Monitoring Land CoverUse Changes at National Parks

1
Monitoring Land Cover/Use Changes at National
Parks

• Eric Brown de Colstoun
• Science Systems and Applications, Inc.
• Biospheric Sciences Branch
• NASA/Goddard Space Flight Center

NASA Biodiversity and Eco-forecasting Science
Team Meeting Washington, D.C. August 29-31, 2005
2
National Park Service Mission
Promote and regulate the use of ...national
parks to conserve the scenery and the natural
and historic objects and wild life therein and to
provide for the enjoyment of the same in such
manner and by such means as will leave them
unimpaired for the enjoyment of future
generations.
NPS Organic Act, 1916
3
NATIONAL PARKS OMNIBUS MANAGEMENT ACT OF 1998
The Secretary shall undertake a program of
inventory and monitoring of National Park System
resources to establish baseline information and
to provide information on the long-term trends in
the condition of National Park System resources.
The monitoring program shall be developed in
cooperation with other Federal monitoring and
information collection efforts to ensure a
cost-effective approach.
The Secretary shall assure the full and proper
utilization of the results of scientific studies
for park management decisions.
4
Science-based Inventories Monitoring Programs
5
Most Common Vital Signs (first 17 networks)
6
Current Activities/Contacts

• NASA/NPS MOA signed.
• NASA/NPS/CSA/Parks Canada meeting in St.
  Petersburg, Fl. March 1-3 2005
• Creation of North American Network for Remote
  Sensing Park Ecosystem Condition (NARSEC).
• Woody Tuner (NASA HQ), John Gross (NPS)
  organizers.
• NPS Vegetation Mapping Program
• http//biology.usgs.gov/npsveg/
• NPS IM Program
• http//science.nature.nps.gov/im/monitor/

7

• National Park Vegetation Mapping and Monitoring
  (J. Irons, Code 614.4)
• The National Park Service is attempting to map
  and monitor vegetation resources within over 270
  park units using aerial photography and manual
  photo interpretation.
• At the current rate progress, 25 to 50 years may
  be needed to map all 270 units at the required
  detail and accuracy. The information is required
  more immediately for many within-park resource
  management decisions. Active monitoring with
  these methods is not realistic.
• NASA/GSFC is collaborating with the NPS to
  develop tools to use Landsat data for
  cost-effective and timely mapping and monitoring
  of park resources.
• Potential Benefits
• Cost-effective, accurate and repeatable
  assessment of National Park vegetation resources
  and surrounding areas (connections with local
  and/or state organizations).
• NPS able to use NASA ESE data on self-sufficient
  basis.
• Tools and products impact resource management
  decisions within the parks (e.g. Invasive
  species, urban growth pressures, water
  management).
• Potential National Application (use national
  standards).

8
National Park Vegetation Mapping and Monitoring

• Phase 1
• Vegetation mapping tools using multi-date Landsat
  7 data demonstrated at Delaware Water Gap
  National Recreation Area (PA, NJ).
• Use National Vegetation Classification Standard,
  and NPS mapping and accuracy assessment
  protocols.
• Use state-of-the-art decision tree classifier.
• Brown de Colstoun , E.C., M.H. Story, C.
  Thompson, K. Commisso, T.G. Smith, and J.R. Irons
  (2003). National Park Vegetation Mapping Using
  Multi-temporal Landsat 7 Data and a Decision Tree
  Classifier, Remote Sens. Environ., 85316-327,
  2003.

• Phase 2
• Use Landsat for active monitoring of natural
  resources at Delaware Water Gap NRA, Upper
  Delaware Scenic and Recreational River and
  vicinity
• -Invasive Species (Woolly Adelgid).
• -Urban growth pressures on the park.
• -Input to urban growth and watershed models.

9
Consequences of Land Cover/Use Changes on
National ParksA Research/Educational
Partnership in the Upper Delaware River Basin

• Proposal submitted to NASAs New Investigator
  Program
• P.I. Eric Brown de Colstoun
• Research/Education support Jessica Robin (SSAI,
  UMD, GLOBE)
• Partners include Elissa Levine (NASA/GSFC), Susan
  Riha (Cornell U.), River Valley GIS consortium
  (UPDE, DEWA), GLOBE program, educators/students
  from area schools.
• 3-year project to develop tools for land
  cover/use change monitoring.
• Model consequences of changes on water/energy
  cycles.
• Structured around existing educational
  connections between NASA, NPS, area schools and
  the GLOBE program.
• Develop pilot curriculum that includes Earth
  system science, remote sensing, modeling, etc...
• Selection April 1, 2004, Project began February
  2005.

10
Research Objectives

• To develop cost-effective, satellite-based
  methods to inventory and monitor land cover/use
  in and around National Parks in support of the
  NPS Inventory and Monitoring Program.
• Measure land cover/use changes and trends in the
  Upper Delaware River Basin from 1984 to the
  present using Landsat.
• Simulate urban growth to 2030 with various growth
  scenarios using SLEUTH urban growth model.
• Examine consequences of land cover/use changes on
  regional water and energy cycles with the GAPS
  model.

11
Research Overview
Landsat Data 1984-2004
Atmos. Correction MODAPS
CUBIST Regression Tree
Training Data
Tree, Grass, Bare 1984-2004
Validation
Land Cover/Use
Land Cover/Use Change 1984-2004
12
Research Overview (cont.)
Land Cover/Use Change 1984-2004
GAPS Model
SLEUTH Model
Consequences of Land cover/use change on energy
and water cycles
Urban growth in the Upper Delaware River Basin
to 2030
13
(No Transcript)
14
Mosaic of two Landsat ETM scenes Acquired on
September 23, 1999 3,2,1
15
LEDAPS Atmospheric Corrections
16
Atmos. Corrections (Cont.)
17
Tree Cover Training Data Development
Color Infrared Air Photo (0.6 m Resolution)
18
Results from Cubist Regression Tree
19
Tree Cover Time Series
20
Summary

• The NPS IM Program offers an excellent
  opportunity for meaningful scientific
  collaborations.
• Data/tool availability for research is
  unprecedented
• Landsat ??
• Inter-Agency funding??
• Tools for decision support are not fully
  developed yet.
• Preliminary analyses show promise for continuous
  fields approaches at Landsat scale but much work
  remains.
• Second phase of project will integrate land cover
  change and student measurements into regional
  modeling efforts.
• Support your Parks!!

21
Regression tree approach
Establish linear models for estimating a bands
values on the target scene. - Values for a
target scene band are the dependent
variable - Values from all bands from a prior
and/or a post date image are used as
independent variables.
22
Divide and Conquer to handle non-linearity and
high order interactions Cubist example
Rule 1/1 10230 cases, mean 85.3, range 60 to
122, est err 4.1 if band01 lt 99 band04 gt
52 band04 lt 77 then dep 10.2 0.66
band01 0.42 band02 - 0.12 band03 - 0.05
band05 Rule 1/2 610 cases, mean 85.7, range
65 to 134, est err 5.6 if band01 lt
103 band04 lt 52 then dep 4.4 0.65
band01 0.48 band02 0.18 band06 - 0.15
band03 - 0.21 band04
Death and Fabricius 2000 http//www.ruleques
t.com/cubist-info.html
23
The Approach used by Cubists
1) Develop classical regression tree -all nodes
mutually exclusive -subdivide data into subsets
which minimize the simple linear model
weighted standard deviation of residuals 2)
Develop Generalized rules from the regression
tree in step 1 -makes trees easier to interpret
-less rules than tree leaves (a.k.a.
terminal nodes) -generalized by deleting
excessive conditions -rule sets can overlap,
predictions are averaged in overlap areas for
smoother output -usually as accurate as a pruned
regression tree 3) Generalized rules are used to
make predictions across the image
Quinlan 1993
24
Cubist linear models

• Construct classical multivariate linear model for
  each generalized rule
• Simplification of linear model
• eliminate parameters to minimize estimated
  error
• uses a greedy search to remove variables that
  contribute little

Quinlan, J.R. 1992. Learning with continuous
classes. In proceedings AI92, Singapore World
Scientific.