314
00:15:15,610 --> 00:15:17,200
We're excited to be working with states

315
00:15:17,200 --> 00:15:19,280
who are participating in
the U.S. Climate Alliance

316
00:15:19,280 --> 00:15:20,263
to do this work.

317
00:15:21,980 --> 00:15:23,890
In future work we're continuing to look

318
00:15:23,890 --> 00:15:27,200
at forest conversions and loss of forests.

319
00:15:27,200 --> 00:15:31,570
And we're also looking at
this at a national scale

320
00:15:31,570 --> 00:15:34,030
to come up with some very
important conclusions.

321
00:15:34,030 --> 00:15:35,890
For example, that forests
are currently holding

322
00:15:35,890 --> 00:15:39,040
only about 50% of their
carbon stock potential;

323
00:15:39,040 --> 00:15:41,530
disturbances are being
shown to push some regions

324
00:15:41,530 --> 00:15:43,610
toward net carbon emission;

325
00:15:43,610 --> 00:15:46,180
we're seeing that avoided
forest carbon loss

326
00:15:46,180 --> 00:15:49,300
could nationwide prevent an emission

327
00:15:49,300 --> 00:15:52,270
of about 20 teragrams of carbon per year;

328
00:15:52,270 --> 00:15:54,710
and that harvest emissions are, as known,

329
00:15:54,710 --> 00:15:58,293
a major offset to carbon
uptake in forest lands.

330
00:16:00,500 --> 00:16:02,090
We're doing a lot of innovations.

331
00:16:02,090 --> 00:16:05,590
We intend to work with repeat measures

332
00:16:05,590 --> 00:16:08,790
of FIA plots to test our framework

333
00:16:08,790 --> 00:16:11,293
as well as to perform
some scenario testing.

334
00:16:14,010 --> 00:16:15,580
How was the final estimate comparing

335
00:16:15,580 --> 00:16:17,160
with field measure data?

336
00:16:17,160 --> 00:16:19,420
And if this assessment can
be used to put arrow bars

337
00:16:19,420 --> 00:16:21,220
in some of the predictions?

338
00:16:21,220 --> 00:16:23,660
Yes. We haven't done that
work yet and we should.

339
00:16:23,660 --> 00:16:24,880
We're looking forward to doing it.

340
00:16:24,880 --> 00:16:27,950
So we've been so fixated
on generating estimates

341
00:16:27,950 --> 00:16:30,870
at pixel scales where
pixel for us is 30 meters

342
00:16:30,870 --> 00:16:32,390
we haven't yet intersected that

343
00:16:32,390 --> 00:16:34,600
with individual plot measurements

344
00:16:34,600 --> 00:16:38,070
either for a given epoch
or repeat measurements.

345
00:16:38,070 --> 00:16:40,330
So as I briefly mentioned,

346
00:16:40,330 --> 00:16:43,660
that's one of the, say, innovations
or pieces of future work

347
00:16:45,420 --> 00:16:46,820
that we'd really like to do.

348
00:16:48,730 --> 00:16:51,650
Some of our maps are
publicly available, James.

349
00:16:51,650 --> 00:16:54,960
So for example, we are, we have posted

350
00:16:54,960 --> 00:16:58,203
with TNC on the Resilient
Lands Mapping Tool.

351
00:17:00,310 --> 00:17:01,883
Many of our data sets, and we're,

352
00:17:03,250 --> 00:17:05,050
and we're also generating this for

353
00:17:07,630 --> 00:17:10,500
the Oak Ridge National Lab DAAC,

354
00:17:10,500 --> 00:17:12,320
Data Archive Distribution Center,

355
00:17:12,320 --> 00:17:14,900
and that should be live within weeks.

356
00:17:14,900 --> 00:17:15,733
Who is we?

357
00:17:15,733 --> 00:17:17,539
This is Clark University,

358
00:17:17,539 --> 00:17:20,530
a team of mine in geography
and environmental science.

359
00:17:20,530 --> 00:17:24,750
And we are generating
state-level results for, well,

360
00:17:24,750 --> 00:17:26,340
like I show here, New Hampshire,

361
00:17:26,340 --> 00:17:28,410
but also all of the New
England and York States.

362
00:17:28,410 --> 00:17:31,063
And ultimately we plan
to do this nationwide.