3_TDWG OSR model new

Download Report

Transcript 3_TDWG OSR model new

Definition of an Observation
In general, an observation represents the measurement of some
attribute, of some thing, at a particular time and place.
Observations can be gathered using sensors or sensor arrays,
direct human observations, or biological specimen collections.
Specifically, an observation characterizes the evidence for the
presence or absence of an organism or set of organisms through a
data collection event at a location.
Observations are not necessarily independent and could be linked via
characteristics such as time, place, protocol, and co-occurring
organisms.
TDWG OSR Interest Group
Focus: Primary biodiversity data from observer-derived
measurements and specimens stored within natural
history collections.
Concept: There is sufficient overlap in observational
data and specimen records to join these data
resources.
Primary Issue: Unique requirements for individual
projects (e.g. protocols, taxonomy, spatial).
Goal: To create an observational data description that
fully integrates specimen and observational data.
Metadata Standards
Darwin Core and Access to Biological Collections Data (ABCD)
Darwin Core: set of basic data element definitions that can be
extended.
ABCD: comprehensive structured hierarchical schema.
Darwin Core and ABCD were developed to facilitate the exchange
biological collections data at an international level.
TDWG OSR will focus on model extensibility and schema
integration.
Priority Issues:
• Look for synergies between specimen and observations
• Taxonomic unity
• Variety of collection methods
• Spatial representation
• Define core data items for data discovery
• Compare data attributes and identify overlap and look for synergies
with other TDWG Interest Groups.
• Support and provide input to the TDWG core ontology.
Key Elements of the Darwin Core
• GUID- Global Unique Identifier
• Record Information- e.g. where, date modified, collection or catalog number
• Taxonomic Organization
• General geographic Information
• Time and Date
• Specimen
Key Extensions to Darwin Core
for the Bird-monitoring Data Exchange
• Project code
• Protocol description
• Survey area description
• Effort information
duration of observation
number of observers
• Number counted
• All species reported
• Data access rights
GBIF
ORNIS
Avian Knowledge Network
Figure 2. The observational data matrix showing that as the type of analysis
(rows) becomes more detailed the information (columns) necessary to be
gathered increases. (http://avianknowledge.net).
The occurrence of Red-breasted Nuthatch
reported during the irruption of November 1998.
While these data are useful for showing the extent
of a species occurrence at the continental scale,
they provide little information at greater
spatial resolution.
(Source:
The Irruptive Bird Survey http://www.birdsource.org/ibs).
Seasonal changes in Common Yellowthroat distribution
in North America during 2006. The frequency of checklists
reported is shown. These maps provide information both
on coverage (e.g. where data were collected) as well as
some indication of how common the species was.
A total of 174,197 checklists were submitted to eBird in
2006 and 10,447 of them included Common Yellowthroat.
The data were collected in eBird (http://www.ebird.org).
Thirty year trends in Northern Flicker occurrence.
This map was generated from observations made over
a 30 year period. Individual Locations (@ 3000) were
sampled repeatedly
The data were collected in Breeding Bird Survey
(http://www.mbr-pwrc.usgs.gov/bbs/bbs.html).
Cross-project Data Integration
Use Data Mining models to predict
the occurrence of grassland species
30,000 observations
•
•
•
Rocky Mountain Bird Observatory
2001 – 2005
10,000 locations
138 Predictors
•
•
NLCD Habitat
– 21 classes
– 6 scales
Climate (EPA)
– Average Precipitation
– Average Snowfall
30 grassland species
• Data federation is crucial for ecological
analysis at appropriate scales.
• Current techniques for organization have
been an excellent first step.
• Creation of a new observational ontology
must build off these early successes.