Transcript Using Remote Sensing to Better Managing Wildlife

Monitoring Landscape Dynamics
John Gross
I&M Annual Meeting
San Diego, California
7 February 2006
• I&M / NPS Highlights
• Lessons
• Into the future …
Highlight 1 – Look what we’re doing!!!
Quality, quantity, breadth, relevance
•
•
•
•
•
•
Disturbance
Vegetation change
Land condition
Phenology (plants, ice, permafrost)
Topography (coasts, reefs, etc)
Pattern and context
Programmatic Goals
•
•
•
•
Wise choices
Consistency
Efficiency
Institutional learning
Shared learning among I&M Networks and collaborators:
Cohen, Kennedy – NCCN, SWAN, NCPN, SCPN
Townsend – APHN, NCRN
Reed –
SWAN, NCPN, SCPN
Wang –
NETN, NCBN
Hansen –
HTLN, GRYN
Brock –
GULN, SFCN, SECN, NCBN
With External Partners:
• Workshop – NASA, PCA, CCRS, CSA, NPS
• Ecosystem modeling – NASA (SIEN – YOSE)
• NASA internship program (SIEN, Fire, USFS)
• NASA Proposals and grants:
• Invasive species & fire (Welch, Paintner, Benson, Morrisette)
• Monitoring proposal (Hansen et al.)
• Land use and climate effects on biodiversity in 70 large parks
(Hansen & Running)
• Park Science paper – I&M and NASA (Turner, Nemani, Gross)
• National Phenological Network
• Heinz Center – terrestrial and coastal groups
NDMI, 1989 to 2004
Networks and Landscape Dynamics
Draft protocols:
• NCCN
• GRYN
• NCRN
Bandelier National Monument
March 4, 1999
NCCN Protocol – Warren Cohen and Robert Kennedy
• Very large and remote parks
• Landsat focus: cheap, consistent, historical, good near short-wave
• Track changes in broad physiognomic classes
• Many changes described as proportional mixture changes
broadleaf/grass/crop
increasing canopy
conifer
shadow
water
burn
DATE 1
soil
snow/ice/cloud
NCCN Protocol – Warren Cohen and Robert Kennedy
• Very large and remote parks
• Landsat is core of effort
• Cheap, consistent, historical, good near short-wave sensor
• Track changes in broad physiognomic classes
• Many changes described as proportional mixture changes
• Multi-tiered validation approach
• Focus on pixel-based products
• Solid foundation for post-map analysis
• Facilitate patch or super-pixel pattern analyses
GRYN Protocol – Hansen / Jones
• Large area
• Land use intensification in critical habitats
• Excellent conceptual models linking landscape change to resources
• Extensive use of remotely sensed and ancillary data
Linking landscape change to park resources
Mechanism
Type of effect
Monitoring
Change in
effective size of
reserve
Species area effect
Minimum dynamic area
Tropic structure
Land use and habitat
area
Disturbance patterns
Wildlife populations
Changes in ecological
flows into and out of
reserve
Disturbance initiation and runout
zones
Placement in watershed or airshed
Disturbance patterns
Loss of crucial habitat
outside of reserve
Ephemeral habitats
Dispersal or migration habitats
Population source / sink habitats
Land use and habitat location
Animal movements
Water & air quality
Impoundments / hydrology
Animal demography
Increased exposure to
human activity at
reserve edge
Poaching
Displacement
Exotics / disease
(modified from Hansen and DeFries in prep)
Human density
Human activity
Exotics / disease
Public data
Spatial Dataset
Source
Housing and population density
U.S. Census Bureau (2000)
Water discharge permit records
State Department of Environmental Quality; U.S.
EPA
Land cover
Conventional water pollution
Hydrologic modification
Cities
Overall population change
Change in farmland acreage
USGS, NLCD; NOAA CCAP, LandFire
EPA National Watershed Characterization
EPA National Watershed Characterization; NPS
impoundments database
National Atlas of the United States
U.S. Census Bureau
U.S. Census of Agriculture; State Agriculture
Statistics Services
Trends in major dam construction
U.S. Army Corp of Engineers and FEMA, National
Inventory of Dams
Changes in housing density
U.S. Census Bureau, “Profile of Selected Housing
Characteristics”
Plus: roads, lights, imagery archives
(modified from Hansen and Gryskiewicz 2003)
NCRN Protocol – Townsend, Gardner, & Lookingbill
• Many small parks in rapidly urbanizing
landscape
• Effects of imagery resolution
• Pattern analysis based on graph
theory
• Comprehensive testing and review of
protocol
(Figure: Townsend et al. draft protocol)
Lessons leaned
Many opportunities for broad-scale analyses
• Core vital signs,
• Major potential to use inexpensive, widely-available data,
• Change detection - use of inexpensive high-frequency, coarseresolution data to strategically acquire expensive data,
• Scale of objectives consistent with USGS, EPA, NOAA, PCA,
• Potential for program-wide efficiencies in data processing and
analysis,
• Potential collaborations at local to international scales.
Finer-scale landscape dynamics (often vegetation change)
• Partnership opportunities at regional, network or biome scale
• Many more network- or park-specific issues
• Change detection is a very big issue (resolution, cost)
Parks Canada Approach
Large collaborative project with limited set of objectives:
• Habitat fragmentation / pattern
• Vegetation succession / retrogression
• Vegetation productivity
• Biodiversity (species richness)
Lesson:
Efficiencies from a highly focused group with clearly
objectives. Very rapid progress and consistency.
Agencies: PCA, CCRS, CSA, Universities
What’s on the horizon?
How can we best monitor linear park units?
NETN, GLKN, HTLN
Appalachian Trail & river-based parks
What’s in the future
Emergence of a National Phenological Network
• Seasonal changes are one of the most pervasive environmental
variations on Earth
• Effects seen in agriculture, transportation, health, hydrology, etc.
• Direct link between monitoring results and broader social values
• http://www.uwm.edu/Dept/Geography/npn/
Why we want a National Phenological Network
• Priority vital sign for multiple networks,
• Standardized protocols,
• Ability to use and contribute to broader context,
• Leverage activities by others,
• Excellent means to link and add value to other measures.
Implementation team meeting – March 22-23, 2006
Involves USGS, USFS, EPA, NOAA, NASA, NPS, universities
What’s in the future
Greater use of ecosystem modeling for monitoring and management
• Rama Nemani, NASA Ames – Terrestrial Observation and Prediction
System (TOPS).
• Current link to NASA internship program
• Pilot project with SIEN – Yosemite NP
• Hope to expand to Island Royale
• Educational process
http://ecocast.arc.nasa.gov/
(figure from http://ecocast.arc.nasa.gov/)
What’s in the future
Coordinated acquisition of regional to national data?
• Focus on broad-scale data sets:
• Land cover, roads, population, agricultural records, pollution, etc.
• Linkages to MRLC, Landfire
• Consistent evaluation, system-wide context
MRLC land cover zones
Landscape Dynamics Breakout Session – focus on partnerships
Landfire and how it’s going to help us – Dr. Kevin Ryan, USFS
Parks Canada’s approach – Dr. Donald McLennan
NASA DEVELOP interns, NPN, modeling – John Gross
Selected Resources
Landscape dynamics
•
Landscape dynamics web site: http://science.nps.gov/im/monitor/lulc/LULC.cfm
•
NASA TOPS - http://ecocast.arc.nasa.gov/
•
NASA DEVELOP Internship program - http://develop.larc.nasa.gov/
Phenology and climate change:
•
National Phenological Network - http://www.uwm.edu/Dept/Geography/npn/
•
European Phenological Network - http://www.dow.wau.nl/msa/epn/index.asp
•
Pacific West Region Climate Change Page http://inside.nps.gov/regions/region.cfm?rgn=223&lv=3
John Gross 970 267-2111, [email protected]
http://science.nature.nps.gov/im/monitor
Remote Sensing and Landscape Dynamics