WEBVTT

80
00:00:00.000 --> 00:00:02.000
Jim, feel free to begin.

81
00:00:02.000 --> 00:00:08.000
>> All right. Thanks, everyone,
for joining

82
00:00:08.000 --> 00:00:11.000
us. I am Jim Rice, the session
moderator. We have quite a few

83
00:00:11.000 --> 00:00:17.000
people logged in from across the
country. We also have a really
great

84
00:00:17.000 --> 00:00:20.000
group of speakers whom I will
introduce in just a couple of
minutes. Let

85
00:00:20.000 --> 00:00:27.000
me first say a couple of words
about this workshop. The EPA
office

86
00:00:27.000 --> 00:00:29.934
of research and development has
designed this series of webinars

87
00:00:29.934 --> 00:00:35.934
to explore some of the
challenges that we face working
with mining

88
00:00:35.934 --> 00:00:36.934
sites and mineral processing
facilities.

89
00:00:36.934 --> 00:00:42.934
For those of you who might not
be familiar, the EPA

90
00:00:42.934 --> 00:00:48.934
research arm, the scientific
research arm helps inform the

91
00:00:48.934 --> 00:00:53.000
agency's decisions and policies.
The objective of the

92
00:00:53.000 --> 00:00:54.000
seminar is to both inform you,
the practitioners and
stakeholders about

93
00:00:54.000 --> 00:00:59.000
some of the latest tools and
techniques and methods are
available. Also,

94
00:00:59.000 --> 00:01:03.000
we want to understand your
research needs

95
00:01:03.000 --> 00:01:07.000
in the areas of Mine and Mineral
processing.

96
00:01:07.000 --> 00:01:10.000
The webinars are designed to be
interactive and generate
discussion,

97
00:01:10.000 --> 00:01:17.000
just like you would any
technical conference.

98
00:01:17.000 --> 00:01:23.000
>> Over the last three workshops
we have asked some questions
about

99
00:01:23.000 --> 00:01:28.934
the challenges you might face in
dealing with Mine and Mineral

100
00:01:28.934 --> 00:01:31.934
processing sites. We got quite a
few answers. One of

101
00:01:31.934 --> 00:01:36.934
the things that we found was
that there is a lot of unique

102
00:01:36.934 --> 00:01:40.934
issues associated with mining
sites, including the fact that
these are

103
00:01:40.934 --> 00:01:45.934
very large sites and complex,
large waste volumes.

104
00:01:45.934 --> 00:01:46.934
Some of the sources located
above and below

105
00:01:46.934 --> 00:01:50.934
ground across entire watersheds.

106
00:01:50.934 --> 00:01:53.000
There is a lot of interaction
between ground and surface

107
00:01:53.000 --> 00:01:57.000
water. Some of the conditions at
the site are

108
00:01:57.000 --> 00:02:03.000
unstable and there is a
potential for a release at any
time.

109
00:02:03.000 --> 00:02:09.000
In general, mining remediation
cost can be much higher than the

110
00:02:09.000 --> 00:02:12.000
traditional fight.

111
00:02:12.000 --> 00:02:13.000
To address the challenges that
we see we defined more effective

112
00:02:13.000 --> 00:02:17.000
ways to address these throughout
the

113
00:02:17.000 --> 00:02:18.000
project lifecycle. Some of these
include innovative
characterization

114
00:02:18.000 --> 00:02:22.000
and data management tools like
we will talk about

115
00:02:22.000 --> 00:02:26.000
today. We have talked about
emergency responses and how
important

116
00:02:26.000 --> 00:02:30.934
they are. Also, applying
adaptive management strategies
to effectively

117
00:02:30.934 --> 00:02:34.934
design remediation systems

118
00:02:34.934 --> 00:02:40.934
and mitigations. The series
brings together a lot

119
00:02:40.934 --> 00:02:45.934
of experts to explore the latest
thinking

120
00:02:45.934 --> 00:02:47.934
in these areas. Your job is to
listen, participate, digest, and
respond

121
00:02:47.934 --> 00:02:51.934
to that information that we have
presented and give us your
feedback

122
00:02:51.934 --> 00:02:56.000
and provide us with your
questions.

123
00:02:56.000 --> 00:03:00.000
Today's session will focus on
the data that we generate and
analyze

124
00:03:00.000 --> 00:03:05.000
at these huge sites. These large
mining sites

125
00:03:05.000 --> 00:03:08.000
often have a lot of data. We are
always looking for

126
00:03:08.000 --> 00:03:12.000
a better way to analyze the data
and present the information we
have

127
00:03:12.000 --> 00:03:15.000
to our site teams.

128
00:03:15.000 --> 00:03:20.000
Our first presentation is from
Katy and Doug from CDM Smith.

129
00:03:20.000 --> 00:03:24.000
They will present us with a tool
to analyze large and complex
datasets.

130
00:03:24.000 --> 00:03:28.000
Katy is a data analytics
specialist who

131
00:03:28.000 --> 00:03:28.934
brings technology solutions to
many different

132
00:03:28.934 --> 00:03:33.934
industries. Doug is an expert in
the transformation

133
00:03:33.934 --> 00:03:38.934
with extensive experience in
artificial intelligence, she
learning, and

134
00:03:38.934 --> 00:03:44.934
predictive and descriptive
analytics.

135
00:03:44.934 --> 00:03:50.934
Next, I'll will look at how we
use large, Metairie datasets

136
00:03:50.934 --> 00:04:07.000
to develop effective mediation
plans.

137
00:04:07.000 --> 00:04:16.000
I will be your final speaker of
this session and this series. I

138
00:04:16.000 --> 00:04:22.000
will present a case study of how
we develop

139
00:04:22.000 --> 00:04:25.000
a three-dimensional
visualization of a contaminated
site in Colorado

140
00:04:25.000 --> 00:04:30.000
and how it was used on the
project team. I am a senior
geologist and

141
00:04:30.000 --> 00:04:34.000
I work with the EPA to provide
technology transfer

142
00:04:34.000 --> 00:04:40.000
and innovative technology
applications to support the
needs of the

143
00:04:40.000 --> 00:04:45.000
EPA regions. Again, we encourage
you

144
00:04:45.000 --> 00:04:51.000
to actively participate.

145
00:04:51.000 --> 00:04:55.000
I would like to turn the
presentation over to our first
beakers,

146
00:04:55.000 --> 00:05:04.000
Katie Deheer and Doug Cushing.

147
00:05:04.000 --> 00:05:10.000
>> As mentioned I am Doug
Cushing. I will be

148
00:05:10.000 --> 00:05:15.000
doing an introduction into data
analytics as we and vision

149
00:05:15.000 --> 00:05:21.000
it being applied to abandon
minds

150
00:05:21.000 --> 00:05:25.000
and facilities. Katy will do a
live aministration to
demonstrate

151
00:05:25.000 --> 00:05:33.934
some of the capabilities with
these tools and the concept of

152
00:05:33.934 --> 00:05:36.934
data expiration.

153
00:05:36.934 --> 00:05:42.934
Jim provided our overview, I
will move on

154
00:05:42.934 --> 00:05:48.934
from there. I want to set the
stage, we hear a taught

155
00:05:48.934 --> 00:05:54.000
about analytics, big data, data
exploration

156
00:05:54.000 --> 00:06:02.000
and visualization. There is just
a huge impact across a whole lot

157
00:06:02.000 --> 00:06:06.000
of industries. There are a lot
of scenarios where there is a
lot

158
00:06:06.000 --> 00:06:10.000
of data, but there is not a lot
of information being pulled from

159
00:06:10.000 --> 00:06:16.000
that. We want to talk about
where data analytics fit into

160
00:06:16.000 --> 00:06:23.000
this field. There are a lot

161
00:06:23.000 --> 00:06:28.000
of fantastic, classic tools in
environmental investigation.
Data

162
00:06:28.000 --> 00:06:31.934
analytics does not seek to
replace or provide the same sort

163
00:06:31.934 --> 00:06:34.934
of insight.

164
00:06:34.934 --> 00:06:39.934
Geospatial analysis has been
around for decades and is
phenomenally

165
00:06:39.934 --> 00:06:48.934
powerful for representing
analytical

166
00:06:48.934 --> 00:06:55.000
chemistry results. These
high-quality results are

167
00:06:55.000 --> 00:06:59.000
extremely important. They are
actually not the focus of what
we

168
00:06:59.000 --> 00:07:05.000
are looking to do. We are
looking for datasets that exist
that have

169
00:07:05.000 --> 00:07:10.000
been overlooked because they
were considered to have
limitations,

170
00:07:10.000 --> 00:07:16.000
such as not being verify data,
missing data or

171
00:07:16.000 --> 00:07:21.000
erroneous data.

172
00:07:21.000 --> 00:07:26.000
>> This is not a different
approach to doing statistical
analysis like

173
00:07:26.000 --> 00:07:31.934
classics the two sticks or --
classic statistics

174
00:07:31.934 --> 00:07:37.934
or analysis. What is in focus is
keeping out trends in these

175
00:07:37.934 --> 00:07:41.934
very large, complex, and often
dirty or incomplete

176
00:07:41.934 --> 00:07:46.934
multidimensional datasets. When
we say large and
multidimensional

177
00:07:46.934 --> 00:07:51.934
we are looking at at least
thousands of rows of data

178
00:07:51.934 --> 00:07:58.000
if not tens of thousands. We
like the datasets wider,

179
00:07:58.000 --> 00:08:04.000
at least 10. often

180
00:08:04.000 --> 00:08:09.000
30-50-100 columns wide. We are
looking for

181
00:08:09.000 --> 00:08:12.000
the concept of directional
correctness and transactional

182
00:08:12.000 --> 00:08:18.000
accuracy. As we drill down we
will often find missing data,
erroneous

183
00:08:18.000 --> 00:08:22.000
or unreasonable data. The great
thing about these

184
00:08:22.000 --> 00:08:28.000
approaches is if we take 80% of
the data randomly we will draw

185
00:08:28.000 --> 00:08:33.934
a conclusion if we randomize it
again. The conclusions do not

186
00:08:33.934 --> 00:08:38.934
change because of the dirtiness
of the data. To overcome

187
00:08:38.934 --> 00:08:43.934
the dirtiness we do need these
large datasets so that you are
not overly

188
00:08:43.934 --> 00:08:48.934
influenced by any single point.

189
00:08:48.934 --> 00:08:58.000
>> We started this process by
searching for datasets in the
abandon

190
00:08:58.000 --> 00:09:01.000
mine space that we could apply
some of these techniques to to
demonstrate

191
00:09:01.000 --> 00:09:06.000
the potential. We did some
digging with people who are very
familiar

192
00:09:06.000 --> 00:09:08.000
with peak areas. We went to
research

193
00:09:08.000 --> 00:09:13.000
institutions like the Colorado
school of mine. We reached out
to

194
00:09:13.000 --> 00:09:19.000
different agencies, park
services and a number of

195
00:09:19.000 --> 00:09:30.934
publicly available abandoned
mine land databases.

196
00:09:30.934 --> 00:09:35.934
None resulted in content that
will provide us the powerful
data analytics

197
00:09:35.934 --> 00:09:39.934
demonstration. I will spoil the
conclusion, at the end we are

198
00:09:39.934 --> 00:09:43.934
looking for help from this
audience. Help for all the

199
00:09:43.934 --> 00:09:49.934
agencies that really want to do
these larger data

200
00:09:49.934 --> 00:09:50.934
analytics analyses.

201
00:09:50.934 --> 00:09:53.000
Help and direction from datasets
that we may use and what they
are

202
00:09:53.000 --> 00:10:02.000
will become clear as we go
through this presentation.

203
00:10:02.000 --> 00:10:06.000
>> With that I would like to
turn this over to Katie and have
her

204
00:10:06.000 --> 00:10:12.000
talk through the dataset we use
and why it

205
00:10:12.000 --> 00:10:19.000
meets the criteria for being
both large and underutilized and

206
00:10:19.000 --> 00:10:25.000
how we went about assessing
this.

207
00:10:25.000 --> 00:10:28.000
She will pop into the tool and
show you live some of the
demonstrations.

208
00:10:28.000 --> 00:10:33.934
>> Sure, thank you, Doug. And

209
00:10:33.934 --> 00:10:37.934
Lou of the great dataset to
demonstrate what is possible

210
00:10:37.934 --> 00:10:41.934
specific to the Atlanta MySpace
in the amount

211
00:10:41.934 --> 00:10:47.934
of time we had, instead we
defaulted to

212
00:10:47.934 --> 00:10:50.934
a dataset that is another
environmental dataset. We
thought that we could

213
00:10:50.934 --> 00:10:57.000
draw parallels. This dataset is
focused on

214
00:10:57.000 --> 00:11:03.000
two overflows. It is
self-reported and covers one
state over

215
00:11:03.000 --> 00:11:09.000
10 years, 20.000 events. That is
about 1 million

216
00:11:09.000 --> 00:11:12.000
or so data points.

217
00:11:12.000 --> 00:11:17.000
Being that it is self-reported

218
00:11:17.000 --> 00:11:21.000
it has free text and free feel,
no formatting issues and things

219
00:11:21.000 --> 00:11:25.000
like that. That wasn't going to
stop us

220
00:11:25.000 --> 00:11:28.934
from at least analyzing it and
trying to draw directionally

221
00:11:28.934 --> 00:11:34.934
correct conclusions. We did a
first pass analysis on

222
00:11:34.934 --> 00:11:38.934
raw data to understand the major
causes and drivers

223
00:11:38.934 --> 00:11:47.934
and see if there was something
there essentially. We found low

224
00:11:47.934 --> 00:11:54.000
hanging fruit. We took on the
order in less than eight hours

225
00:11:54.000 --> 00:11:58.000
to really do some easy mining
data cleanup. We reprocessed the

226
00:11:58.000 --> 00:12:04.000
data and found that the
conclusions were so strong that
if we were going

227
00:12:04.000 --> 00:12:10.000
to present in a public setting,
just out of

228
00:12:10.000 --> 00:12:14.000
respect for the agency without
we ought to anonymize the data
so as

229
00:12:14.000 --> 00:12:18.000
not to call out anyone for
things they have

230
00:12:18.000 --> 00:12:23.000
reported in the publicly
reported dataset.

231
00:12:23.000 --> 00:12:28.934
>> Again, we are going to draw
some parallels. We are going to

232
00:12:28.934 --> 00:12:34.934
try to draw some parallels and
open your mines to what

233
00:12:34.934 --> 00:12:39.934
is possible. With that we will
jump

234
00:12:39.934 --> 00:12:42.934
into the demonstration. You
should be

235
00:12:42.934 --> 00:12:46.934
seeing my screen here in just a
second.

236
00:12:46.934 --> 00:12:53.000
>> As Katie is calling up her
screen I will remind

237
00:12:53.000 --> 00:12:55.000
everyone that your controls
still exist in the

238
00:12:55.000 --> 00:13:02.000
upper right corner. You can use
those buttons to increase or
enlarge

239
00:13:02.000 --> 00:13:07.000
the size of the space as well as
zooming and further. Katie, I
can

240
00:13:07.000 --> 00:13:11.000
confirm that we have visual and
can see

241
00:13:11.000 --> 00:13:13.000
your screen.

242
00:13:13.000 --> 00:13:17.000
>> What you're seeing on the
screen is a

243
00:13:17.000 --> 00:13:21.000
dashboard. We are going to walk
through a series of scenarios so

244
00:13:21.000 --> 00:13:23.000
that you can see how we have
taken large,

245
00:13:23.000 --> 00:13:28.934
complex and dirty datasets and
use it to generate insight that
is essentially

246
00:13:28.934 --> 00:13:32.934
at your fingertips. First I will
walk

247
00:13:32.934 --> 00:13:37.934
you through everything that you
see on

248
00:13:37.934 --> 00:13:40.934
your screen so that it all comes
together and make sense before
we

249
00:13:40.934 --> 00:13:43.934
start interacting with the data.

250
00:13:43.934 --> 00:13:47.934
In the top left we are looking
at a

251
00:13:47.934 --> 00:13:54.000
geographic view. You will notice
here we are

252
00:13:54.000 --> 00:14:00.000
looking at anonymize data. We

253
00:14:00.000 --> 00:14:02.000
actually put the data points
from the one on

254
00:14:02.000 --> 00:14:07.000
the left and we shifted them
over to the UK. They are still
in

255
00:14:07.000 --> 00:14:11.000
the same place proportionate to
one another, but they just
shifted

256
00:14:11.000 --> 00:14:17.000
over to a less obvious location.

257
00:14:17.000 --> 00:14:23.000
On the map you will see a lot of
bubbles in different colors and

258
00:14:23.000 --> 00:14:28.067
sizes. The bubble size indicates
the volume of gallons

259
00:14:28.067 --> 00:14:36.000
associated with an overflow of
it that occurred in that
particular

260
00:14:36.000 --> 00:14:38.000
postal code.

261
00:14:38.000 --> 00:14:44.000
This bubble is bigger than all
the rest. It has many

262
00:14:44.000 --> 00:14:50.000
more millions of gallons
associated.

263
00:14:50.000 --> 00:14:56.000
The color of the bubbles as the

264
00:14:56.000 --> 00:14:59.000
scale suggests indicates the
number of overflows that have
occurred

265
00:14:59.000 --> 00:15:03.000
in a particular location. The
bigger and more wider a bubble
gets the

266
00:15:03.000 --> 00:15:08.000
worst it is. The smaller and the
more Lou

267
00:15:08.000 --> 00:15:13.000
gets by comparison it is not as
bad. Of course we're talking

268
00:15:13.000 --> 00:15:18.000
fewer overflows, one is bad. If
we

269
00:15:18.000 --> 00:15:22.000
are talking two or more, that is
where this analysis can

270
00:15:22.000 --> 00:15:24.000
help us.

271
00:15:24.000 --> 00:15:28.934
>> I will come back to this in a
minute. I want to draw your
attention

272
00:15:28.934 --> 00:15:32.934
to the visualization bullet.

273
00:15:32.934 --> 00:15:38.934
It is going to show the number
of overflows by

274
00:15:38.934 --> 00:15:43.934
year. This is total, about 10
years of data. To

275
00:15:43.934 --> 00:15:49.934
add dimensionality to the trend
year we have two dimensional
heat

276
00:15:49.934 --> 00:15:55.000
maps that look at the overflows
by month across

277
00:15:55.000 --> 00:16:01.000
the top and years on the side.

278
00:16:01.000 --> 00:16:05.000
One box in the heat map will
give us a year and a month and
the

279
00:16:05.000 --> 00:16:11.000
number of overflows so that we
can really start to see more
about

280
00:16:11.000 --> 00:16:16.000
what happened within a
particular year. We can look at

281
00:16:16.000 --> 00:16:22.000
certain months within several
years for the

282
00:16:22.000 --> 00:16:26.000
whole dataset. That is why that
amortization can really

283
00:16:26.000 --> 00:16:31.934
come into place.

284
00:16:31.934 --> 00:16:36.934
We could if we wanted to, we can
start by just

285
00:16:36.934 --> 00:16:40.934
seeing it on the screen. You
will see some of

286
00:16:40.934 --> 00:16:42.934
that as I walk through the
scenarios.

287
00:16:42.934 --> 00:16:48.934
>> Finally, we have a bar chart
that is going to

288
00:16:48.934 --> 00:16:59.000
show you the number of
overflows. I can also

289
00:16:59.000 --> 00:17:03.000
change this parameter to show
municipality, facility, the
collection system,

290
00:17:03.000 --> 00:17:09.000
the watershed. Keep in mine you
are saying

291
00:17:09.000 --> 00:17:18.000
anonymize labels. These are

292
00:17:18.000 --> 00:17:19.000
intentionally anonymize.

293
00:17:19.000 --> 00:17:25.000
I just also wanted to point out
we have some key

294
00:17:25.000 --> 00:17:30.934
performance cues. They give you
a sense of

295
00:17:30.934 --> 00:17:36.934
aggregate numbers for the entire
dataset. We are looking at

296
00:17:36.934 --> 00:17:44.934
about 20.000. it is actually
18.000.

297
00:17:44.934 --> 00:17:53.000
We can also use these filters to
look at a

298
00:17:53.000 --> 00:17:59.000
particular municipality. I can
do that by visually clicking

299
00:17:59.000 --> 00:18:04.000
and interacting with these
filters, which makes it

300
00:18:04.000 --> 00:18:05.000
very intuitive.

301
00:18:05.000 --> 00:18:10.000
>> Now that I have oriented you
to what is available on the
screen

302
00:18:10.000 --> 00:18:14.000
I will walk you through a couple
of

303
00:18:14.000 --> 00:18:17.000
scenarios. First of all, I want
to draw your attention to this
huge

304
00:18:17.000 --> 00:18:23.000
white bubble. You may have been
looking at it the

305
00:18:23.000 --> 00:18:28.934
whole time.

306
00:18:28.934 --> 00:18:34.934
We are going to dig into it. I
can

307
00:18:34.934 --> 00:18:43.934
simply click the map and I can
circle this. Just

308
00:18:43.934 --> 00:18:49.934
to make it clear I have now
selected a few postal codes that
I

309
00:18:49.934 --> 00:18:55.000
wanted to dig deeper into. I
filtered everything on the
screen

310
00:18:55.000 --> 00:19:01.000
by the postal code. No heat map
and my bar chart

311
00:19:01.000 --> 00:19:07.000
are all going to reflect that

312
00:19:07.000 --> 00:19:13.000
postal code. Right away I can
see the areas that I

313
00:19:13.000 --> 00:19:19.000
am interested in looking at. The
number of overflows appear

314
00:19:19.000 --> 00:19:25.000
to be decreasing. Overall the
trend

315
00:19:25.000 --> 00:19:30.934
downward is decreasing. It was
really difficult

316
00:19:30.934 --> 00:19:35.934
to put this analysis together. I
can stop there

317
00:19:35.934 --> 00:19:39.934
and say the problem seems to be
improving. I can think of some

318
00:19:39.934 --> 00:19:45.934
reasons why and stop there, but
since we have modern tools and

319
00:19:45.934 --> 00:19:50.934
the ability to dive deeper you
are not going to

320
00:19:50.934 --> 00:19:57.000
stop there. If we go to the two
dimensional heat map you can
kind

321
00:19:57.000 --> 00:20:03.000
of start to see how from 2006

322
00:20:03.000 --> 00:20:12.000
until about 2012 we have a lot
of overflows occurring. After
2012

323
00:20:12.000 --> 00:20:18.000
they started to increase a
little bit. We still have a
persistent

324
00:20:18.000 --> 00:20:23.000
seasonal problem in the late
spring and early summer. I

325
00:20:23.000 --> 00:20:26.000
can isolate these years by
visually

326
00:20:26.000 --> 00:20:32.934
interacting and it becomes even
more parents. There is still a

327
00:20:32.934 --> 00:20:41.934
problem here. This tells a very
different story.

328
00:20:41.934 --> 00:20:50.934
As someone who is potentially

329
00:20:50.934 --> 00:20:57.000
trying to fix the problems in
the area this could be really

330
00:20:57.000 --> 00:21:02.000
valuable information.
Furthermore we can look at the
causes associated

331
00:21:02.000 --> 00:21:08.000
with the trends.

332
00:21:08.000 --> 00:21:11.834
It is no surprise in this
particular geography that the
leading

333
00:21:11.834 --> 00:21:13.234
cause is precipitation.

334
00:21:13.234 --> 00:21:20.000
That makes sense because it is a
seasonal problem. We and totally

335
00:21:20.000 --> 00:21:24.000
knew that. But it might be
difficult to help others

336
00:21:24.000 --> 00:21:28.000
understand that in order to seek
funding for support in

337
00:21:28.000 --> 00:21:37.934
order to do something about the
problem. Now, within the and are

338
00:21:37.934 --> 00:21:42.934
of minutes you can see what is
going on in this area. I

339
00:21:42.934 --> 00:21:47.934
will draw your attention to the
bar chart. This makes a point
about

340
00:21:47.934 --> 00:21:54.000
data quality. We did not need
pristine data to get to

341
00:21:54.000 --> 00:22:00.000
this very powerful conclusion.
The second-leading cause of
water flow

342
00:22:00.000 --> 00:22:06.000
in this area. Looking at

343
00:22:06.000 --> 00:22:12.000
this as a separate category in
the dataset I am

344
00:22:12.000 --> 00:22:18.000
thinking what do they mean by
excessive flow?

345
00:22:18.000 --> 00:22:23.000
My analyst brain is going in so
many different directions.

346
00:22:23.000 --> 00:22:30.934
>> It does not matter. I just
need to have that analysis in my

347
00:22:30.934 --> 00:22:33.934
thought process in trying to
clean up these categories and
make them

348
00:22:33.934 --> 00:22:49.934
into something meaningful. I can
just drum my conclusions and
move on here.

349
00:23:33.934 --> 00:23:42.934
>> I do not see anything that
stands out to me as being

350
00:23:42.934 --> 00:23:45.934
obvious outliers.

351
00:23:45.934 --> 00:23:51.934
To me this is not an outlier
that I can properly

352
00:23:51.934 --> 00:23:57.000
rely on to make a conclusion.
With

353
00:23:57.000 --> 00:24:03.000
that example I just want to
check on time here

354
00:24:03.000 --> 00:24:06.000
really quickly.

355
00:24:06.000 --> 00:24:12.000
I am going to clear all of the
filters. You can see the
dashboard

356
00:24:12.000 --> 00:24:17.000
looks like it did in the very
beginning. I will draw your

357
00:24:17.000 --> 00:24:24.000
attention back to the map.

358
00:24:24.000 --> 00:24:28.067
Maybe I want to dive into this
area, I am not sure. Something
that

359
00:24:28.067 --> 00:24:37.000
struck as interesting is

360
00:24:37.000 --> 00:24:45.000
anything over 100 is terrible.

361
00:24:45.000 --> 00:24:50.000
Say I want to lower the
threshold and see what jumps out
at me, I

362
00:24:50.000 --> 00:24:56.000
can do that here. This cluster
turns much

363
00:24:56.000 --> 00:25:05.000
writer read and much more
profound. If I change

364
00:25:05.000 --> 00:25:10.000
it back it tells me very clearly
beyond

365
00:25:10.000 --> 00:25:14.000
this area and this area you have
a lot going

366
00:25:14.000 --> 00:25:20.000
on. Let's expand the map and see
what exactly is going on. I

367
00:25:20.000 --> 00:25:26.000
consume in. I can select each
one or just circle

368
00:25:26.000 --> 00:25:37.934
them to get a better sense.

369
00:25:37.934 --> 00:25:42.934
The dashboard looks

370
00:25:42.934 --> 00:25:47.934
much different. I can see right
off the bat that after 2011
there

371
00:25:47.934 --> 00:25:53.000
was a huge uptick for these
particular areas. That

372
00:25:53.000 --> 00:26:01.000
is also evident in the heat map.
This

373
00:26:01.000 --> 00:26:06.000
is where domain expertise comes
into

374
00:26:06.000 --> 00:26:11.000
place. It can be dirty and not
pristine, but then you have to
have someone

375
00:26:11.000 --> 00:26:20.000
challenging them and poking
holes. Let's say the main

376
00:26:20.000 --> 00:26:24.000
expert says you cannot look at
anything

377
00:26:24.000 --> 00:26:29.934
before 2012. I know for a fact
facilities just

378
00:26:29.934 --> 00:26:37.934
started recording their after
2012.

379
00:26:37.934 --> 00:26:43.934
Let's say we just want to look
at 2012. Anything

380
00:26:43.934 --> 00:26:53.000
before that we have specific
geo-locations.

381
00:26:53.000 --> 00:26:57.000
Now I am looking at recent
years. I can see that in

382
00:26:57.000 --> 00:27:06.000
this area there have obviously
been problems. Once again I can

383
00:27:06.000 --> 00:27:12.000
stop here and say I have two

384
00:27:12.000 --> 00:27:18.000
more dimensions to look at.
First of all I can look

385
00:27:18.000 --> 00:27:25.000
at cause. Let's break it down by
cause. It is very different from

386
00:27:25.000 --> 00:27:30.934
what we previously saw. If I
click on

387
00:27:30.934 --> 00:27:36.934
to Greece, I'm actually

388
00:27:36.934 --> 00:27:44.934
going to get out of this. It is
a little

389
00:27:44.934 --> 00:27:51.934
bit different when we are
looking at all overflows in this

390
00:27:51.934 --> 00:27:55.000
area.

391
00:27:55.000 --> 00:27:59.000
>> Let's say I am a facility and
I have been communicating

392
00:27:59.000 --> 00:28:05.000
with the public about adverse
effects of pouring grease down

393
00:28:05.000 --> 00:28:09.000
the drain and I am really trying
to get people to

394
00:28:09.000 --> 00:28:18.000
stop doing that and to be more
aware. Let's say I started

395
00:28:18.000 --> 00:28:23.000
that communication in 2016. You
can see a profound

396
00:28:23.000 --> 00:28:27.000
impact started in 2017. For
anyone wondering if

397
00:28:27.000 --> 00:28:29.934
we should reinvest in that
outreach program, was a really
related

398
00:28:29.934 --> 00:28:38.934
to that? It is just more and and
is to make the

399
00:28:38.934 --> 00:28:44.934
point clear. These tools can be
pretty

400
00:28:44.934 --> 00:28:50.934
powerful vehicles. Looking at
just a couple

401
00:28:50.934 --> 00:28:57.000
more things.

402
00:28:57.000 --> 00:29:03.000
The geographic areas affected by
these

403
00:29:03.000 --> 00:29:06.000
overflows are different. You are
trend is

404
00:29:06.000 --> 00:29:12.000
different. It is stationary and
not increasing.

405
00:29:12.000 --> 00:29:18.000
It is important that you are
able to quickly isolate these

406
00:29:18.000 --> 00:29:26.000
causes. Each one should be
looked at individually and

407
00:29:26.000 --> 00:29:32.934
addressed individually. I just
want to make one more point to
you all

408
00:29:32.934 --> 00:29:35.934
about the dirty data that we
have been mentioning

409
00:29:35.934 --> 00:29:39.934
again and again. Having a dirty
data

410
00:29:39.934 --> 00:29:47.934
set is different from exploring
and gaining value. I just popped

411
00:29:47.934 --> 00:29:57.000
open the our chart. This is a
free-form

412
00:29:57.000 --> 00:30:03.000
field obviously.

413
00:30:03.000 --> 00:30:08.000
Again, if I am looking at the

414
00:30:08.000 --> 00:30:14.000
data set and I haven't had a way
to Chris Ali

415
00:30:14.000 --> 00:30:20.000
-- quickly grasp, there is

416
00:30:20.000 --> 00:30:23.000
no way that we can gain
meaningful conclusions without

417
00:30:23.000 --> 00:30:28.934
going through every single one
and trying to group them
together and

418
00:30:28.934 --> 00:30:34.934
creating clear categories. That
would've been a

419
00:30:34.934 --> 00:30:38.934
terrible mistake. As we can see,
going back to what I was

420
00:30:38.934 --> 00:30:46.934
saying about directional
correctness, the trends

421
00:30:46.934 --> 00:30:54.000
are directionally correct and
they are very clear. We do not
need to

422
00:30:54.000 --> 00:30:56.000
worry about this. It does not
matter. It is

423
00:30:56.000 --> 00:31:02.000
the second and not the most
leading cause, especially within
the

424
00:31:02.000 --> 00:31:08.000
different geographies. The point
is we

425
00:31:08.000 --> 00:31:11.000
get very powerful conclusions.

426
00:31:11.000 --> 00:31:20.000
With that I will send it over to
our

427
00:31:20.000 --> 00:31:24.000
slides again.

428
00:31:24.000 --> 00:31:40.000
>> I am going to switch a view
on you.

429
00:31:40.000 --> 00:31:48.934
>> With that, I hope that I have
giving you an

430
00:31:48.934 --> 00:31:55.000
idea about what is possible. It
was an

431
00:31:55.000 --> 00:32:01.000
environmental dataset and we
hope that you are able to draw

432
00:32:01.000 --> 00:32:02.000
some parallels.

433
00:32:02.000 --> 00:32:05.000
We would like to put it to our
audience: how can we help
agencies with their

434
00:32:05.000 --> 00:32:14.000
interest and utilizing data
analytics?

435
00:32:14.000 --> 00:32:19.000
We are really asking you to
imagine the

436
00:32:19.000 --> 00:32:25.000
possibilities. If you can
consider other data sources out
there that

437
00:32:25.000 --> 00:32:32.934
can be analyzed in real time
things like assessing potential
risk,

438
00:32:32.934 --> 00:32:36.934
looking at physical and chemical
changes along different seasons

439
00:32:36.934 --> 00:32:42.934
or looking at potential impacts
from wildfires

440
00:32:42.934 --> 00:32:48.934
and studies. Any type of
analysis that you

441
00:32:48.934 --> 00:32:51.934
can imagine.

442
00:32:51.934 --> 00:32:54.000
We want the audience to consider
that.

443
00:32:54.000 --> 00:33:01.000
With that, I believe it is time
for questions.

444
00:33:01.000 --> 00:33:07.000
>> Thanks, Katie. This seems
like a really powerful tool. I
would

445
00:33:07.000 --> 00:33:10.000
love to hear from the
participants, if anyone has

446
00:33:10.000 --> 00:33:15.000
used the tool or has experience
with this type of

447
00:33:15.000 --> 00:33:21.000
data analytics we would love to
hear

448
00:33:21.000 --> 00:33:24.000
from you. If you have any
experience, please,

449
00:33:24.000 --> 00:33:31.934
respond in the Q&A section.
Katie, I do have one question
for you,

450
00:33:31.934 --> 00:33:35.934
can you find a little more
information about the software?
What exactly

451
00:33:35.934 --> 00:33:39.934
is the software that you are
using to do this?

452
00:33:39.934 --> 00:33:45.934
>> The software's call -- is one
of the leading

453
00:33:45.934 --> 00:33:55.000
that a software package is
available. Some folks might be

454
00:33:55.000 --> 00:34:01.000
familiar with Power

455
00:34:01.000 --> 00:34:05.000
BI they allow you to explore
large datasets

456
00:34:05.000 --> 00:34:08.000
in real time with minimal
effort.

457
00:34:08.000 --> 00:34:12.000
>> Thanks. I have a question
kind of specific to Mine and
Mineral

458
00:34:12.000 --> 00:34:18.000
processing. It looks like you
use a bunch of data that came
from essentially

459
00:34:18.000 --> 00:34:22.000
a single source. What if we
wanted to bring

460
00:34:22.000 --> 00:34:26.000
in another dataset? You showed
that sewer overflow

461
00:34:26.000 --> 00:34:32.000
was a leading cause under one
set of scenarios, could we

462
00:34:32.000 --> 00:34:36.000
pulled in for example
civilization data from

463
00:34:36.000 --> 00:34:39.000
this area?

464
00:34:39.000 --> 00:34:43.000
>> You absolutely can. It is so
funny that you

465
00:34:43.000 --> 00:34:47.000
should ask. We had attempted to
do that at one point. We did
come

466
00:34:47.000 --> 00:34:52.000
up with a pretty good version of
drawing parallels

467
00:34:52.000 --> 00:34:58.000
between the two data sets. We
decided to just leave

468
00:34:58.000 --> 00:35:02.000
this one, just for the point of
demonstrating one

469
00:35:02.000 --> 00:35:11.000
dataset. As long as we have two
common points you can put

470
00:35:11.000 --> 00:35:14.000
them together.

471
00:35:14.000 --> 00:35:19.000
You can totally bring in as many
as

472
00:35:19.000 --> 00:35:20.000
you have.

473
00:35:20.000 --> 00:35:26.000
>> That is exactly the process
that we look to do,

474
00:35:26.000 --> 00:35:30.934
find the directionally correct
trend and look for these

475
00:35:30.934 --> 00:35:34.934
external sources of data, the
weather is

476
00:35:34.934 --> 00:35:39.934
one of the ones that is often at
the top of the list. There are
other

477
00:35:39.934 --> 00:35:45.934
data aggregators that can
provide other data. That is

478
00:35:45.934 --> 00:35:50.934
a big part of it in a
straightforward process

479
00:35:50.934 --> 00:35:57.000
with any of these tools.

480
00:35:57.000 --> 00:35:59.000
>> We do have a couple more
questions. We will come back
around

481
00:35:59.000 --> 00:36:05.000
to those where we have
additional time at the end.
Right now I would

482
00:36:05.000 --> 00:36:13.000
like to move to our next
speaker. Thank you, Katie and

483
00:36:13.000 --> 00:36:14.000
Doug. That was really
interesting.

484
00:36:14.000 --> 00:36:23.000
Our next speaker will talk about
medal transport

485
00:36:23.000 --> 00:36:30.934
in watersheds.

486
00:36:30.934 --> 00:36:36.934
>> Hello, everyone.

487
00:36:36.934 --> 00:36:42.934
As mentioned before I will talk
about Metal

488
00:36:42.934 --> 00:36:48.934
partitioning and I also will
talk about large

489
00:36:48.934 --> 00:36:55.000
datasets and this

490
00:36:55.000 --> 00:37:01.000
particular dataset is actually
generated in

491
00:37:01.000 --> 00:37:10.000
our labs. My name is Dr.

492
00:37:10.000 --> 00:37:15.000
Al-Abed. I am located in
Cincinnati, Ohio. Today the talk

493
00:37:15.000 --> 00:37:21.000
about the Metal partitioning
with a focus

494
00:37:21.000 --> 00:37:26.000
on Superfund sites have become

495
00:37:26.000 --> 00:37:29.934
really important in these large
watersheds affected by

496
00:37:29.934 --> 00:37:35.934
such problems. How do we study
the transport in such

497
00:37:35.934 --> 00:37:39.934
watersheds and we will show some
examples of how we collect

498
00:37:39.934 --> 00:37:45.934
the data and then

499
00:37:45.934 --> 00:37:51.934
illustrate why a large dataset
is important

500
00:37:51.934 --> 00:37:58.000
to select appropriate options of
best

501
00:37:58.000 --> 00:38:07.000
management practices.

502
00:38:07.000 --> 00:38:13.000
>> Contamination is a common
problem in

503
00:38:13.000 --> 00:38:19.000
the US and worldwide. Today my
example is here in

504
00:38:19.000 --> 00:38:25.000
the US. More than half 1 million
abandoned mine sites

505
00:38:25.000 --> 00:38:28.934
west of Mississippi. There are a
lot

506
00:38:28.934 --> 00:38:37.934
of others with lots of issues.
Mining operations

507
00:38:37.934 --> 00:38:41.934
have less oars, waste

508
00:38:41.934 --> 00:38:47.934
materials and exposing these
materials to air, water, and the
climate

509
00:38:47.934 --> 00:38:53.000
in general

510
00:38:53.000 --> 00:38:58.000
has caused a lot of weathering,
physical

511
00:38:58.000 --> 00:39:04.000
or chemical. This weathering
ends up affecting

512
00:39:04.000 --> 00:39:10.000
whether the contamination, in
this case

513
00:39:10.000 --> 00:39:19.000
the Metal contamination is
particulate or dissolved

514
00:39:19.000 --> 00:39:22.000
Metal.

515
00:39:22.000 --> 00:39:26.000
The way that we use the water,
rather it

516
00:39:26.000 --> 00:39:34.934
is for drinking or fishery or

517
00:39:34.934 --> 00:39:40.934
even irrigation. I find this

518
00:39:40.934 --> 00:39:46.934
is challenging in many aspects.

519
00:39:46.934 --> 00:39:53.000
The change of the climate, the
change of the

520
00:39:53.000 --> 00:39:57.000
different hydrology's in these
watersheds, it does affect

521
00:39:57.000 --> 00:40:03.000
the water treatment design that
is being used and also

522
00:40:03.000 --> 00:40:11.000
what kind of aspect we could do
for these

523
00:40:11.000 --> 00:40:17.000
different sites.

524
00:40:17.000 --> 00:40:26.000
This is a good example to show
you.

525
00:40:26.000 --> 00:40:31.934
You see these deposits or
mineral

526
00:40:31.934 --> 00:40:36.934
processing waste. If you look at
is everything

527
00:40:36.934 --> 00:40:42.934
pretty much dead. The landscape
looks like the moon.

528
00:40:42.934 --> 00:40:48.934
Once that is removed then you
expose what is

529
00:40:48.934 --> 00:40:56.000
below it that is because the
topsoil is

530
00:40:56.000 --> 00:40:59.000
all gone.

531
00:40:59.000 --> 00:41:05.000
That change does affect the
mineralogy of the

532
00:41:05.000 --> 00:41:12.000
waste and how aggressive some

533
00:41:12.000 --> 00:41:17.000
of these materials are being
exposed into

534
00:41:17.000 --> 00:41:23.000
the environment. Why do we need
metal partitioning? This is

535
00:41:23.000 --> 00:41:31.934
the issue that is really
important since there are many
rivers

536
00:41:31.934 --> 00:41:37.934
in these watersheds and creeks.
These could

537
00:41:37.934 --> 00:41:42.934
actually become a transport
mechanism for a lot of this
metal contamination

538
00:41:42.934 --> 00:41:48.934
through the watershed. One

539
00:41:48.934 --> 00:41:51.934
form or another, dissolve or
particulate

540
00:41:51.934 --> 00:41:57.000
could become extremely important
in

541
00:41:57.000 --> 00:42:02.000
looking at long-term data. That
will help us in finding
solutions.

542
00:42:02.000 --> 00:42:07.000
Dissolved metals can be
difficult to remove and
typically requires

543
00:42:07.000 --> 00:42:15.000
a lot of treatment. It requires
chemical

544
00:42:15.000 --> 00:42:21.000
and biological.

545
00:42:21.000 --> 00:42:26.000
It kind of

546
00:42:26.000 --> 00:42:31.934
depends on these waterways
within the watershed.

547
00:42:31.934 --> 00:42:37.934
It requires a different type of

548
00:42:37.934 --> 00:42:45.934
management practice.

549
00:42:45.934 --> 00:42:50.934
>> In finding solutions, why do
we need a

550
00:42:50.934 --> 00:42:57.000
large dataset? That is one of
the important assets

551
00:42:57.000 --> 00:43:03.000
that relate to the previous talk
about

552
00:43:03.000 --> 00:43:08.000
the necessity of having
long-range data

553
00:43:08.000 --> 00:43:14.000
sets. In this case the
statistics tell us large sample
sizes

554
00:43:14.000 --> 00:43:20.000
and the offense,

555
00:43:20.000 --> 00:43:26.000
the smaller margin of area --

556
00:43:26.000 --> 00:43:31.934
error we will have including
certain parameters. Those

557
00:43:31.934 --> 00:43:37.934
changes will help us

558
00:43:37.934 --> 00:43:41.934
in deciding how we go about
finding solutions. Often times
we

559
00:43:41.934 --> 00:43:45.934
need to validate that assumption
to avoid

560
00:43:45.934 --> 00:43:50.934
costly errors. Trial and error

561
00:43:50.934 --> 00:43:55.000
in a large area like this based
on very small datasets could be

562
00:43:55.000 --> 00:44:01.000
very costly and has a

563
00:44:01.000 --> 00:44:10.000
long range effect on the
environment's and the

564
00:44:10.000 --> 00:44:19.000
developments of these particular
sites.

565
00:44:19.000 --> 00:44:25.000
In finding these kinds of
solutions, this particular

566
00:44:25.000 --> 00:44:27.000
cartoon illustrates what is a
part of the

567
00:44:27.000 --> 00:44:33.000
watershed that we actually need
to look at.

568
00:44:33.000 --> 00:44:42.000
It presents the physical
mechanisms that involve the
metal

569
00:44:42.000 --> 00:44:48.000
transport. These are an example
of a

570
00:44:48.000 --> 00:44:54.000
chat pile. Anything that could
be dragged into

571
00:44:54.000 --> 00:45:00.000
the streams, it could either
precipitate it,

572
00:45:00.000 --> 00:45:06.000
it could be partitioning to
resolve metals or also stay

573
00:45:06.000 --> 00:45:11.000
as articulate, but a smaller
size could actually be
transported

574
00:45:11.000 --> 00:45:21.000
from one area to another. This
is a very

575
00:45:21.000 --> 00:45:27.000
dynamic process. Most of these
metals could become

576
00:45:27.000 --> 00:45:32.934
bioavailable and it could
actually

577
00:45:32.934 --> 00:45:37.934
affect human health. The metal
concentration

578
00:45:37.934 --> 00:45:43.934
is typically used to determine
the metal

579
00:45:43.934 --> 00:45:49.934
mobility. The solid sentiment --

580
00:45:49.934 --> 00:45:58.000
sediment concentration, the
reason for

581
00:45:58.000 --> 00:46:05.000
those particular sediments, they
transport much either.

582
00:46:05.000 --> 00:46:11.000
The difference between total and
dissolved concentration in the

583
00:46:11.000 --> 00:46:15.000
water and the suspended
particles traveling in the
stream. When using

584
00:46:15.000 --> 00:46:23.000
these parameters we have more
complete pictures of how to

585
00:46:23.000 --> 00:46:27.000
make decisions about the
remediation of the watershed.
Now

586
00:46:27.000 --> 00:46:33.934
we are moving to the example
that I would like to

587
00:46:33.934 --> 00:46:39.934
present here. This is an area,
we

588
00:46:39.934 --> 00:46:43.934
call it the tri-state mining
district. Today I'm just

589
00:46:43.934 --> 00:46:50.934
going to talk about Missouri and
Kansas. Oklahoma is also a part

590
00:46:50.934 --> 00:46:59.000
of this. In this study there are
several tributaries

591
00:46:59.000 --> 00:47:05.000
or creeks into this particular
watershed that end up in a lake

592
00:47:05.000 --> 00:47:09.000
that you see over here in this

593
00:47:09.000 --> 00:47:12.000
flat circle.

594
00:47:12.000 --> 00:47:16.000
>> Most of the water that is
collected in these watersheds
where a lot

595
00:47:16.000 --> 00:47:20.000
of these mine blowing either
still

596
00:47:20.000 --> 00:47:24.000
exist or have been removed. Most
of the water is being collected

597
00:47:24.000 --> 00:47:32.934
and end up in this

598
00:47:32.934 --> 00:47:38.934
particular lake. The objective
here is

599
00:47:38.934 --> 00:47:44.934
to monitor the zinc transport.
In this case

600
00:47:44.934 --> 00:47:50.934
we are concentrating

601
00:47:50.934 --> 00:47:57.000
on other metals associated with
this

602
00:47:57.000 --> 00:47:59.000
particular site.

603
00:47:59.000 --> 00:48:05.000
For the time that we have for
this talk will talk

604
00:48:05.000 --> 00:48:10.000
about zinc.

605
00:48:10.000 --> 00:48:16.000
>> The sample collected is done
twice

606
00:48:16.000 --> 00:48:20.000
a month. We are still continuing
with the monitoring, but the
data

607
00:48:20.000 --> 00:48:28.934
that I will show you today is
for that timeframe.

608
00:48:28.934 --> 00:48:34.934
They were all measured for this

609
00:48:34.934 --> 00:48:37.934
particular work.

610
00:48:37.934 --> 00:48:43.934
On the left here this can

611
00:48:43.934 --> 00:48:48.934
show you the average

612
00:48:48.934 --> 00:48:54.000
discharged sediments for each of
these streams. The discharge is

613
00:48:54.000 --> 00:48:58.000
calculated by the suspended
sediment concentration

614
00:48:58.000 --> 00:49:04.000
and the flow rate. This is for
the whole watershed, all of the

615
00:49:04.000 --> 00:49:13.000
creeks that you see.

616
00:49:13.000 --> 00:49:19.000
All of these tributaries have
been collecting a lot of

617
00:49:19.000 --> 00:49:21.000
these sediments.

618
00:49:21.000 --> 00:49:27.000
The graph on the right shows the
same concentration. As

619
00:49:27.000 --> 00:49:33.934
you see very clearly there is an
partitioning and there is

620
00:49:33.934 --> 00:49:37.934
also zinc in the suspended
sediment.

621
00:49:37.934 --> 00:49:43.934
What does that tell us? It tells
us that each of these

622
00:49:43.934 --> 00:49:49.934
particular creeks have certain
dynamics a son either physical

623
00:49:49.934 --> 00:49:56.000
or also the chemistry of

624
00:49:56.000 --> 00:50:03.000
these particulates and rather
they partition to dissolve or is

625
00:50:03.000 --> 00:50:14.000
associated with the suspended
loading.

626
00:50:14.000 --> 00:50:17.000
>> Now we moved to zinc in the
water.

627
00:50:17.000 --> 00:50:26.000
This is the average for the
dissolve

628
00:50:26.000 --> 00:50:29.934
and total in stream sediments.
As you see, the

629
00:50:29.934 --> 00:50:35.934
average is two very based on the
Creek we are talking

630
00:50:35.934 --> 00:50:40.934
about. It does not tell us about
variability. It is very easy

631
00:50:40.934 --> 00:50:46.934
to look at this and say this has
major

632
00:50:46.934 --> 00:50:51.934
issues, but it does not really
show why these changes are
happening

633
00:50:51.934 --> 00:50:56.000
without a larger dataset.

634
00:50:56.000 --> 00:51:00.000
That is what you see on the
right side. We show you the
variability

635
00:51:00.000 --> 00:51:09.000
that exists for these

636
00:51:09.000 --> 00:51:25.000
particular creeks.

637
00:51:25.000 --> 00:51:33.934
How much these variations will
affect the decisions we need

638
00:51:33.934 --> 00:51:39.934
to make. This graph zero's in on
this

639
00:51:39.934 --> 00:51:45.934
particular Creek. I took the
same graph that you saw earlier.
As

640
00:51:45.934 --> 00:51:50.934
you see, the variation through
time is

641
00:51:50.934 --> 00:51:57.000
very prominent. That is another
advantage of a large dataset
that

642
00:51:57.000 --> 00:52:09.000
we need to look at.

643
00:52:09.000 --> 00:52:15.000
Here is another example where
you see the variability that
does

644
00:52:15.000 --> 00:52:20.000
occur between total zinc and the
dissolved

645
00:52:20.000 --> 00:52:25.000
zinc. This dataset shows a a lot
of variability. When we narrow
it

646
00:52:25.000 --> 00:52:30.934
down we wanted to show is it
suspended sediment or a

647
00:52:30.934 --> 00:52:39.934
sink content? Sometimes they
line up. Sometimes they

648
00:52:39.934 --> 00:52:43.934
don't. That means if they do

649
00:52:43.934 --> 00:52:51.934
not, like in this case that
could be the dissolve fraction
for this

650
00:52:51.934 --> 00:52:58.000
particular Creek. Now we are
going to move, we now know more

651
00:52:58.000 --> 00:53:04.000
about these creeks and how they
behave and what is

652
00:53:04.000 --> 00:53:08.000
the zinc contamination. So what
are we going to do about that?
This

653
00:53:08.000 --> 00:53:14.000
is where this slide shows since

654
00:53:14.000 --> 00:53:19.000
we did understand the
partitioning between sediments,

655
00:53:19.000 --> 00:53:26.000
exposed surfaces that there are
several options that could

656
00:53:26.000 --> 00:53:31.934
be used to actually help remove
some of

657
00:53:31.934 --> 00:53:37.934
these contaminants from the
water. One of those

658
00:53:37.934 --> 00:53:41.934
could be sediment

659
00:53:41.934 --> 00:53:47.934
traps, bioreactors, it could be
several things. Today

660
00:53:47.934 --> 00:53:54.000
I am going to focus on using

661
00:53:54.000 --> 00:53:57.000
the biochar.

662
00:53:57.000 --> 00:54:01.000
>> To do this you cannot do
trial

663
00:54:01.000 --> 00:54:07.000
and error. You have to take the
water chemistry

664
00:54:07.000 --> 00:54:13.000
specifically for the water
stream. In this

665
00:54:13.000 --> 00:54:19.000
case we picked the short Creek,
it has the higher concentration,

666
00:54:19.000 --> 00:54:22.000
just to give you an example. To
do this you have

667
00:54:22.000 --> 00:54:31.000
to do some bench work and then
you move into the pilot and

668
00:54:31.000 --> 00:54:37.000
eventually field scale.

669
00:54:37.000 --> 00:54:43.000
To do this for the biochar,
again we needed to,

670
00:54:43.000 --> 00:54:48.000
if we pick any of these
materials we need to understand

671
00:54:48.000 --> 00:54:54.000
more about what they are and how
they are made. A bio chart

672
00:54:54.000 --> 00:55:03.000
is regulated carbon products. It
is sometimes even

673
00:55:03.000 --> 00:55:08.000
found naturally. It has been

674
00:55:08.000 --> 00:55:14.000
used as [ Indiscernible ] with
very hyper awfully. The core
volume is

675
00:55:14.000 --> 00:55:22.000
quite large for absorptive
material. It could be

676
00:55:22.000 --> 00:55:33.934
considered as beneficial use of
ways. You

677
00:55:33.934 --> 00:55:39.934
could take a lot of the

678
00:55:39.934 --> 00:55:44.934
burn forests, large forest fires
for

679
00:55:44.934 --> 00:55:48.934
example and that could generate
a lot of

680
00:55:48.934 --> 00:55:54.000
biochar naturally that could be
used for this

681
00:55:54.000 --> 00:55:57.000
particular purpose.

682
00:55:57.000 --> 00:56:02.000
>> Now we move into if this
material does

683
00:56:02.000 --> 00:56:08.000
work how do we apply it? This
cartoon takes that cross-section

684
00:56:08.000 --> 00:56:14.000
that I showed you earlier. You
could either use

685
00:56:14.000 --> 00:56:18.000
it as traffic for reaching the
Creek.

686
00:56:18.000 --> 00:56:24.000
It could be what we call teabags
or the more

687
00:56:24.000 --> 00:56:29.934
technical term, instream
absorption backs. These are

688
00:56:29.934 --> 00:56:32.934
usually materials that you put
in these bags that are made

689
00:56:32.934 --> 00:56:38.934
from geosynthetic materials that
is porous. It allows a contact

690
00:56:38.934 --> 00:56:50.934
between the bio chart and the
contaminated water.

691
00:56:50.934 --> 00:56:56.000
>> What is the characteristic of
these

692
00:56:56.000 --> 00:56:59.000
different biochar's? We look at
these different ones from
different

693
00:56:59.000 --> 00:57:05.000
manufacturers and the cost. That
is why

694
00:57:05.000 --> 00:57:11.000
in any engineering approach you
have

695
00:57:11.000 --> 00:57:16.000
to include the cost of the
material and how much you are
going to use

696
00:57:16.000 --> 00:57:20.000
and what purpose. Are they going
to be effective

697
00:57:20.000 --> 00:57:24.000
for how long? This is the
density and the porosity.

698
00:57:24.000 --> 00:57:31.934
You can see they very between
the surface area, porosity and
so

699
00:57:31.934 --> 00:57:36.934
on. Overall we all feel they are
applicable for the purpose

700
00:57:36.934 --> 00:57:39.934
that we are trying to do. To do
this

701
00:57:39.934 --> 00:57:45.934
test, one of the things that we
did in this initial

702
00:57:45.934 --> 00:57:50.934
experience is we used a batch
test to find

703
00:57:50.934 --> 00:57:58.000
the amount of the biochar to
absorb the ink. We use water
from the short

704
00:57:58.000 --> 00:58:04.000
Creek at pH 7. As you can see,

705
00:58:04.000 --> 00:58:08.000
when the Creek goes flat there
is no benefit of adding more
solvent.

706
00:58:08.000 --> 00:58:14.000
That means it is optimal before
he

707
00:58:14.000 --> 00:58:19.000
gets into the flat area.

708
00:58:19.000 --> 00:58:25.000
The Douglas for biochar
performed the best in

709
00:58:25.000 --> 00:58:28.000
this case.

710
00:58:28.000 --> 00:58:33.934
It just show the final pH is
8.4. good

711
00:58:33.934 --> 00:58:37.934
for keeping the zinc in the
precipitator form rather

712
00:58:37.934 --> 00:58:43.934
than the resolve. When we moved

713
00:58:43.934 --> 00:58:48.934
to the Connecticut test overtime
we could see that

714
00:58:48.934 --> 00:58:55.000
it did reach equilibrium around
24 hours. It is

715
00:58:55.000 --> 00:59:03.000
very quick and then it flattened
out. That means in any

716
00:59:03.000 --> 00:59:09.000
design for recycling of these
materials

717
00:59:09.000 --> 00:59:15.000
it depends on the concentration
and how fast it is going

718
00:59:15.000 --> 00:59:21.000
to absorb contamination from the
water.

719
00:59:21.000 --> 00:59:25.000
In this case it is about 24
hours. When we did the
isotherm's

720
00:59:25.000 --> 00:59:30.934
on this, which is really
important to determine the

721
00:59:30.934 --> 00:59:36.934
absorption capacity, again, some
of these bio

722
00:59:36.934 --> 00:59:41.934
charts stood up for higher

723
00:59:41.934 --> 00:59:47.934
absorption capacity and some are
less. Some of them were

724
00:59:47.934 --> 00:59:57.000
very low. That is another
decision for

725
00:59:57.000 --> 01:00:03.000
remediation. Finally, we wanted
to know

726
01:00:03.000 --> 01:00:08.000
not only did this dissolve, but

727
01:00:08.000 --> 01:00:14.000
also particulates. We did an
experiment where we used

728
01:00:14.000 --> 01:00:20.000
dissolve particular --
particulate of

729
01:00:20.000 --> 01:00:24.000
zinc only. The way we
distinguished is by changing

730
01:00:24.000 --> 01:00:28.934
the pH. We also changed

731
01:00:28.934 --> 01:00:33.934
the concentration. As you see
here, there is a

732
01:00:33.934 --> 01:00:39.934
reduction in the core volume of
this reticular

733
01:00:39.934 --> 01:00:45.934
bio forest. By calculating the
area for each

734
01:00:45.934 --> 01:00:51.934
of these, what this is telling
us is

735
01:00:51.934 --> 01:00:57.000
this reduction is not only
taking up

736
01:00:57.000 --> 01:01:04.000
the zinc as dissolve, it also
particulate.

737
01:01:04.000 --> 01:01:10.000
That means there is a good
option here for using these
materials and

738
01:01:10.000 --> 01:01:16.000
this approach for these
particular absorptive

739
01:01:16.000 --> 01:01:25.000
materials in instream. Given all
of

740
01:01:25.000 --> 01:01:30.934
that information, we have to
look in

741
01:01:30.934 --> 01:01:37.934
terms of how much data could we
actually collect

742
01:01:37.934 --> 01:01:43.934
to understand contamination and
how did

743
01:01:43.934 --> 01:01:46.934
they partition?

744
01:01:46.934 --> 01:01:53.000
And then be innovative enough
after

745
01:01:53.000 --> 01:01:58.000
understanding the condition of
these contaminated

746
01:01:58.000 --> 01:02:07.000
Superfund sites is to be
innovative enough to choose
different

747
01:02:07.000 --> 01:02:13.000
test management practices. In
this case we use the

748
01:02:13.000 --> 01:02:22.000
bio charts to actually start
thinking about these

749
01:02:22.000 --> 01:02:26.000
problems in -- you

750
01:02:26.000 --> 01:02:31.934
cannot apply the same thing
everywhere on every

751
01:02:31.934 --> 01:02:37.934
site. Collecting specific

752
01:02:37.934 --> 01:02:42.934
datasets for a specific site is
the right way of approaching the

753
01:02:42.934 --> 01:02:48.934
problem and finding solutions.
Finally, I would

754
01:02:48.934 --> 01:02:55.000
like to recognize a lot of my
colleagues

755
01:02:55.000 --> 01:03:11.000
who helped out with this
project.

756
01:03:25.000 --> 01:03:30.934
I would also like to recognize
Steve Camp

757
01:03:30.934 --> 01:03:36.934
from region 7 and Todd Campbell.
And also

758
01:03:36.934 --> 01:03:42.934
Robert Weber. Thank you.

759
01:03:42.934 --> 01:03:46.934
>> Thank you, that was good. We
did have a couple of

760
01:03:46.934 --> 01:03:53.000
questions, a couple related to
biochar. You referred to one

761
01:03:53.000 --> 01:03:58.000
of the product as beetle kill,
can you explain

762
01:03:58.000 --> 01:04:01.000
what that is?

763
01:04:01.000 --> 01:04:07.000
>> The origin of these actually

764
01:04:07.000 --> 01:04:13.000
came from the USDA.

765
01:04:13.000 --> 01:04:19.000
They collect these

766
01:04:19.000 --> 01:04:25.000
in the different farm operations
where

767
01:04:25.000 --> 01:04:31.000
they have waste and then they
heated at

768
01:04:31.000 --> 01:04:35.000
different temperatures. This
particular one, I don't have it

769
01:04:35.000 --> 01:04:41.000
in front of me, but it is a
specific temperature

770
01:04:41.000 --> 01:04:47.000
being used for generating this
ticket a

771
01:04:47.000 --> 01:04:50.000
bio chart.

772
01:04:50.000 --> 01:04:56.000
>> Okay. Thank you. And of a
follow-on

773
01:04:56.000 --> 01:04:59.000
question, is biochar available?
Where do you

774
01:04:59.000 --> 01:05:03.000
find this type of interior? Is
it

775
01:05:03.000 --> 01:05:09.000
readily available for folks who
might want to investigate?

776
01:05:09.000 --> 01:05:13.000
>> Yes. It is commercially
available with some of them.
Some

777
01:05:13.000 --> 01:05:21.000
of them are generated by certain
agencies

778
01:05:21.000 --> 01:05:27.000
as research. Most of them are

779
01:05:27.000 --> 01:05:29.934
commercially available.

780
01:05:29.934 --> 01:05:37.934
That is why mentioned cost. Cost
play a big role.

781
01:05:37.934 --> 01:05:42.934
For example, you use that high
temperature and that means more
energy and that

782
01:05:42.934 --> 01:05:49.934
means the cost is going to go
up.

783
01:05:49.934 --> 01:05:55.000
>> Okay. And then a final
question, in your absorption

784
01:05:55.000 --> 01:05:59.000
test did you say you use water
from the site

785
01:05:59.000 --> 01:06:05.000
or did you prepare laboratory
solutions mimicking those water

786
01:06:05.000 --> 01:06:08.000
quality components?

787
01:06:08.000 --> 01:06:17.000
>> The water that we used was
water from

788
01:06:17.000 --> 01:06:23.000
the creek. The reason we did
have more than added to it

789
01:06:23.000 --> 01:06:28.000
was because we wanted to test
the

790
01:06:28.000 --> 01:06:31.934
capacity at a higher zinc
loading.

791
01:06:31.934 --> 01:06:36.934
Some of the loadings, we did see
them at least from our data

792
01:06:36.934 --> 01:06:42.934
that you saw earlier on the
graphs. It is a

793
01:06:42.934 --> 01:06:51.934
combination of the actual
watershed data.

794
01:06:51.934 --> 01:06:58.000
In addition it is water in
certain treatments.

795
01:06:58.000 --> 01:07:03.000
>> All right. Thank you very
much. If folks have additional

796
01:07:03.000 --> 01:07:04.000
questions, please, post them
online.

797
01:07:04.000 --> 01:07:13.000
We will get back around to them
if we have time at the end.

798
01:07:13.000 --> 01:07:17.000
Thank you.

799
01:07:17.000 --> 01:07:21.000
>> I will be the last presenter.
I am going to talk

800
01:07:21.000 --> 01:07:27.000
about a project that we did for
EPA

801
01:07:27.000 --> 01:07:29.934
region 8.

802
01:07:29.934 --> 01:07:33.934
What we did is prepare a
three-dimensional visualization

803
01:07:33.934 --> 01:07:37.934
and analysis for a site out near

804
01:07:37.934 --> 01:07:41.934
Breckenridge, Colorado. We
called it the

805
01:07:41.934 --> 01:07:47.934
French coach site because it was
on French goals. I am going to
talk

806
01:07:47.934 --> 01:07:53.000
about a little bit about the
background as well as some of

807
01:07:53.000 --> 01:07:58.000
the project needs and some of
the things the project team
wanted

808
01:07:58.000 --> 01:08:00.000
to see out of this
visualization.

809
01:08:00.000 --> 01:08:06.000
Most of the time I will spend
today is on the datasets

810
01:08:06.000 --> 01:08:10.000
themselves, what data was
available for the

811
01:08:10.000 --> 01:08:19.000
site layout and features as well
as the water chemistry

812
01:08:19.000 --> 01:08:23.000
and hydrology. As a talk about
these I will talk about some of

813
01:08:23.000 --> 01:08:24.000
the datasets, the origin of the
datasets as well as the
challenges

814
01:08:24.000 --> 01:08:29.934
that we face with using some of
these datasets and some

815
01:08:29.934 --> 01:08:34.934
of the solutions that we found.
Finally we will take a

816
01:08:34.934 --> 01:08:39.934
look at the final product. For
we get going

817
01:08:39.934 --> 01:08:43.934
I would like to give a big
thanks to cascade technical

818
01:08:43.934 --> 01:08:50.934
service who provided and
prepared the 3-D visualization
for

819
01:08:50.934 --> 01:08:57.000
us. They use a product called
EVS, which is operated

820
01:08:57.000 --> 01:09:13.000
by CTEC studios.

821
01:09:57.000 --> 01:10:01.000
>> The folks at the city up
Breckenridge had developed

822
01:10:01.000 --> 01:10:06.000
a little bit in this floodplain
area

823
01:10:06.000 --> 01:10:12.000
by 1982. Here is a photograph of
the site area that we were

824
01:10:12.000 --> 01:10:18.000
working in. You can see some of

825
01:10:18.000 --> 01:10:22.000
the infrastructure, the shaft
locations and the dredged area
as well as

826
01:10:22.000 --> 01:10:26.000
the stream channel. Zooming out
a

827
01:10:26.000 --> 01:10:31.934
little bit let me show you a
couple other interesting things.
Again,

828
01:10:31.934 --> 01:10:35.934
here is French Creek. You can
see the

829
01:10:35.934 --> 01:10:40.934
dredged area. You can see where
we placed the

830
01:10:40.934 --> 01:10:47.934
mine workings. It is also
interesting to note this is

831
01:10:47.934 --> 01:10:50.934
a more recent photograph and the
city of Breckenridge has
developed

832
01:10:50.934 --> 01:10:57.000
much closer to the mine. There
is a

833
01:10:57.000 --> 01:11:01.000
potential impact to these

834
01:11:01.000 --> 01:11:05.000
homes. The interesting thing
that we found, when you look at
the close-up

835
01:11:05.000 --> 01:11:08.000
you can see the dredged
direction came

836
01:11:08.000 --> 01:11:14.000
up the stream Valley, turned
around and started heading back.
If you

837
01:11:14.000 --> 01:11:18.000
look really closely you will see
the dredged

838
01:11:18.000 --> 01:11:23.000
came to rest and a small
drainage pond. We thought

839
01:11:23.000 --> 01:11:28.934
that was pretty cool and
historic looking.

840
01:11:28.934 --> 01:11:31.934
>> The site we are looking at is
a combination of two mines that

841
01:11:31.934 --> 01:11:40.934
eventually grew together. They
started out mining lead

842
01:11:40.934 --> 01:11:45.934
and zinc. They operated until
the 1970s. Most of the mining
was

843
01:11:45.934 --> 01:11:49.934
completed by the 1950s. Just

844
01:11:49.934 --> 01:11:56.000
about 12 miles, most of it was
below the

845
01:11:56.000 --> 01:11:57.000
valley floor. They use a
combination of underground
mining as well as

846
01:11:57.000 --> 01:12:03.000
plaster mining in the stream to
pull

847
01:12:03.000 --> 01:12:08.000
things out. They also produced a
little copper, silver,

848
01:12:08.000 --> 01:12:12.000
and gold. There were a lot of
previous investigations

849
01:12:12.000 --> 01:12:18.000
that concluded the mine pool was
the primary source of

850
01:12:18.000 --> 01:12:24.000
contaminants leading into the
French Creek. There is a water
treatment

851
01:12:24.000 --> 01:12:27.000
plant that treats some of these.

852
01:12:27.000 --> 01:12:32.934
It treats the acid rock drainage
that

853
01:12:32.934 --> 01:12:38.934
is collected. The surface and
groundwater are connected
through the mine workings

854
01:12:38.934 --> 01:12:45.934
and a series of fractured rock.
That made understanding

855
01:12:45.934 --> 01:12:50.934
of the chemistry pretty
complicated. The other thing

856
01:12:50.934 --> 01:12:57.000
we found interesting was that
originally the

857
01:12:57.000 --> 01:13:03.000
more bodies were extracted from
these mineralized

858
01:13:03.000 --> 01:13:10.000
zones. These are no longer ORO
bodies worth money,

859
01:13:10.000 --> 01:13:22.000
but they are sources of
contamination.

860
01:13:22.000 --> 01:13:28.000
>> The project objectives were
ultimately to create a

861
01:13:28.000 --> 01:13:32.934
3-D visualization. For those of
you unfamiliar I have shown a

862
01:13:32.934 --> 01:13:35.934
couple examples on the right. On
the top is a geologic

863
01:13:35.934 --> 01:13:39.934
visualization showing layering
at a site and then below

864
01:13:39.934 --> 01:13:47.934
some of you might be familiar
with the more traditional maps
that can

865
01:13:47.934 --> 01:13:51.934
be created using these 3-D
visualizations.

866
01:13:51.934 --> 01:13:58.000
Our objective was to construct a
3-D BA that shows the
hydrogeology

867
01:13:58.000 --> 01:14:03.000
and if possible the contaminant
distributions. On top

868
01:14:03.000 --> 01:14:07.000
of that, to address the features
including where the mine pool
was

869
01:14:07.000 --> 01:14:11.000
and all the workings.

870
01:14:11.000 --> 01:14:16.000
The stream channels, the
fractures, the faults, and any

871
01:14:16.000 --> 01:14:23.000
other features that might be
important to understanding. The

872
01:14:23.000 --> 01:14:27.000
site team wanted this
visualization so that they could
get a better

873
01:14:27.000 --> 01:14:32.000
understanding to integrate or
interconnect all of the

874
01:14:32.000 --> 01:14:41.000
different aspects into a single
visualization package.

875
01:14:41.000 --> 01:14:42.000
The team was going to use this
to make decisions about where to

876
01:14:42.000 --> 01:14:48.000
go next in their investigation
and what the loading sources

877
01:14:48.000 --> 01:14:52.000
might be. They are also going to
use this

878
01:14:52.000 --> 01:14:56.000
to determine the feasibility of
additional actions might be,

879
01:14:56.000 --> 01:15:03.000
based on the interaction and
availability of water in some of
the subsurface

880
01:15:03.000 --> 01:15:07.000
mine workings. Secondly, and

881
01:15:07.000 --> 01:15:13.000
just as important was to provide
a tool that stakeholders could
use

882
01:15:13.000 --> 01:15:18.000
to visualize the site and get an
understanding of what might be

883
01:15:18.000 --> 01:15:22.000
going on. As mentioned, the city
of Breckenridge was very

884
01:15:22.000 --> 01:15:28.000
much involved with understanding
the

885
01:15:28.000 --> 01:15:29.934
conditions here, they were not
necessarily geologists

886
01:15:29.934 --> 01:15:35.934
and scientists. There has to be
a good platform for everyone to

887
01:15:35.934 --> 01:15:40.934
understand. This slide talks
about the

888
01:15:40.934 --> 01:15:45.934
process used. It comes directly
from the

889
01:15:45.934 --> 01:15:50.934
EPA best practices for site
characterization scores. Let me
walk you through

890
01:15:50.934 --> 01:15:59.000
the process. The first things
necessary are to really
understand

891
01:15:59.000 --> 01:16:01.000
project goals. What questions
are to be answered? That will
help to

892
01:16:01.000 --> 01:16:04.000
drive the datasets we might
need.

893
01:16:04.000 --> 01:16:10.000
The second is to acquire this
data,

894
01:16:10.000 --> 01:16:12.000
to review it, continuously check
it and process the data. This
can

895
01:16:12.000 --> 01:16:18.000
take a tremendous amount of
time, even if the data is
already in databases.

896
01:16:18.000 --> 01:16:22.000
As our first speaker talked
about, dirty data can be

897
01:16:22.000 --> 01:16:28.000
a really important aspect of
reviewing and

898
01:16:28.000 --> 01:16:29.934
managing data.

899
01:16:29.934 --> 01:16:36.934
The third step is to create
databases from this data. From
that, start

900
01:16:36.934 --> 01:16:40.934
developing individual
visualizations of the geology,
hydrogeology

901
01:16:40.934 --> 01:16:46.934
and chemistry. For us we had an
additional component of the

902
01:16:46.934 --> 01:16:49.934
mine features.

903
01:16:49.934 --> 01:16:50.934
Once we develop the individual
component and checked with the
site

904
01:16:50.934 --> 01:16:56.000
team that they were comfortable,
the next step would be to
integrate

905
01:16:56.000 --> 01:17:02.000
and pull these components
together into a single,

906
01:17:02.000 --> 01:17:06.000
viewable structure. One that
would help us understand the
integration

907
01:17:06.000 --> 01:17:13.000
of these components. This will
also help to identify outliers
and check

908
01:17:13.000 --> 01:17:16.000
to see if they might be related
to some other aspect

909
01:17:16.000 --> 01:17:32.000
of the components.

910
01:17:39.934 --> 01:17:42.934
the last up was to present the
findings

911
01:17:42.934 --> 01:17:47.934
to the stakeholder team. We had
a lot of data to look at, but
unfortunately

912
01:17:47.934 --> 01:17:54.000
we had very little digital data,
especially for the mine
features.

913
01:17:54.000 --> 01:17:56.000
Almost all the data that we had
to work with

914
01:17:56.000 --> 01:18:01.000
developed the framework of the
geology and the

915
01:18:01.000 --> 01:18:07.000
mine infrastructure and it was
from a 1934 USGS paper centered
on

916
01:18:07.000 --> 01:18:13.000
the entire record Ridge mining
district. There were a couple

917
01:18:13.000 --> 01:18:19.000
great sections on the Wellington
mine and how they were

918
01:18:19.000 --> 01:18:22.000
developed. Again, this was a
1934 paper that we had

919
01:18:22.000 --> 01:18:26.000
to read and pull from. We really
did not have any digital mining

920
01:18:26.000 --> 01:18:34.934
data. Near the end of the
project

921
01:18:34.934 --> 01:18:35.934
more data did become available
and we were able to incorporate

922
01:18:35.934 --> 01:18:37.934
that to enhance some of our QC
and the topography and
understanding

923
01:18:37.934 --> 01:18:39.934
of some of the mining features.

924
01:18:39.934 --> 01:18:48.934
>> There was hydrology data that
happen generated, starting as
early

925
01:18:48.934 --> 01:18:51.934
as the 1980s from the EPA and
USGS investigations. Most of the
data

926
01:18:51.934 --> 01:18:58.000
was fairly sporadic with a short
burst of intense

927
01:18:58.000 --> 01:19:03.000
activity at the in of the 1990s.
The water level data

928
01:19:03.000 --> 01:19:10.000
was not consistent either and
locations where the data was
taken or

929
01:19:10.000 --> 01:19:12.000
consistent across 1980 through
the current. We did not

930
01:19:12.000 --> 01:19:17.000
have a really good understanding
of the long-term water level.

931
01:19:17.000 --> 01:19:23.000
Similarly, the chemistry data
was a little sporadic and most
of it

932
01:19:23.000 --> 01:19:30.934
was focus on surface water.

933
01:19:30.934 --> 01:19:36.934
>> As I mentioned, most of our
geology and mining
infrastructure

934
01:19:36.934 --> 01:19:42.934
data came from this 1934 paper,
which was an excellent

935
01:19:42.934 --> 01:19:43.934
paper written by some pretty
tremendous geologists. There
were

936
01:19:43.934 --> 01:19:48.934
five plates in the back of that
USGS paper that we used to
generate

937
01:19:48.934 --> 01:19:58.000
most of the mining at the
structure data and a lot of the
fault

938
01:19:58.000 --> 01:19:59.000
data also.

939
01:19:59.000 --> 01:20:02.000
The first plate was what you see
in the top left and it was what

940
01:20:02.000 --> 01:20:06.000
we call the spaghetti map. This
is a single map that shows all
the

941
01:20:06.000 --> 01:20:12.000
layers of both mines on one map.

942
01:20:12.000 --> 01:20:18.000
Fortunately we had another plate
that went level

943
01:20:18.000 --> 01:20:24.000
bilayer -- level by level as

944
01:20:24.000 --> 01:20:29.934
well as the tunnels, the shafts,
and all of the

945
01:20:29.934 --> 01:20:33.934
mine workings. We were able to
digitize each and every one of
these levels

946
01:20:33.934 --> 01:20:37.934
and place them in appropriate
locations,

947
01:20:37.934 --> 01:20:43.934
based on some of the other maps
in

948
01:20:43.934 --> 01:20:45.934
these plates.

949
01:20:45.934 --> 01:20:51.934
>> Another set of data that we
found were these

950
01:20:51.934 --> 01:20:55.000
geologic cross-sections. They
were really helpful and they
provided

951
01:20:55.000 --> 01:20:58.000
a general understanding of the
geology as well as the location
of some

952
01:20:58.000 --> 01:21:04.000
of the falls. There were other
cross-sections that were

953
01:21:04.000 --> 01:21:10.000
provided as part of the mine
level maps. You can see the
level

954
01:21:10.000 --> 01:21:14.000
of detail that was done. We have
some major

955
01:21:14.000 --> 01:21:18.000
faults here that were mapped
here and here

956
01:21:18.000 --> 01:21:24.000
and here, but there is also a
series of much smaller faults in
very detailed

957
01:21:24.000 --> 01:21:28.934
geology that was provided and we
had access to,

958
01:21:28.934 --> 01:21:33.934
but we had to digitize all of
this data.

959
01:21:33.934 --> 01:21:37.934
The first thing we did was we
took the cross-sections and
place them

960
01:21:37.934 --> 01:21:42.934
in the appropriate locations on

961
01:21:42.934 --> 01:21:45.934
the map. What we came up with
was a construct

962
01:21:45.934 --> 01:21:53.000
of geologic cross-sections that
circled around the

963
01:21:53.000 --> 01:21:59.000
Wellington and Oro mines. We
also placed the maps in their
correct

964
01:21:59.000 --> 01:22:03.000
orientation and digitize them to
create the

965
01:22:03.000 --> 01:22:07.000
mine infrastructure. We also
placed the cross-sections that

966
01:22:07.000 --> 01:22:13.000
we saw associated with the
different levels, we

967
01:22:13.000 --> 01:22:17.000
placed them in this 3-D
visualization. You can

968
01:22:17.000 --> 01:22:21.000
see we have three or four in
this figure.

969
01:22:21.000 --> 01:22:28.934
>> Ultimately, what we came up
with is this picture of the mine
infrastructure

970
01:22:28.934 --> 01:22:31.934
that was advertised from all of
those plates. In this figure we

971
01:22:31.934 --> 01:22:36.934
show the mine infrastructure
that is the tunnels

972
01:22:36.934 --> 01:22:41.934
and stokes -- excuse me, just
the tunnels here. We also show
some

973
01:22:41.934 --> 01:22:49.934
of the shafts. These were
important

974
01:22:49.934 --> 01:22:55.000
to understand location and depth
when we started to talk about
water

975
01:22:55.000 --> 01:22:59.000
level, geology, and movement of
contaminants.

976
01:22:59.000 --> 01:23:05.000
>> Faults were a pretty
important component

977
01:23:05.000 --> 01:23:08.000
according to the site team. We
know there were lots of falls at
the

978
01:23:08.000 --> 01:23:10.000
site and they had the potential
for both transmitting
contaminants

979
01:23:10.000 --> 01:23:16.000
as well as generating additional
contaminated nodes

980
01:23:16.000 --> 01:23:18.000
in groundwater.

981
01:23:18.000 --> 01:23:24.000
Our first question was could we
even visualize

982
01:23:24.000 --> 01:23:26.000
these false?

983
01:23:26.000 --> 01:23:31.934
-- Faults? We were not sure if
we could create faults in this

984
01:23:31.934 --> 01:23:33.934
3-D visualization package. We
tried a couple things.

985
01:23:33.934 --> 01:23:35.934
In the picture on the right we
show some of the major faults
that were

986
01:23:35.934 --> 01:23:40.934
located at the site. For you
geologists the bull

987
01:23:40.934 --> 01:23:46.934
hide fault on the left and the
great northern on the

988
01:23:46.934 --> 01:23:53.000
rights described where the mine

989
01:23:53.000 --> 01:23:58.000
was located. One of the most
important faults that we looked
at was this

990
01:23:58.000 --> 01:24:00.000
one in yellow. It went through
some

991
01:24:00.000 --> 01:24:06.000
of the most highly mineralized
zones and also crossed the
French

992
01:24:06.000 --> 01:24:10.000
Gulch approximately where we saw
the highest loading

993
01:24:10.000 --> 01:24:14.000
content. We wanted to make sure
we could visualize that to help

994
01:24:14.000 --> 01:24:23.000
understand the integration of my
working, faults, and

995
01:24:23.000 --> 01:24:27.000
surface water.

996
01:24:27.000 --> 01:24:33.000
>> We went back to the only
dataset that we

997
01:24:33.000 --> 01:24:37.000
really had from those incredible
maps created

998
01:24:37.000 --> 01:24:43.000
in the 1920s and 30s. We took
those maps and identify the
location

999
01:24:43.000 --> 01:24:48.000
of the faults and we created
these planar features

1000
01:24:48.000 --> 01:24:52.000
along those fall planes that
were identified in the
cross-sections.

1001
01:24:52.000 --> 01:25:00.000
We link them together and were
able to place them in 3-D. Here
are a

1002
01:25:00.000 --> 01:25:03.000
couple examples.

1003
01:25:03.000 --> 01:25:08.000
One of the interesting things we
observed was the USGS

1004
01:25:08.000 --> 01:25:13.000
surface map showed a different
location at

1005
01:25:13.000 --> 01:25:19.000
the surface from what we had
interpreted based on

1006
01:25:19.000 --> 01:25:22.000
the subsurface geologic
cross-sections.

1007
01:25:22.000 --> 01:25:27.000
We thought that was an important
conclusion that the team I want

1008
01:25:27.000 --> 01:25:29.934
to understand.

1009
01:25:29.934 --> 01:25:33.934
>> After we presented the
information on the geology

1010
01:25:33.934 --> 01:25:39.934
and faults, the site team
thought that it might be
important to

1011
01:25:39.934 --> 01:25:45.934
understand the mined out areas
or

1012
01:25:45.934 --> 01:25:49.934
the stopes sum could represent
long-term sources

1013
01:25:49.934 --> 01:25:56.000
or additional source material.
All

1014
01:25:56.000 --> 01:26:02.000
we had was a couple Stope

1015
01:26:02.000 --> 01:26:06.000
maps from the 1934 and 1950 USGS
paper. We took these maps and we

1016
01:26:06.000 --> 01:26:13.000
rubber she did them into the 3-D
model. We were able to place

1017
01:26:13.000 --> 01:26:17.000
them in approximate locations.
When we say

1018
01:26:17.000 --> 01:26:24.000
rubber sheet that literally
means stretching parts of the
map so that

1019
01:26:24.000 --> 01:26:26.000
it matched the other
infrastructure that we

1020
01:26:26.000 --> 01:26:30.934
had already developed. We did
this in a few places and we came
up with

1021
01:26:30.934 --> 01:26:35.934
a combination of mine workings
and Stope

1022
01:26:35.934 --> 01:26:38.934
areas.

1023
01:26:38.934 --> 01:26:43.934
>> The next thing we did was
develop the overburdened
geology. It was

1024
01:26:43.934 --> 01:26:48.934
important to cut the dredged
materials could be a potential
source and

1025
01:26:48.934 --> 01:26:55.000
it could be a transport pathway.

1026
01:26:55.000 --> 01:27:00.000
We generated the alluvium extent
and

1027
01:27:00.000 --> 01:27:04.000
the depth based on boring logs
and some

1028
01:27:04.000 --> 01:27:08.000
geologic maps. We also took air
photos and we

1029
01:27:08.000 --> 01:27:14.000
looked very closely for outcrops
so that we could very closely

1030
01:27:14.000 --> 01:27:20.000
bound the materials within the
stream Valley. We also do ties
the

1031
01:27:20.000 --> 01:27:24.000
dredged materials of the air
photo, using a technique that

1032
01:27:24.000 --> 01:27:28.000
I will talk about

1033
01:27:28.000 --> 01:27:31.934
in a little bit, we estimated
the depth. We were able to, with
a three-dimensional

1034
01:27:31.934 --> 01:27:39.934
map. This is a contour map
showing the thickness of the
material within

1035
01:27:39.934 --> 01:27:44.934
the stream. Ultimately what we
came up with was this
three-dimensional

1036
01:27:44.934 --> 01:27:49.934
view of the dredged material,
the and

1037
01:27:49.934 --> 01:27:56.000
the bedrock. You can see from
the upper right that this
material

1038
01:27:56.000 --> 01:28:05.000
is pretty shallow.

1039
01:28:05.000 --> 01:28:10.000
The component was hydrology.

1040
01:28:10.000 --> 01:28:15.000
>> We created a series of
potential metric surfaces from

1041
01:28:15.000 --> 01:28:21.000
the synoptic water level reams
that we had. There was a

1042
01:28:21.000 --> 01:28:26.000
really good set between
1998-2000. but outside of that
it was

1043
01:28:26.000 --> 01:28:29.934
pretty sporadic. We did not
really have a good

1044
01:28:29.934 --> 01:28:32.934
feeling for what was going on
with the surfaces other than
through

1045
01:28:32.934 --> 01:28:38.934
that timeframe. To generate

1046
01:28:38.934 --> 01:28:41.934
an average surface for the mine
we took an

1047
01:28:41.934 --> 01:28:48.934
average of all of the readings
and generated a surface using
all of

1048
01:28:48.934 --> 01:28:58.000
the monitoring wells and we
generated this layer. We also
took

1049
01:28:58.000 --> 01:29:02.000
a look at some of the individual
water levels. For example, there

1050
01:29:02.000 --> 01:29:06.000
was a lot of water for the Oro
shaft over time. One of the
interesting

1051
01:29:06.000 --> 01:29:11.000
things that we noted is that
between 1998-2000 there

1052
01:29:11.000 --> 01:29:19.000
appeared to be about a 6 foot
increase in water

1053
01:29:19.000 --> 01:29:20.000
level. Unfortunately, we did not
have any significant data after

1054
01:29:20.000 --> 01:29:24.000
that that we could ascertain
whether or not that was just a
localized

1055
01:29:24.000 --> 01:29:34.934
event or a longer-term event.

1056
01:29:34.934 --> 01:29:38.934
>> We took that surface map from
the monitoring wells, centered
in

1057
01:29:38.934 --> 01:29:43.934
just a small area and we
extrapolated that across the

1058
01:29:43.934 --> 01:29:48.934
entire area, across the entire
study area. We placed that

1059
01:29:48.934 --> 01:29:53.000
within the 3-D visualization of
mine workings

1060
01:29:53.000 --> 01:29:59.000
and we colored the area below
that surface

1061
01:29:59.000 --> 01:30:03.000
dark and the area above it light
to indicate the areas that would

1062
01:30:03.000 --> 01:30:10.000
likely be flooded versus the
areas that might

1063
01:30:10.000 --> 01:30:15.000
be accessible.

1064
01:30:15.000 --> 01:30:19.000
>> Finally, we took the surface
water data.

1065
01:30:19.000 --> 01:30:24.000
There was quite a bit of surface
water data, we plotted each and

1066
01:30:24.000 --> 01:30:28.000
every event over time. We
plotted the two primary

1067
01:30:28.000 --> 01:30:33.934
contaminants of concern think

1068
01:30:33.934 --> 01:30:37.934
and cadium act created this
visualization so that the team

1069
01:30:37.934 --> 01:30:43.934
could see the cleaner areas, the
cooler colors and

1070
01:30:43.934 --> 01:30:49.934
the darker or brighter colors
where there

1071
01:30:49.934 --> 01:30:53.000
were contaminants.

1072
01:30:53.000 --> 01:30:59.000
We did this for each monitoring
event that

1073
01:30:59.000 --> 01:31:04.000
we had.

1074
01:31:04.000 --> 01:31:10.000
>> I mentioned that most of the
data

1075
01:31:10.000 --> 01:31:12.000
was nondigital.

1076
01:31:12.000 --> 01:31:13.000
A couple interesting pieces were
really valuable to us for
creating

1077
01:31:13.000 --> 01:31:15.000
and quality checking the model.

1078
01:31:15.000 --> 01:31:20.000
First of all, my level elevation
and the extent, there was a
table

1079
01:31:20.000 --> 01:31:26.000
in a 1934 reports that we used
to generate the elevation

1080
01:31:26.000 --> 01:31:31.934
of the entry points. That was

1081
01:31:31.934 --> 01:31:36.934
really helpful in developing the
infastructure. For

1082
01:31:36.934 --> 01:31:42.934
the faults we had some data from
inside the mine levels

1083
01:31:42.934 --> 01:31:51.934
created from the cross-sections,

1084
01:31:51.934 --> 01:31:55.000
but we needed additional data to
find out what the orientation
was

1085
01:31:55.000 --> 01:31:56.000
outside of where we had those
cross-sections.

1086
01:31:56.000 --> 01:32:00.000
When we looked into the report
we could see there

1087
01:32:00.000 --> 01:32:06.000
was some information and
elevation differences and
thicknesses of some

1088
01:32:06.000 --> 01:32:07.434
of these false that we could

1089
01:32:07.434 --> 01:32:09.000
use the quality check our data.

1090
01:32:09.000 --> 01:32:16.000
That was really helpful in
expanding and extrapolating the
location and

1091
01:32:16.000 --> 01:32:21.000
direction of these fall planes
in 3-D. The other really
valuable

1092
01:32:21.000 --> 01:32:27.000
piece of information came from a
1919

1093
01:32:27.000 --> 01:32:30.934
Mining Journal.

1094
01:32:30.934 --> 01:32:34.334
It indicated the depth to
bedrock in the main channel

1095
01:32:34.334 --> 01:32:35.767
was about 45-50 feet.

1096
01:32:35.767 --> 01:32:43.934
That was really valuable,
because we had essentially no
geophysical

1097
01:32:43.934 --> 01:32:49.934
data or geologic boring data
within the area to estimate the

1098
01:32:49.934 --> 01:32:56.000
depth to bedrock. Using 45-50

1099
01:32:56.000 --> 01:33:08.000
feet we were able to generate
that visualization.

1100
01:33:08.000 --> 01:33:09.000
There was a lot of important
data that we cannot include in
the 3-D

1101
01:33:09.000 --> 01:33:11.000
visualization that was really
helpful to the CSM. For example,
the USGS

1102
01:33:11.000 --> 01:33:16.000
had done a number of tests to
look at the groundwater flow

1103
01:33:16.000 --> 01:33:22.000
directions and the different
impacts of

1104
01:33:22.000 --> 01:33:25.000
surface water and run off. We
were not able to

1105
01:33:25.000 --> 01:33:30.934
incorporate that easily into
this particular model, but it
was valuable

1106
01:33:30.934 --> 01:33:33.934
in interpretation. Similarly
there

1107
01:33:33.934 --> 01:33:37.934
was data from deeper mine
shafts, but we couldn't really
find a good

1108
01:33:37.934 --> 01:33:43.934
way to present that data, the
cause there was not enough of
it.

1109
01:33:43.934 --> 01:33:46.934
It was limited spatially and
temporally.

1110
01:33:46.934 --> 01:33:50.934
We did not have a good story to
tell.

1111
01:33:50.934 --> 01:33:56.000
>> As you saw, there was a
tremendous and complex
mineralogy

1112
01:33:56.000 --> 01:34:04.000
associated with a number of
different zones and the
different

1113
01:34:04.000 --> 01:34:06.000
geologic formations.

1114
01:34:06.000 --> 01:34:10.000
We were not able to reasonably
correlate all of these different

1115
01:34:10.000 --> 01:34:17.000
layers like you would normally
do and an section if you are
doing

1116
01:34:17.000 --> 01:34:23.000
a traditional transport. The
level of detail of

1117
01:34:23.000 --> 01:34:26.000
the geology and the changes in
geology was really

1118
01:34:26.000 --> 01:34:30.000
not necessary. The team did not
feel like it was a significant
contributing

1119
01:34:30.000 --> 01:34:33.000
factor to something they wanted
to

1120
01:34:33.000 --> 01:34:37.000
look at. There were
interpretations that

1121
01:34:37.000 --> 01:34:40.000
were used to support our
visualization.

1122
01:34:40.000 --> 01:34:46.000
For example, knowing that the
mine and the were

1123
01:34:46.000 --> 01:34:52.000
hydraulically connected were
really valuable in allowing

1124
01:34:52.000 --> 01:34:58.000
us to extrapolate across the
entire

1125
01:34:58.000 --> 01:35:01.000
study area. Similarly, knowing
there was drawdown across

1126
01:35:01.000 --> 01:35:06.000
the 1110 fault told us that was
not abounding fall, but that it

1127
01:35:06.000 --> 01:35:13.000
was transmissive and we could
create a cross-section or a
potential metric

1128
01:35:13.000 --> 01:35:19.000
surface that went across that.

1129
01:35:19.000 --> 01:35:25.000
The final product that we have
for the

1130
01:35:25.000 --> 01:35:29.934
team was a visualization package
that

1131
01:35:29.934 --> 01:35:33.934
included the four components as
well as an integrated
visualization

1132
01:35:33.934 --> 01:35:36.934
and a memo that explained how we
came

1133
01:35:36.934 --> 01:35:42.934
up with this and what was in the
3-D

1134
01:35:42.934 --> 01:35:43.934
the eighth. A couple other
useful products that came

1135
01:35:43.934 --> 01:35:49.934
out were presentation to the
stakeholders using this 3-D VA
to help them better

1136
01:35:49.934 --> 01:35:54.000
understand the site

1137
01:35:54.000 --> 01:35:56.000
and a database. Since we spend
so much time cleaning up the
monitoring

1138
01:35:56.000 --> 01:36:00.000
well and chemistry database this
was a great set of

1139
01:36:00.000 --> 01:36:03.000
point -- starting point.

1140
01:36:03.000 --> 01:36:07.000
>> Finally, we did create some

1141
01:36:07.000 --> 01:36:09.000
PDFs, they are actually 3-D PDFs
similar to the product I'm going

1142
01:36:09.000 --> 01:36:15.000
to show it a minute that could
be used in meetings and

1143
01:36:15.000 --> 01:36:18.000
other medications.

1144
01:36:18.000 --> 01:36:27.000
I am going to switch over and do
a quick walk-through of the

1145
01:36:27.000 --> 01:36:40.934
final product.

1146
01:36:40.934 --> 01:36:44.934
>> I will just remind the
audience that you each have
individual controls

1147
01:36:44.934 --> 01:36:51.934
in the upper right window.

1148
01:36:51.934 --> 01:36:56.000
>> I am just going to walk
quickly through this

1149
01:36:56.000 --> 01:36:58.000
final integrated VA that we had.

1150
01:36:58.000 --> 01:37:06.000
We draped the air photo over the
lidar imagery. We will step
through

1151
01:37:06.000 --> 01:37:12.000
the different stages that we
provided the

1152
01:37:12.000 --> 01:37:17.000
client with. This is a depiction
of the. If you

1153
01:37:17.000 --> 01:37:24.000
look at it from the side you can
see

1154
01:37:24.000 --> 01:37:27.000
the thickness.

1155
01:37:27.000 --> 01:37:30.934
And then you can see

1156
01:37:30.934 --> 01:37:34.934
the faults. We laid the surface
location and we also showed the

1157
01:37:34.934 --> 01:37:39.934
3-D planar surfaces in the
subsurface. We added

1158
01:37:39.934 --> 01:37:43.934
each fall individually so that
the teams

1159
01:37:43.934 --> 01:37:53.000
could flu -- view the
significance and take them on
and off at

1160
01:37:53.000 --> 01:37:59.000
their leisure. Then we added the
infastructure,

1161
01:37:59.000 --> 01:38:04.000
including the Stope areas. If
you turn it sideways you can
manipulate

1162
01:38:04.000 --> 01:38:08.000
this. You can expand, you can
see how the

1163
01:38:08.000 --> 01:38:17.000
infastructure is related to the
fault and the.

1164
01:38:17.000 --> 01:38:23.000
On top of that we placed the
monitoring wells that

1165
01:38:23.000 --> 01:38:28.934
were at the site, they are
indicated

1166
01:38:28.934 --> 01:38:31.934
in yellow.

1167
01:38:31.934 --> 01:38:40.934
We show the total depth of the
well as well as any

1168
01:38:40.934 --> 01:38:42.934
screen intervals.

1169
01:38:42.934 --> 01:38:46.934
Here is another shot showing the
dressed materials, looking

1170
01:38:46.934 --> 01:38:50.934
at it from the edge you can look
at the 3-D version and see how

1171
01:38:50.934 --> 01:38:57.000
that material relates to the
stream and

1172
01:38:57.000 --> 01:39:06.000
the faults.

1173
01:39:06.000 --> 01:39:07.000
Then we place the chemistry data
for the streams and the
monitoring

1174
01:39:07.000 --> 01:39:13.000
wells, identify each individual
point where chemistry data

1175
01:39:13.000 --> 01:39:17.000
was taken.

1176
01:39:17.000 --> 01:39:22.000
This is for zinc and then we
added the Cadium data.

1177
01:39:22.000 --> 01:39:28.000
>> Ultimately this was the final
integrated VA that the team

1178
01:39:28.000 --> 01:39:36.934
could use.

1179
01:39:36.934 --> 01:39:42.934
All right, that concludes my
presentation.

1180
01:39:42.934 --> 01:39:48.934
I will be happy to entertain

1181
01:39:48.934 --> 01:39:49.934
any questions.

1182
01:39:49.934 --> 01:39:54.000
>> I remind everyone the Q&A
window is available in the lower
right

1183
01:39:54.000 --> 01:39:57.000
corner of your screen. We did
get a few of them they came in
while

1184
01:39:57.000 --> 01:40:01.000
you were presenting. I will just
read one or two out while you
are

1185
01:40:01.000 --> 01:40:04.000
getting rear-ended in the
environment.

1186
01:40:04.000 --> 01:40:09.000
A number of people have asked
about the software that

1187
01:40:09.000 --> 01:40:12.000
you used for the visualization.
If you could

1188
01:40:12.000 --> 01:40:18.000
confirm or repeat the name and
if it

1189
01:40:18.000 --> 01:40:21.000
is proprietary.

1190
01:40:21.000 --> 01:40:28.934
>> It is a commercial software
called EVS. It is produced by a

1191
01:40:28.934 --> 01:40:37.934
company called C-Tech studios.
It is

1192
01:40:37.934 --> 01:40:42.934
extremely sophisticated and very
statistically, you really need

1193
01:40:42.934 --> 01:40:48.934
an expert to run.

1194
01:40:48.934 --> 01:40:54.000
There are a lot of add-ons that
you can do and a lot of

1195
01:40:54.000 --> 01:40:57.000
proprietary modules that people
generate to

1196
01:40:57.000 --> 01:41:01.000
help facilitate developing these
different models. It is

1197
01:41:01.000 --> 01:41:07.000
a commercially available
software with licensing. It is

1198
01:41:07.000 --> 01:41:10.000
not cheap. It is a very
sophisticated product.

1199
01:41:10.000 --> 01:41:15.000
>> Could you talk a little bit
about the amount of time it took
to complete

1200
01:41:15.000 --> 01:41:19.000
this project? And number of
people commented on the

1201
01:41:19.000 --> 01:41:25.000
digitizing element and that it
could've taken them years.

1202
01:41:25.000 --> 01:41:29.934
Any interested -- information on
the

1203
01:41:29.934 --> 01:41:33.934
time required.

1204
01:41:33.934 --> 01:41:36.934
>> It was pretty intensive,
although it was

1205
01:41:36.934 --> 01:41:39.934
not the digitizing itself. Once
you have a map laid out on the
table

1206
01:41:39.934 --> 01:41:45.934
you can just zip right through
them pretty quickly. It did take

1207
01:41:45.934 --> 01:41:51.934
some time. What really took the
time was making sure that all

1208
01:41:51.934 --> 01:41:58.000
of the layers lined up with each
other in the

1209
01:41:58.000 --> 01:42:02.000
3-D space. That did take many
hours. Also, data cleanup for

1210
01:42:02.000 --> 01:42:05.000
this particular site took a long
time.

1211
01:42:05.000 --> 01:42:11.000
That is going through the
databases, extracting data

1212
01:42:11.000 --> 01:42:14.000
from reports.

1213
01:42:14.000 --> 01:42:17.000
It took 18 months or so to do
the entire process, we had a
number

1214
01:42:17.000 --> 01:42:24.000
of times where we have to stop
for various things for months

1215
01:42:24.000 --> 01:42:27.000
on in. The key thing is we
always had a product and a
component available

1216
01:42:27.000 --> 01:42:31.934
for the site team to look at and
respond to so that we could do
our

1217
01:42:31.934 --> 01:42:39.934
next phase of work.

1218
01:42:39.934 --> 01:42:43.934
>> Could you go back to one of
the illustrations, they are
interested

1219
01:42:43.934 --> 01:42:48.934
in what the axes was on the XYZ?
They

1220
01:42:48.934 --> 01:42:51.934
are specifically looking at some
of the numbers or values or they

1221
01:42:51.934 --> 01:42:56.000
were not sure if that was even
relevant in the viewer that you
were

1222
01:42:56.000 --> 01:42:59.000
showing.

1223
01:42:59.000 --> 01:43:05.000
>> It is quite relevant. The X
and

1224
01:43:05.000 --> 01:43:13.000
Y were coordinates.

1225
01:43:13.000 --> 01:43:14.000
That was the basis of our
developing these locations and
getting everything

1226
01:43:14.000 --> 01:43:19.000
properly GR reference. The Z is
actually the elevations, the
true

1227
01:43:19.000 --> 01:43:27.000
elevation. Some from the table
that indicated what the entry
point

1228
01:43:27.000 --> 01:43:30.934
elevation was for each of the
levels. That was a great
question.

1229
01:43:30.934 --> 01:43:35.934
>> Okay. We had a few other
questions that came

1230
01:43:35.934 --> 01:43:39.934
in about the actual application
and decision-making. They were
wondering

1231
01:43:39.934 --> 01:43:45.934
if you could give more detail on
how your customer or whoever you

1232
01:43:45.934 --> 01:43:49.934
built a visualization for was
using it to make decisions

1233
01:43:49.934 --> 01:43:53.000
when it came to
characterization.

1234
01:43:53.000 --> 01:43:58.000
>> I can't really speak a lot to
that. They are still in the
process

1235
01:43:58.000 --> 01:44:02.000
of determining what to do next,
but it did give them some

1236
01:44:02.000 --> 01:44:08.000
insight as to what areas they
might want to focus on.

1237
01:44:08.000 --> 01:44:15.000
For example, where the faults
cross the screen

1238
01:44:15.000 --> 01:44:17.000
those could be loading areas for
contaminants. Those might be
areas

1239
01:44:17.000 --> 01:44:20.000
they want to look at in more
detail.

1240
01:44:20.000 --> 01:44:29.000
In terms of optimizing and
improving the collection and

1241
01:44:29.000 --> 01:44:30.000
treatments of Meadowlake waters
there may be additional areas
that

1242
01:44:30.000 --> 01:44:34.000
they want to look at. For
example, the fact that the 1110

1243
01:44:34.000 --> 01:44:40.000
fault as maps was different from
what

1244
01:44:40.000 --> 01:44:43.000
we interpreted might give an
idea of where there might be a
secondary

1245
01:44:43.000 --> 01:44:48.000
loading area into thethat had
not already been identified.

1246
01:44:48.000 --> 01:44:55.000
The site team could then do some
investigative approaches to see

1247
01:44:55.000 --> 01:44:59.000
if there were additional loading
points.

1248
01:44:59.000 --> 01:45:05.000
>> I do not see any other
questions in

1249
01:45:05.000 --> 01:45:11.000
the queue. One last one

1250
01:45:11.000 --> 01:45:12.000
came in in terms of data
sources.

1251
01:45:12.000 --> 01:45:17.000
Can you remind the audience were
you pulled some of the

1252
01:45:17.000 --> 01:45:20.000
information from?

1253
01:45:20.000 --> 01:45:26.000
>> For the my workings it was
literally some plates, paper

1254
01:45:26.000 --> 01:45:29.934
copies of large maps that we

1255
01:45:29.934 --> 01:45:35.934
digitized. For the groundwater
chemistry water level data these
were

1256
01:45:35.934 --> 01:45:43.934
Excel spreadsheets and databases
and tables from

1257
01:45:43.934 --> 01:45:47.934
various reports. It was quite a
range of different types of data

1258
01:45:47.934 --> 01:45:50.934
that have to be integrated
together.

1259
01:45:50.934 --> 01:45:55.000
>> I think in the interest of
time

1260
01:45:55.000 --> 01:45:59.000
we might wrap up your specific
question Q and I will sort of
switch back

1261
01:45:59.000 --> 01:46:03.000
to you and that you

1262
01:46:03.000 --> 01:46:04.000
finish out with facilitation
duties.

1263
01:46:04.000 --> 01:46:11.000
>> Thanks, you are keeping me
busy today. [ laughter ]

1264
01:46:11.000 --> 01:46:17.000
>> We did have one question that
came in regarding Katie and

1265
01:46:17.000 --> 01:46:20.000
Doug's presentation.

1266
01:46:20.000 --> 01:46:26.000
In terms of the outlier analysis
talks a lot about dirty data.
Correcting

1267
01:46:26.000 --> 01:46:31.934
dirty data can be a

1268
01:46:31.934 --> 01:46:35.934
real challenge. You started an
explanation of looking at your

1269
01:46:35.934 --> 01:46:39.934
outlier analysis. Was that done
manually or did you have an

1270
01:46:39.934 --> 01:46:45.934
algorithm that help you to
identify those

1271
01:46:45.934 --> 01:46:51.934
data points?

1272
01:46:51.934 --> 01:46:53.000
>> Just to

1273
01:46:53.000 --> 01:46:56.234
clarify, this was related to
Doug and I?

1274
01:46:56.234 --> 01:47:02.000
>> Correct.

1275
01:47:02.000 --> 01:47:07.000
>> We didn't have an hour with
them we identify them visually.

1276
01:47:07.000 --> 01:47:13.000
>> To go a little deeper, it was
started

1277
01:47:13.000 --> 01:47:21.000
by just doing frequency. We
first looked

1278
01:47:21.000 --> 01:47:28.934
at the frequency of occurrence
and solve rain, snow, sleet, and
precipitation

1279
01:47:28.934 --> 01:47:32.934
were in the top 20.

1280
01:47:32.934 --> 01:47:37.934
We grouped those together and
did a more of those groupings
until

1281
01:47:37.934 --> 01:47:42.934
the top 10 or so

1282
01:47:42.934 --> 01:47:48.934
causes comprised 85-90+ percent
of

1283
01:47:48.934 --> 01:47:58.000
the data. In most cases it was
valid and

1284
01:47:58.000 --> 01:48:01.000
not standard.

1285
01:48:01.000 --> 01:48:06.000
If it had rain-snow mix that
would've been valid, but there
was no good

1286
01:48:06.000 --> 01:48:21.000
way and no tool to match that
in.

1287
01:48:21.000 --> 01:48:24.000
>> Thanks. You know, we did pose
a question to the

1288
01:48:24.000 --> 01:48:28.934
audience about participants who
might have used this tool and we

1289
01:48:28.934 --> 01:48:34.934
didn't get any responses from
anyone who confirmed they have

1290
01:48:34.934 --> 01:48:37.934
used it, but it was interesting,
we got a couple

1291
01:48:37.934 --> 01:48:42.934
of responses from folks who said
they could imagine using the
tool

1292
01:48:42.934 --> 01:48:50.934
either for facilities, to look
at ongoing issues or even to

1293
01:48:50.934 --> 01:48:57.000
look at the differences in
evaluating natural versus mine

1294
01:48:57.000 --> 01:49:01.000
influence metals or conditions
over time. Does that sound like
something

1295
01:49:01.000 --> 01:49:08.000
this tool might be suitable for?

1296
01:49:08.000 --> 01:49:13.000
>> Definitely, the place where I
could see that the

1297
01:49:13.000 --> 01:49:18.000
most applicable would be in one
watershed where you had many
sources

1298
01:49:18.000 --> 01:49:23.000
that were changing over time and
then you might be collecting

1299
01:49:23.000 --> 01:49:29.934
data at different time intervals
and somewhat

1300
01:49:29.934 --> 01:49:30.934
different contaminant analyses.

1301
01:49:30.934 --> 01:49:36.934
The ability to pull these data
sources together and look at

1302
01:49:36.934 --> 01:49:42.934
a broader watershed wide or
statewide change in composition
over time

1303
01:49:42.934 --> 01:49:48.934
definitely would be.

1304
01:49:48.934 --> 01:49:54.000
>> Okay. One of the other areas
tools like this have

1305
01:49:54.000 --> 01:49:59.000
been used in the natural
environment is there is a case
study

1306
01:49:59.000 --> 01:50:05.000
about three years ago in
Indonesia, which floods all

1307
01:50:05.000 --> 01:50:10.000
the time, using twitter to
estimate flooding depth

1308
01:50:10.000 --> 01:50:18.000
all over the city. At first they
just took

1309
01:50:18.000 --> 01:50:23.000
natural twitter and made it a
civic duty. To take all of that
and pull

1310
01:50:23.000 --> 01:50:28.000
it all together to map out in a
location where they just did not

1311
01:50:28.000 --> 01:50:31.934
have anything in the way of
water height sensors that

1312
01:50:31.934 --> 01:50:37.934
they could get.

1313
01:50:37.934 --> 01:50:44.934
I love the thought of it being
any way to garner

1314
01:50:44.934 --> 01:50:47.934
specific information?

1315
01:50:47.934 --> 01:50:55.000
>> It sounds like this data
analytic tool could

1316
01:50:55.000 --> 01:50:59.000
be used real-time in evaluating
situations and

1317
01:50:59.000 --> 01:51:01.000
mining sites?

1318
01:51:01.000 --> 01:51:04.000
>> If you have the data set,
definitely.

1319
01:51:04.000 --> 01:51:11.000
The whole basis of this data
expiration

1320
01:51:11.000 --> 01:51:13.000
and the value of having a very
wide dataset with different
parameters,

1321
01:51:13.000 --> 01:51:18.000
you can see what parameters
tended to

1322
01:51:18.000 --> 01:51:25.000
be influential. It could be real
time

1323
01:51:25.000 --> 01:51:29.934
evaluation informing on-site
analysis or other data

1324
01:51:29.934 --> 01:51:30.934
collection methods.

1325
01:51:30.934 --> 01:51:36.934
>> One of the magic parts of
these tools are the tool

1326
01:51:36.934 --> 01:51:42.934
itself, every data element is
related. When

1327
01:51:42.934 --> 01:51:48.934
you want to slice and dice
differently or you want

1328
01:51:48.934 --> 01:51:53.000
quantitative information in the
bucket of 0-100. 100-1000.

1329
01:51:53.000 --> 01:51:58.000
1000-10.000. there is just no
back end work

1330
01:51:58.000 --> 01:52:04.000
to do. There are things that can
be done on the fly

1331
01:52:04.000 --> 01:52:08.000
very easily.

1332
01:52:08.000 --> 01:52:11.000
>> Okay.

1333
01:52:11.000 --> 01:52:17.000
Any other questions? I don't see
anything else. We did

1334
01:52:17.000 --> 01:52:24.000
get a comment, it looks like
region 10 is developing a tool
to

1335
01:52:24.000 --> 01:52:28.934
show where or D research
development products and
projects are located

1336
01:52:28.934 --> 01:52:34.934
and developed. Not only can you
use this

1337
01:52:34.934 --> 01:52:38.934
for site-specific, it appears
you can use it for programmatic

1338
01:52:38.934 --> 01:52:44.934
tools also.

1339
01:52:44.934 --> 01:52:51.934
>> Definitely.

1340
01:52:51.934 --> 01:52:57.000
>> I do not see additional
questions coming in at

1341
01:52:57.000 --> 01:53:00.000
this point. Jean, I think we can
wrap up.

1342
01:53:00.000 --> 01:53:04.000
>> Thank you very much. I think
that we will go ahead and

1343
01:53:04.000 --> 01:53:08.000
walk through our final reminders
for today's session. This was
session

1344
01:53:08.000 --> 01:53:15.000
number four in the mining
seminar series that was

1345
01:53:15.000 --> 01:53:18.000
hosted by the office of science
policy. This is the last session

1346
01:53:18.000 --> 01:53:25.000
in the series. We want to thank
everyone for joining us in

1347
01:53:25.000 --> 01:53:28.000
these sessions. I encourage you
to stay connected with us by
visiting

1348
01:53:28.000 --> 01:53:31.934
us at our website or signing up
for our free monthly newsletter

1349
01:53:31.934 --> 01:53:36.934
if you are interested in
learning about related webinars
on this topic

1350
01:53:36.934 --> 01:53:38.934
or additional resources when it
comes

1351
01:53:38.934 --> 01:53:47.934
to character revising and
cleaning up contaminated sites.
I will

1352
01:53:47.934 --> 01:53:50.934
remind everyone that copies of
the presentation

1353
01:53:50.934 --> 01:53:55.000
materials are currently posted
on the links or resources page.
You

1354
01:53:55.000 --> 01:54:10.000
can follow the links under
related URLs.

1355
01:54:10.000 --> 01:54:12.000
Click on the seminar resources
link and press the browse button

1356
01:54:12.000 --> 01:54:13.000
to open it up. I have also
included those in the Q&A
window. Contact

1357
01:54:13.000 --> 01:54:14.000
information for our presenters
and organizers can also be found
there.

1358
01:54:14.000 --> 01:54:17.000
I will ask each of you to fill
out our brief feedback form to
let us

1359
01:54:17.000 --> 01:54:18.000
know what you thought of today
session.

1360
01:54:18.000 --> 01:54:19.000
One of the most common questions
that I get about our sessions is

1361
01:54:19.000 --> 01:54:20.000
if we offer credits or
certificates?

1362
01:54:20.000 --> 01:54:26.000
We do not issue city use, but we
do provide for dissipation
certificates

1363
01:54:26.000 --> 01:54:30.000
when you submit the back. Please
be sure to enter

1364
01:54:30.000 --> 01:54:34.000
your email address correctly and
check

1365
01:54:34.000 --> 01:54:41.000
the box at the very bottom.

1366
01:54:41.000 --> 01:54:43.000
>> [ Event has exceeded
scheduled

1367
01:54:43.000 --> 01:54:46.000
time. Captioner must proceed to
captioner's next scheduled
event.

1368
01:54:46.000 --> 01:54:51.000
Disconnecting at 300. ]