What is the Future of the Brazilian Amazon The Challenges of Spatial Information Modelling

1
What is the Future of the Brazilian Amazon? The
Challenges of Spatial Information Modelling

• Gilberto Câmara
• Director for Earth Observation
• National Institute for Space Research
• Brazil

2
About...

• Gilberto Câmara is Director for Earth Observation
  at INPE.
• Eletronics Engineer (ITA, 1979) with a PhD in
  Computer Science (INPE, 1995).
• Research interests
• Geographical information science, spatial
  databases, spatial analysis and remote sensing
  image processing
• Achievements
• Leader in the development of GIS and Image
  Processing technology in Brazil
• Co-chair of the Brazilian Research Network on
  Environmental Modelling of the Amazon

3
INPE - brief description

• National Institute for Space Research
• main civilian organization for space activities
  in Brazil
• staff of 1,800 ( 800 Ms.C. and Ph.D.)
• Areas
• Space Science, Earth Observation, Meteorology and
  Space Engineering

4
Environmental activities at INPE

• Numerical Weather Prediction Centre
• medium-range forecast and climate studies
• LANDSAT/SPOT Receiving and Processing Station
• in operation since 1974
• China-Brazil Earth Resources Satellite
• 5 bandas (3 visible, 1 IR) at 20 m resol.
• Research Activities in Remote Sensing
• 300 MsC and PhD graduates
• ONU-funded Center for Africa and S. America

5
What is an Information Science Problem?

• Multidisciplinary issue
• Different agents with conflicting interests
• Computer representation is only part of the
  problem
• Rôle of the information science expert
• Bring together expertise in different field
• Make the different conceptions explicit
• Make sure these conception are represented in the
  information system

6
The Future of Brazilian Amazon

• Why is this an information science problem?
• Amazonia is a key environmental resource
• Many different concerns
• Environment and biodiversity conservation
• Economic development
• Native population

7
The forest...
Source Carlos Nobre (INPE)
8
Source Carlos Nobre (INPE)
9
The rains...
Source Carlos Nobre (INPE)
10
The rivers...
Source Carlos Nobre (INPE)
11
Amazonia at a glance ... The Natural System

• Almost 6 million km2 of contiguous tropical
  forests
• Perhaps 1/3 of the planet's biodiversity
• Abundant rainfall (2.2 m annually)
• 18 of freshwater input into the global oceans
  (220,000 m3/s)
• Over 100 G ton C stored in vegetation and soil
• A multitude of ecosystems, biological and ethnic
  diversity

Source Carlos Nobre (INPE)
12
Population Growth and Land Use Change

• Modern occupation of Amazonia (since 1500)
  negligible land use change up to the 1960's, but
  large loss of ethnic diversity due to
  colonization
• Large land use change in the last 30 years
• Close to 600,000 km2 deforested in Brazilian
  Amazonia (15)
• High annual rates of deforestation (15,000 to
  30,000 km2/year)

Source Carlos Nobre (INPE)
13
Understanding Deforestation in Amazonia
14
Deforestation...
Source Carlos Nobre (INPE)
15
Fire...
Source Carlos Nobre (INPE)
16
Fire...
Source Carlos Nobre (INPE)
17
But there are millions of the beings All so well
disguised That no-one asks From where such people
come Chico Buarque
Source Carlos Nobre (INPE)
18
Amazon Deforestation 2003
Deforestation 2002/2003
Deforestation until 2002
Fonte INPE PRODES Digital, 2004.
19
Scientific Challenges

• Third culture
• Modelling of physical phenomena
• Understanding of human dimensions
• How to combine man-climate-earth?

20
Challenges of Sustainable Development

• Unlike other factors of production (such as
  capital and labor), natural resources are
  inflexible in their location. The Amazonian
  Forest is where it is the water resources for
  our cities cannot be very far away from them.
• The challenge posed by sustainable development
  is that we can no longer consider natural
  resources as indefinitely replaceable, and move
  people and capital to new areas when existing
  resources become scarce or exhausted there are
  no new frontiers in a globalized world.
• (Daniel Hogan)

21
Sustainability Science Core Questions

• How can the dynamic interactions between nature
  and society be better incorporated in emerging
  models and conceptualizations that integrate the
  earth system, human development and
  sustainability?
• How are long-term trends in environment and
  development, including consumption and
  population, reshaping nature-society interactions
  in ways relevant to sustainability?
• What determines vulnerability/resilience of
  nature-society interactions for particular places
  and for particular types of ecosystems and human
  livelihoods?

Source Sustainability Science Workshop,
Friibergh, SE, 2000
22
Sustainability Science Core Questions

• Can scientifically meaningful limits or
  boundaries be defined that would provide
  effective warning of conditions beyond which the
  nature-society systems incur a significantly
  increased risk of serious degradation?
• How can todays relatively independent activities
  of research planning, monitoring, assessment and
  decision support be better integrated into
  systems for adaptative management and societal
  learning?

Source Sustainability Science Workshop,
Friibergh, SE, 2000
23
Public Policy Issues

• What are the acceptable limits to land cover
  change activities in the tropical regions in the
  Americas?
• What are the future scenarios of land use?
• How can food production be made more efficient
  and productive?
• How can our biodiversity be known and the
  benefits arising from its use be shared fairly?
• How can we manage our water resources to sustain
  our expected growth in urban population?

24
The Importance of Environmental Data

• Our knowledge of earth system science is very
  incomplete
• Support for earth science modelling
• Understanding of processes
• Supporting conjectures and refutations
• Helps address sustainability science questions
• From scientific questions to public policy issues
• Data collection brings new questions and helps
  formulate new ones
• Breaking the five orders of ignorance

25
Causes for Land Use Change

• Government plans to integrate Amazonia
• Build road network throughout the region
• Population growth in Amazonia 3,5 million in
  1970, up to 20 million in 2000, though 65 living
  in large and mid-size cities and towns
• Colonization projects rush of landless people to
  small scale, low tech agriculture
• Subsidized cattle ranching
• Destructive logging as a vector to subsequent
  deforestation
• Large-scale soybean agriculture

Source Carlos Nobre (INPE)
26
Deforestation in Amazonia
PRODES (Total 1997) 532.086 km2 PRODES (Total
2001) 607.957 km2
27
1 9 7 3
28
1 9 9 1
Courtesy INPE/OBT
29
1 9 9 9
Courtesy INPE/OBT
30
LBA Flux Towers on Amazonia
Source Carlos Nobre (INPE)
31
Source Carlos Nobre (INPE)
Biodiversity...
32
CBERS Image
33
What do we do with so much spatial data?

• First, we collect it...
• GPS, remote sensing, field surveys
• Data conversion
• Then, we organize it...
• Spatial modelling
• Spatial databases
• Spatial visualization
• But more important is to analyse and understand
  it!

34
Space
Objects
Actions
Material world
Events
Space is a system of entities and a system of
actions Milton Santos
35
Spatial Data
Natural Domain
Human Domain
INFRASTRUCTURE -roads -utilities -dams
CADASTRAL DATA -parcels -streets -land use
CENSUS DATA -Demographics -Economics
IMAGES -planes -satellites
ENVIRONMENTAL DATA -topography -soils -temperatur
e -hidrography -geology
36
FROM DATA TO COMPUTER REPRESENTATION
X,Y,Z
X,Y,Z
X,Y,Z
X,Y,Z
X,Y,Z
EVENTS / POINT SAMPLES
SURFACES / REGULAR GRIDS
AREA DATA / POLIGONS
FLUX DATA / NETWORKS
37
Remote Sensing
LANDSAT 5 TM image of São Paulo, 1997
38
Aerial Photos
Favela da maré, Rio de Janeiro - 2001
39
Choropletic Maps
São Paulo - 96 districts per capita income
São Paulo 270 survey areas per capita income
40
Trend Surfaces
iex
Social Exclusion 1995
Social Exclusion 2002
41
FLUXES
42
The First Law of Geography

• Toblers Law
• Everything is related to everything else, but
  near things are more related than distant things
• We call this spatial dependence
• Can we see Toblers law in action?
• Yes, there are lots of exemples...Here are
  some....

43
The Future of Brazilian Amazonia?

• Scenarios for Amazônia in 2020 (Laurance et al.,
  Science)
• Optimistic scenario
• 28 of deforestation
• Pessimistic scenario
• 42 of deforestation
• Whats the real science behind this work?

44
The Future of Brazilian Amazonia(Laurance)

• Optimistic scenario
• Complete degradation up to 20 km from roads
  (existing and projected)
• Moderate degradation up to 50 km from roads
• Reduced degradation up to 100 km from roads
• Pessimistic scenario
• Complete degradation up to 50 km from roads
  (existing and projected)
• Moderate degradation up to 100 km from roads
• Whats wrong with this approach?

45
Scenarios and Models

• Scenarios require models!
• Models
• Describe quantitatively a phenomenon and predict
  its evolution in space and time
• A model must answer
• What changes?
• When changes take place?
• Where changes take place?
• Why are there changes?

46
Modelling and Laurances work

• The Future of the Brazilian Amazon?
• What changes?
• Is constrast forest-deforestation enough?
• Where changes take place?
• Model is spatially explicit - OK
• When changes take place?
• No change equations
• Why are there changes?
• Model does not indicate causes

47
Alternatives to Simplistic Models

• Multidisciplinary work
• Geography, Demography, Antropology, Computer
  Science, Statistics, Ecology
• Use of empirical evidence
• Census surveys
• On-situ visit
• Remote Sensing
• Models grounded on hard data

48
Competition for Space
Loggers
Competition for Space
Source Dan Nepstad (Woods Hole)
49
What Drives Tropical Deforestation?
of the cases
? 5 10 50
Underlying Factors driving proximate causes
Causative interlinkages at proximate/underlying
levels
Internal drivers
If less than 5of cases, not depicted here.
sourceGeist Lambin (Université Louvain)
50
Source LUCC
51
Modelling and Public Policy
External Influences
System Ecology Economy Politics
Desired System State
Decision Maker
Scenarios
Policy Options
52
Modelling Tropical Deforestation
Coarse 100 km x 100 km grid
Fine 25 km x 25 km grid
53
Factors Affecting Deforestation
54
Coarse resolution candidate models
55
Coarse resolution Hot-spots map
56
Modelling Deforestation in Amazonia

• High coefficients of multiple determination were
  obtained on all models built (R2 from 0.80 to
  0.86).
• The main factors identified were
• Population density
• Connection to national markets
• Climatic conditions
• Indicators related to land distribution between
  large and small farmers.
• The main current agricultural frontier areas, in
  Pará and Amazonas States, where intense
  deforestation processes are taking place now were
  correctly identified as hot-spots of change. 

57
Fatores Correlacionados ao Desmatamento

• Sete fatores estão relacionados à variação de 83
  das taxas de desmatamento na Amazônia nos últimos
  anos (a) Estrutura Agrária (2 fatores)
  percental de área ocupada por grandes fazendas e
  número de pequenas propriedades. (b) Ocupação
  Populacional (1 fatores) densidade de população.
  (c) Condições do Meio Físico (2 fatores)
  Precipitação média e percentual de solos férteis.
  (d) Infraestrutura (1 fator) distância a
  estradas. (e) Presença do Estado (1 fator)
  percentagem de áreas indígenas

58
Clocks, Clouds or Ants?

• Clocks
• Paradigms Netwons laws (mechanistic,
  cause-effect phenomena describe the world)
• Clouds
• Stochastic models
• Suporte Teoria de sistemas caóticos
• Formigas
• Modelos emergentes
• Suporte teoria de sistemas complexos
• Exemplos automata celulares

59
Modelos Espaciais Princípios Básicos
f ( I (t) )
f ( I (t1) )
f ( I (t2) )
f ( I (tn ))

• Célula localização
• Input processo ocorre no lugar (ex. chuva)
• Função entrada -gt estado

F
F
. . .
60
Ambientes Computacionais para Modelagem
superfície discreta de células retangulares
multivaloradas possivelmente não contíguas
61
O modelo ambiental
X

• Um ambiente possui 3 submodelos
• Modelo Espacial espaços celulares regiões
  GPM
• Modelo Comportamental teoria de sistemas
  autômatos celulares híbridos agentes situados
• Modelo Temporal simulador de eventos discretos
  definidos de forma recorrente
• A estrutura espacial e temporal é compartilhada
  por vários agentes.

62
A estrutura do espaço é heterogênea
Ambientes definidos de forma recorrente
É possível construir modelos multiescalas
Porções distintas do espaço podem ter escalas
diferentes
63
Ambiente Computacional de Modelagem TerraLib
Geoinfo (Aguiar, 2003), Submetido GIScience
(Câmara et al, 2004)
64
Deforestation 2002/2003
Deforestation until 2002
Laurance et al., 2001 Pessimist scenario
(2020)
Savannas, non-forested areas, deforested or
heavely degrated
Moderately degrated
Lightly degrated
Pristine
Fonte INPE PRODES Digital, 2004.
65
Conjectures and Refutations on Third Culture...

• Amazon Deforestation Models Challenging the
  Only-Roads Approach
• Deforestation predictions presented by Laurance
  et al. are based on the assumption that the road
  infrastructure is the prime factor driving
  deforestation.
• Deforestation rates have increased significantly
  in the last two years, but very few Federal
  investments on roads have effectively been made
  since the 80s.
• Simplistic models such as Laurance et al. may
  deviate attention from real deforestation causes,
  being potentially misleading in terms of
  deforestation control
• There is an urgent need to understand the genesis
  of the new Amazon frontiers.

66
How Ethical is Science Judgment?

• From Brian White ltmailtobwhite_at_aaas.orggt gt
• Date 09/02/04 095522 gt
• TO laurancew_at_tivoli.si.edu ltmailtolaurancew_at_tivo
  li.si.edugt gt
• Dear Dr. Laurance,
• We have recently sent letters about your Policy
  Forum published in Science to which you have
  responded. Following is another letter we have
  received about the same paper. If possible, we
  would like your response to this comment as well.
  
• Sincerely, Etta Kavanagh Associate Letters Editor

67
Environmental Modelling in Brasil

• GEOMA Rede Cooperativa de Modelagem Ambiental
• Cooperative Network for Environmental Modelling
• Established by Ministry of Science and Technology
• INPE/OBT, INPE/CPTEC, LNCC, INPA, IMPA, MPEG
• Long-term objectives
• Develop computational-mathematical models to
  predict the spatial dynamics of ecological and
  socio-economic systems at different geographic
  scales, within the framework of sustainability
• Support policy decision making at local, regional
  and national levels, by providing decision makers
  with qualified analytical tools.  

68
The Road Ahead Can Technology Help?

• Advances in remote sensing are giving computer
  networks new eyes and ears.
• Sensors detect physical changes and then send a
  signal to a computer.
• Scientists expect that billions of these devices
  will someday put the environment itself online.
• (Rand Corporation, The Future of Remote Sensing)

69
The Road Ahead Smart Sensors
SMART DUST Autonomous sensing and communication
in a cubic millimeter
Sources Silvio Meira and Univ Berkeley,
SmartDust project
70
Limits for Models
Uncertainty on basic equations
Social and Economic Systems
Quantum Gravity
Particle Physics
Living Systems
Global Change
Chemical Reactions
Applied Sciences
Meteorology
Solar System Dynamics
Complexity of the phenomenon
source John Barrow
71
The Road Ahead...

• Producing environmental data in the Americas
• Tremendous impact of in the management of our
  natural resources
• Task outside of the resources and capabilities of
  a single country
• Breaking the bottleneck
• Establishment of continental research networks
• Adherence to agreed international protocols
  (Biodiversity Convention, Kyoto Protocol)

72
The Rôle of Science and Scientists

• Science is more than a body of knowledge it is a
  way of thinking. ...The method of science ...
  is far more important than the findings of
  science. (Carl Sagan)
• Scientists have to understand the sensitivities
  involved in collecting, using and disseminating
  environmental data