Phragmites Adaptive Management Framework: Model Overview

Video Transcript
Download Video
Right-click and save to download

Detailed Description

This animation is the third video in a series explaining different aspects of the Phragmites Adaptive Management Framework (PAMF). This video focuses on the machine learning model that PAMF uses to systematically learn from participant's data and then provide site-specific management guidance in return. More information about PAMF can be found at our websites: http://www.greatlakesphragmites.net/pamf/.

Also see: https://www.usgs.gov/media/videos/pamf-participation-cycle and https://www.usgs.gov/media/videos/phragmites-adaptive-management-framework-pamf

Details

Image Dimensions: 1561 x 881

Date Taken:

Length: 00:03:19

Location Taken: Ann Arbor, MI, US

Video Credits

 

Great Lakes Commission;
Kindea Labs

Transcript

{9.20 - 15.72}
Non-native Phragmites
australis, or common reed, is
invasive in the Great Lakes
basin

{15.72 - 18.04}
and controlling it can be
expensive.

{18.04 - 22.84}
Because it often isn't clear
what management approaches
will succeed, many managers
are

{22.84 - 26.60}
interested in science-based
decision support.

{26.60 - 32.64}
The Phragmites Adaptive
Management Framework, or PAMF,
invites Phragmites managers to
contribute

{32.64 - 35.40}
to and benefit from collective
learning.

{35.40 - 39.68}
At the core of the program is
a mathematical model that
weighs the same factors that
you

{39.68 - 44.88}
might consider when managing
Phragmites-- tradeoffs between
efficacy, uncertainty and

{44.88 - 45.92}
cost.

{45.92 - 50.96}
It draws on experience and
data from participants across
the basin to determine the
best-value

{50.96 - 54.00}
guidance for each management
unit.

{54.00 - 58.92}
Each complete set of reports
that you upload to the PAMF
Web Hub teaches the model
about

{58.92 - 61.40}
Phragmites' response to
management.

{61.40 - 66.44}
Monitoring reports track the
severity of each Phragmites
invasion through time.

{66.44 - 70.12}
Management reports tell the
model what was done, and how
much it cost.

{70.12 - 75.20}
Together, these reports show
the model how Phragmites in
your management unit responded

{75.20 - 77.28}
to the actions you carried out.

{77.28 - 82.32}
PAMF's model keeps track of
what is known about management
outcomes by categorizing

{82.32 - 88.64}
invasions into six states,
increasing in severity from
one to six, and predicting how
likely

{88.64 - 93.96}
an invasion is to transition
from one state to another
after management.

{93.96 - 99.20}
Imagine that you have a highly
invaded management unit in
State 6, and plan to control
it using

{99.20 - 104.40}
Management Combination A.
Before you take action, the
model predicts that your unit

{104.40 - 111.84}
is 40% likely to improve to
State 4, 20% likely to improve
to State 3, and 10% likely to
end

{111.84 - 113.76}
up in another state.

{113.76 - 118.32}
After managing, you see the
unit's infestation improve to
State 3.

{118.32 - 123.56}
The model responds by
increasing its prediction that
similar units managed the same
way will

{123.56 - 129.20}
improve to State 3, and
reducing predictions that they
will transition to any other
state.

{129.20 - 131.04}
As you and others file
reports, the model's
predictions.

{131.04 - 134.28}
As you and others file
reports, you add to the pool
of data that informs the
model's

{134.28 - 135.32}
predictions.

{135.32 - 1387.60}
Now suppose you have another
combination available.

{138.60 - 143.68}
The model predicts that
Combination B is 40% likely to
improve your unit's
infestation

{143.68 - 150.92}
to State 2, 20% likely to
improve it to State 3, and 10%
likely to result in another
state.

{150.92 - 156.40}
Since Combination B is more
likely to improve your unit to
a less invaded state, the
model

{156.40 - 161.88}
will recommend Combination B
over Combination A, provided
that it isn't too expensive.

{161.88 - 166.92}
As the model collects data
about cost and effectiveness
of each of PAMF's sixteen

{166.92 - 172.12}
combinations, its guidance
changes to reflect the best
information available.

{172.12 - 173.56}
You make this happen!

{173.56 - 178.44}
The model is a critical part
of the adaptive management
process and reduces the
uncertainty

{178.44 - 182.28}
surrounding Phragmites
management throughout the
Great Lakes basin.

{182.28 - 187.80}
Your data drive the collective
learning process and help
reduce the spread of invasive
Phragmites.