Databases and Global Environmental Change

1
Databases and Global Environmental Change

• Gilberto Câmara
• Diretor, INPE

2
The fundamental question of our time
How is the Earths environment changing, and what
are the consequences for human civilization?
source IGBP
3
Earth is a system of systems
Human actions are changing the balance!
4
(No Transcript)
5
Earth as a system
6
Impacts of global environmental change By 2020 in
Africa, agriculture yields could be cut by up to
50
sources IPCC and WMO
7
Precipitation anomalies (2071-2100)- (1961-90)
in mm/day
Seco
Seco
Seco
Seco
A2
B2
Climate change scenarios in Brazil
Quente
Quente
B2
A2
Temperature anomalies (2071-2100)- (1961-90) in
oC
8
Average temp raised 0,7 C in 50 years in Brazil
Tmin up 1 C!
Source (Obregón e Marengo, 2007)
9
Impacts on Agriculture
Fonte Eduardo Assad, Embrapa
10
Collapse of Amazon Rain Forest?
2100
2000
savanna
forest
caatinga
pastures
desert
Is there a tipping point for Amazonia?
source Oyama and Nobre, 2003
11
Impacts on Water Availability in NE Brazil
Hidrological Balance NE Brazil
1961-1990

Less Water for Agriculture!
2071-2100
Source Marengo and Salati, 2007
12
Amazônia in 2005
source Greenpeace
13
Amazônia in 2015?
fonte Aguiar et al., 2004
14
Great challengeDatabase support for earth
system science
source NASA
15
Global Change
Where are changes taking place? How much change
is happening? Who is being impacted by the
change?
16
Global Land Project

• What are the drivers and dynamics of variability
  and change in terrestrial human-environment
  systems?
• How is the provision of environmental goods and
  services affected by changes in terrestrial
  human-environment systems?
• What are the characteristics and dynamics of
  vulnerability in terrestrial human-environment
  systems?

17
Data chain in Earth System Science
fonte NASA
18
INPEs supercomputers and worlds TOP 500
1 trend
Sum top 500
150 TF
500 trend
5 TF
2 TF
40 GF
INPE (MPP equivalent peak performance)
8 GF
19
Earth System Science Data Handling
PetaFlop Centres
Large Scale Data
Megascenarios
Índice de Vegetação
Regional Centers
Regional Scenarios
Policy Options
20
Global Earth Observation System of Systems
Vantage Points
Capabilities
L1/HEO/GEO TDRSS Commercial Satellites
Far-Space
Permanent
LEO/MEO Commercial Satellites and Manned
Spacecraft
Near-Space
Airborne
Aircraft/Balloon Event Tracking and Campaigns
Deployable
Terrestrial
User Community
Forecasts Predictions
21
Weather and climate
11,000 land stations (3000 automated) 900
radiosondes, 3000 aircraft 6000 ships, 1300
buoys 5 polar, 6 geostationary satellites
source WMO
22
ARGOS Data Collection System (16000 plats)
650,000 messages processed daily
23
Data collection services
Tracking Positions collected over a fixed period
of time
Monitoring Data from remote stations, fixed or
mobile
24
Argo bouy network
25
I am the Walrus
26
Models From Global to Local
Athmosphere, ocean, chemistry climate model
(resolution 200 x 200 km) Atmosphere only
climate model (resolution 50 x 50 km) Regional
climate model (resolution e.g 10 x 10
km) Hydrology, Vegetation Soil Topography (e.g,
1 x 1 km) Regional land use change Socio-economic
changes Adaptative responses (e.g., 10 x 10 m)
27
Data integration enables crucial links between
nature and society
Nature Physical equations Describe processes
Society Decisions on how to Use Earths
resources
28
ST DBMS-21

• augmented reality

• mobile devices

• Data-centered, mobile-enabled, contribution-based,
  field-based modelling

• sensor networks

• ubiquitous images and maps

29
Slides from LANDSAT
source USGS
Databases and Change A Research
Programme Understanding how humans use
space Predicting changes resulting from human
actions Modeling the interaction between
society and nature
Aral Sea
1973
1987
2000
Bolivia
1975
1992
2000
30

• How can DBMS technology handle Earth System
  Science data?
• What algebra is needed for spatio-temporal data?
  How can this algebra be handled in an
  object-relational DBMS?

31
Identity conditions on ST data

• Average temp for IPCC scenarios

• Continuous fields (x,y,z,t)

• Continuous fields (x,y,z,t)

32
Identity conditions on ST data
land_cover cells in 1985
land_cover cells in 2000
Individual objects (id, t,(x,y,z))
33
Identity conditions on ST data Images
M. Silva, G.Câmara, M.I. Escada, R.C.M. Souza,
Remote Sensing Image Mining Detecting Agents of
Land Use Change in Tropical Forest Areas.
International Journal of Remote Sensing, vol 29
(16) 4803 4822, 2008.
Remotely sensed images are ontologically
instruments for capturing landscape dynamics
34
Identity conditions on ST data Images
Landsat Image 13/Ago/2003
35
Identity conditions on ST data Images
Deforestation 13/Ago/2003 until 07/Mai/2004
Deforestation in 13/Aug/2003 (yellow)
deforestation from 13/Aug/2003 until 07/mai/2004
(red)
36
Identity conditions on ST data Images
Deforestation on 21/May/2004
Deforestation in 13/Aug/2003 (yellow)
deforestation from 13/Aug/2003 until 07/May/2004
(red) deforestation on 21/May/2004 (orange)
37
Identity conditions have uncertain cases!
Furacão Catarina (março/2004) Imagem NASA
38
Modelling changefrom practice to theory
Outiline of a theory for change modelling in
spatio-temporal data
39
What is a geo-sensor?
What is a geo-sensor?
Basic spatio-temporal types S set of locations
(space) T set of intervals (time) I set of
identifiers (objects) V set of values
(attributes)
measure (s,t) v s ? S - set of locations in
space t ? T - is the set of times. v ? V - set
of values
40
What is a geo-sensor?
What is a geo-sensor?
Field (static) field S?V The function field
gives the value of every location of a space
measure (s,t) v s ? S - set of locations in
space t ? T - is the set of times. v ? V - set
of values
41
Slides from LANDSAT
snap (1973)
snap (1987)
snap (2000)
Aral Sea
Time-varying fields are modelled by
snapshots snap T ? Field snap T ? (S ? V)
The function snap produces a field with the
state of the space at each time.
Bolivia
snap (1975)
snap (1992)
snap (2000)
42
Sensors sources of continuous information
43
Sensors water monitoring in Brazilian Cerrado
Wells observation 50 points 50 semimonthly
time series (11/10/03 06/03/2007)
Rodrigo Manzione, Gilberto Câmara, Martin Knotters
44
Fixed sensors time series (histories)
Well 30 Well 40 Well 56 Well 57
hist S ?(T ? V) each sensor (fixed location)
produces a time series
45
Evolving (modifiable) object
life I ? (T ? (S,V)) The function life produces
the evolution of a modifiable object
46
A lifes trajectory
life I ?(T?(S,V)) The life of the object is
also a trajectory
47
Which objects are alive at time T and where are
they?
exist T ? (I?(S,V))
48
Models From Global to Local
snap T ? (S ? V) evolution of a
landscape hist S ? (T ? V) History of a
location
exist T ?(I ? (S,V)) objects alive in a time T
life I ? (T ? (S,V)) the life of an object in
space-time
49
A model for time-varying geospatial data....
set
Temporal entity
is-a
is-a
T-field (coverage set)
T-object hist(oi) (feature)
has-a
snap(t) (coverage t)
has-a
Feature instancet
has-a
location
T-fields have snapshots
T-objects have histories
50
ST DBMS as a basis for data integration
Visualization (TerraView)
Modelling (TerraME)
Spatio-temporal Database (TerraLib)
Data Mining(GeoDMA)
Statistics (aRT)
51
GIS-21 Dynamical modelling integrated in a
spatio-temporal database
G. Câmara, L. Vinhas, G. Queiroz, K. Ferreira,
A.M.V. Monteiro, M. Carvalho, MA Casanova.
TerraLib An open-source GIS library for
large-scale environmental and socio-economic
applications. In B. Hall, M. Leahy (eds.),
Open Source Approaches to Spatial Data
Handling. Berlin, Springer, 2008.
Spatio-temporal database
52
GIS-21 Dynamical modelling integrated in a
spatio-temporal database
53
GIE-21 Network-based analysis
Emergent area
Consolidated area

• Modelling beef chains in Amazonia

54
GIS-21 Dynamical spatial modellingwith Agents
in Cell Spaces
Tiago Garcia de Senna Carneiro, "Nested-CA A
Foundation for Multiscale Modelling of Land Use
and Land Cover Change. PhD Thesis, INPE, june
2006
TerraME Based on functional programming
concepts (second-order functions) to develop
dynamical models
55
TerraAmazon open source software for
large-scale land change monitoring
116-112
116-113
Spatial database (PostgreSQL with vectors and
images) 2004-2008 5 million polygons, 500 GB
images
166-112
56
R-Terralib interface
R data from geoR package.
Loaded into a TerraLib database, and visualized
with TerraView.
57
Conclusions

• Earth System Science data management poses a
  major challenge for the database community
• We need new algebras and data representation and
  handling techniques to deal with ESS data