Uncertainities on land cover and land ue data sets designed for global models

(1)

Uncertainities on land cover

and land ue data sets

designed for global models

Gilberto Câmara

National Institute for Space Research (INPE)

Brazil

IFGI, University of Münster, Germany

http://www.dpi.inpe.br/gilberto

(2)

Why are most remote sensing researchers

like Indiana Jones?

(3)

(4)

The world is divided in cells. Each cell has a single class.

There is a correct classification. The more our classification

approaches the ideal, the better.

(5)

Statistics on land change modelling

Measure correlation between causes and effects

Causes: food production, population increases

Effects: land changes (remote sensing)

y=a₀ + a₁x₁ + a₂x₂ + ... +a_ix_i+E

(6)

Statistics: Assessment of land use drivers

A. Aguiar, G. Câmara, I. Escada, “Spatial statistical analysis of land-use determinants in the Brazilian Amazon”. Ecological Modelling, 209(1-2):169–188, 2007.

G. Espíndola, A. Aguiar, E. Pebesma, G. Câmara, L. Fonseca, “Agricultural land use dynamics in the Brazilian Amazon based on remote sensing and census data”, Applied Geography, 32(2):240-252, 2012.

Land use models are good at allocating change in space. Their problem is: how much change will happen?

(7)

How good are statistical models?

0 10 20 30 40 50 60 70 0 1 2 3 4 5 6 7 E rr o (% )

Log₂(size of comparison window)

Amazonia: estimated x real deforestation

Laurance O Laurance NO

SIMAMAZONIA BAU SIMAMAZONIA GOV Simplificado Vizinhança

(8)

~230 scenes Landsat/year

Yearly detailed estimates of clear-cut areas (wall-to-wall)

Data released openly after 2003

INPE: Clear-cut deforestation mapping of

Amazonia since 1988

(9)

The unbearable lightness of PRODES

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

(37)

Land cover

wetlands tropical forest

Non-natural vegetation shrublands

(38)

Land use

cattle production non-managed forest

Temporary agriculture shifting cultivation

“the arrangements, activities and inputs people

undertake in a certain land cover type to produce,

change or maintain it”

(39)

What happened with 720.000 km2 deforested?

TerraClass - first map of land use and land cover of

Amazonia

(40)

(41)

(42)

(43)

(44)

How are we using the forest?

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% AC MT PA RO Degraded Land Second Veg Degrad Pasture Pasture Small Farms Grains

(45)

What’s in an image?

“Remote sensing images describe landscape dynamics”

(46)

Annual Crop – 1 Crop per season Annual Crop – 2 Crops per season Sugarcane

Crop Year 2002/2003

(47)

Annual Crop – 1 Crop per season Annual Crop – 2 Crops per season Sugarcane

Crop Year 2010/2011

(48)

MATO GROSSO – Sorriso: Crop Year 2002/2003

Annual Crop - 1 Crop Annual Crop - 2 Crops

(49)

MATO GROSSO – Sorriso: Crop Year 2010/2011

Annual Crop - 1 Crop Annual Crop - 2 Crops

(50)

Vegetation index

time series

Área 1

Área 2

Área 3

source: Victor Maus (INPE)

Time series analysis of land change

Forest

Pasture

Forest

(51)

Unique repository of knowledge and data about global land change

Global Land Observatory:

describing change in a connected world

40 years of LANDSAT + 12 years of MODIS + SENTINELs + CBERS Free satellite images Methods for land

change for foresty and agriculture uses

(52)

TerraME Runtime Environment

Eclipse & LUA plugin • model description • model highlight syntax

TerraView • data acquisition • data visualization • data management • data analysis TerraLib database d a ta Model source code MODEL DATA m od e l

• model syntax semantic checking • model execution TerraME INTERPRETER LUA interpreter TerraME framework TerraME/LUA interface m o d e l _data

Uncertainities on land cover and land ue data sets designed for global models