The Geostat Hypotheses In search of a high resolution foundation for Geostatistics in Europe an inte

1
The Geostat HypothesesIn search of a high
resolution foundation for Geostatistics in
Europe an interactive response to operational
user needs in a time of transition.

• Lars H. Backerlars.backer_at_scb.se
• Statistics Sweden
• Nordic Forum for GeoStatistics (NFGS)
• European Forum for GeoStatistics (EFGS)
• European GridClub

2
7 Proposals / Hypotheses

• Message
• Spaceship Earth
• Cybernetics
• Paradigm
• STDI (Spatial and Temporal Data
  Infrastructure)
• GIIS (Global Integrated Information System)
• Extended ESS
• ESSgrid (The GEOSTAT project)
• Two Conclusions

3
1. Spaceship Earth
4
1. Spaceship Earth

• That the earth is, due to the success of the
  modernistic project, no longer to be regarded as
  self-regulative deterministic. The Earth is a
  closed system.
• Blinded by our greed and lack of regard for our
  condition, we have soiled our nest and threaten
  destroy the prospects for future generations.
• This means that the community of public
  authorities have to take command and
  responsibility, and spread the wings of our
  intellect to fly or perish.
• This is not so much a metaphysical as a very
  pragmatic quest

5
2. Cybernetics
6
2. Cybernetics

• The global system cannot be re-built, but we know
  how to develop complex man environmental systems
• Man-machine interactions are controlled with
  systems of human knowledge (private and/or shared
  systems of reference)
• Cybernetics is the science of regulating
  information in closed systems. With the
  informational revolution we can build powerful
  cybernetic systems where both complex actions and
  feedback systems interact.
• This will require the cooperation between a more
  scientific practice and a more practical science
  to build and use a shared operational model of
  vital structures and processes.

7
3. Paradigm
W. BRADFORD PALEY Map of science in the journal
Nature, SEED and Discover Magazines
8
3. Paradigm
9
3. Paradigm

• To develop a closed system we will need a
  holistic or ecological approach that may
  require the cooperation of all key fields of
  human knowledge
• If we cannot describe it we cannot manage it.
  We will require a system of models that is as
  simple as possible, but not too simple.
• Although only the most critical information
  should be included, this will require close
  cooperation between sciences and professions that
  have lived in vitro.
• We need to formulate no less than an outline for
  a world culture

10
4. STDI
11
4. STDI

• Dynamic Models

12
4. STDI

• Narratives (Background information (grand
  narratives) that describes a problem in its
  context.).
• Description of the reference system upon which
  the project is founded
• Description of the project or projects designed
  to develop targeted systems
• Object (concept) based semantic models. Objects
  may be described as systems of systems that form
  hierarchical structural and functional entities).
  
• Spatial object models (Kits of parts
  structural models)
• Spatial and temporal object models
  (Interacting whole dynamic models)
• Spatial and Temporal Model1
• Models 1 Statistics. (Statistics and systems of
  statistics)
• Non spatial object at a given point in time
  (static)
• Non spatial objects over a time period (dynamic)
• Models 2 Cartographics2 (Maps and systems of
  maps)Objects described with focus on spatial
  representations (Points, pixels (2d),voxels (3d))
• Spatial object at a given point in time (static)
• Spatial objects over a time period (dynamic)
• Reference information. (Documents and collections
  of related documents)
• Reference information 1 (What to do) Information
  from the humanities (efforts to discuss values
  etc. in relation to actions)
• Reference information 2 (How to do) Information
  from science (efforts to describe and explain
  systems)

13
4. STDI

• We need to share the same models of the world
  worldwide to serve coordinated actions. This must
  be done through the internet. This requires a
  shared spatial data infrastructure suited for
  this purpose.
• We need the necessary infrastructure to discuss
  whole paradigms. This requires a data
  infrastructure that can handle both narratives
  and models.
• Narratives require conventional alpha- numeric
  data infrastructures (texts and databases with
  reference catalogues)
• Models require data infrastructures for spatial
  data (maps) integrated with empirical
  observations (statistics)

14
5. GIIS

• Data infrastructures (Infrastructures of data)
• Focus
• Man-made environment
• Narrative (digital and/or analogue documents)
• Properties (statistics)
• Cartographies (geodata)
• Reference literature (reference library)
• Context
• Sociocultural system
• Narrative (digital and/or analogue documents)
• Properties (statistics)
• Cartographies (geodata)
• Reference literature (reference library)
• Economic system
• Narrative (digital and/or analogue documents)
• Properties (statistics)

15
5. GIIS

• Modeling will require that complex man
  environmental systems are broken down to
  subsystems and subsystems of subsystems.
• One method is first human systems vs. natural
  systems. Further to be divided into natural
  manmade environment vs. social, cultural
  economic human systems (of systems).
• The GIIS proposal is a proposal for the
  structuring of spatial and temporal models of man
  environmental systems. The state of the art for
  modeling man environmental systems is to use an 2
  or 3d object strategy that is based on semantic
  models / networks.
• Spatial and Temporal models are processed with
  GIS / GIT. GI(I)S models are based on a layer
  method where all objects belonging to the same
  class is presented with the same variables in the
  same layer. Each object has its own distinct
  geography.

16
Extended ESS

• Iteration 1
• Analyse 1 (Exante information for engineering)
• Design 1
• Produce 1
• Manage 1 (Expost information for accountant)
• Iteration 2
• Analyse 2 (Exante information for engineering)
• Design 2
• Etc
• Iteration n
• How to integrate D. (Ex post information for
  accountant) and A. (Ex ante information for
  engineering) into one integrated system!

17
Extended ESS
From Torsten Hägerstrands paper from 1955
Statistiska primäruppgifter flygkartering och
data processing maskiner. (Statistical micro
data, aircraft surveillance and data processing
machines An integrating project.
18
6. Extended EES

• Harmonized foundation for both grids and
  blobs
• A harmonized system for small area statistics for
  accounting purposes. (an ESSreg system for
  official statistics)
• A harmonized system of small area statistics for
  engineering purposes. (an ESSgrid system for
  official statistics)

19
7. ESS_grid

• Modelling society

A city is not a tree.
20
7. ESS_grid
21
7. ESS_grid

• We have a well functioning statistical system of
  statistics aggregated to administrative areas
  designed and built serve accounting purposes.
• For building a global infrastructure of spatial
  information these traditional statistical systems
  represents an obstacle. Their data are not
  suitable for the engineering purposes required.
• The current global challenge will a statistical
  systems that allows for integrated spatial
  analysis that involves qualified statistics from
  any available source. This calls for a shared of
  statistics that covers the whole earth, is
  liberated from administrative areas, and that
  allows for hierarchical analysis of datasets on
  all scale ranges from global to local.

22
Two Conclusions

• The ESS should stand on two legs. The current to
  aggregate statistics to administrative areas for
  official accounting purposes, and the other a
  general, full coverage hierarchical system for
  any kind of spatial statistics to serve current
  pressing engineering tasks.
• The GEOSTAT, hopefully to be an ESSnet project
  aims to introduce a general purpose grid based
  system of spatial statistics for the NSIs and
  Europe

23
Population explosion? Resource
constraints?Climate Change?

• Solvable Problems?
• Not without GeoStatistics !

24
Thank You!

• The Geostat ProjectEuropean Forum for
  GeoStatistics
• Statistics Sweden
• lars.backer_at_scb.se