Invasive Species, Vegetation, and the Semantic Web

1
Invasive Species, Vegetation, and the Semantic Web
Jim Quinn, Jim Thorne, Allan Hollander, Christine
Meynard, David Waetjen, Josh Viers, Deanne
DiPietro Information Center for the Environment
and California Information Node, NBII University
of California, Davis
2
Collaborators

• NSF-ITR SPIRE
• UM (Hendler)
• UMBC (Finan)
• RMBL (Martinez)
• Goddard (Schnase)
• USGS/NBII
• NIISS
• Caltrans
• Cal Resources Agency

• CalEPA
• CALFED
• TNC
• Sonoma Ecology Center
• Conservation International
• (plus many other data contributors)

3
Model
Species locality Data (plots)
Range Models GAM/ANN/ CART For current and
future climate scenarios
Demographic Dispersal Species tolerances parameter
s
Climate grids Current and Future, dynamic
monthly GCMs
Input Data
Base Geographic Data- DEM, soils Veg, study extent
Spatially Explicit Demographic Model-
Biomove Midgley et al.
Species range modeling in California
Model
4
(No Transcript)
5
(No Transcript)
6
Species Locality Data Plot data collected by
various groups for different projects- compiled
for modeling effort.
7
Conversion of plot species data to input data for
modeling
8
Selection of study area exent
9
Sample Range Model Current and future ranges
both shown in one image.
10
CalJep Map Units 283 units for California
11
(No Transcript)
12
Range Map Allium parvum
13
California Floristic Province- Early basis for
ecoregional approach
14
(No Transcript)
15
The Semantic WebTools for Creating a Standard
Data Reporting Structure

• XML Extensible Markup Language
• Emerging as a standard way to exchange structured
  data Expresses semantics (vocabularies, etc.)
  of a network
• RDF Resource Description Framework
• Good for expressing the ontologies (relations
  between data resources) ina network
• Other Open-Source technologies
• Web Ontology Language (OWL)
• Open GIS (OGIS) and Geographic Mark-up Language
  (GML)
• Ecological and sensor markup languages?

16
CRISISMaps An Open Source Invasive Species
Mapping System
17
Yellow Starthistle by Township in California
18
Project Catalogs

• California efforts mostly project catalogs
• Natural Resource Project Inventory (NRPI)
• Initiated by California Biodiversity Council
• Projects, Restoration Projects, CalWeed
• CALFED/SFEI
• Prop. 50 projects
• FGDC Clearinghouse / NBII catalog / NRIS
• Federally funded projects
• EML Clearinghouses (LTER, NCEAS)
• NSF-style research projects
• others
• Redundant and high-maintenance
• Limited information content best for
  information discovery

19
http//www.ice.ucdavis.edu/nrpi/NRPIDescription.as
p?ProjectPK622
20
(No Transcript)
21
Arundo donax

• Has invaded thousands of acres in Southern
  California
• Destroys riparian habitat
• Flood and fire hazard

22
The Study SiteCamp Pendleton Marine Corps Base
23
AVIRIS Data

• Low-altitude AVIRIS
• 3.9 m pixels
• Collected Sept., 2000
• Calibrated to reflectance with ACORN (Atmospheric
  Correction Now)

24
Ground Spectral Data

• Spectral data collected for several species and a
  large parking lot
• Locations recorded with GPS
• Used as input into atmospheric calibration and to
  evaluate its success, and for ROIs

25
Maximum Likelihood Classification

• Very good results obtained with .98 threshold
• Visual inspection shows that most areas are
  classifying correctly

26
Techniques Compared

• Continuum removal
• Unsupervised classifications
• MNF/Supervised classifications
• MNF/Spectral Angle Mapper
• Methods explored for guiding endmember selection

27
RipTopo Riparian Corridor
28
RipTopo Aquatic Conservation
29
Citizen monitoring

• Critical for applications such as detecting new
  invasive species
• Data quality concerns
• Treat with metadata, not filters
• Attributes?
• Annotation
• Digital field forms Standardized output
• Example NAWMA
• Digital field identification tools

North American Weed Management Association
30
(No Transcript)
31
Citizen monitoring

• Information / IT Needs
• Better help materials
• Standard vocabularies
• Methods to automatically upload data
• Or better yet, post locally and crawl
• Semantic Web technologies
• Methods to push data to recipients
• RSS feeds?
• Spatial modeling and map-based visualization

32

Any Questions?