GISdevelopment.net --> Policy --> Geographic Information Infrastructure


Geo-Spatial Data Infrastructure in Brazil : Some Keynotes

Tania Maria Sausen
INPE-CEP, Av. dos Astronuats 1758, Cx.P. 515
CEP 12201-970 São José dos Campos, SP, Brasil
Email: Tania@ltid.inpe.br

The National Insitute for Space Resources-INPE, in Brasil, is the main intitution in space area in Latin America. Since its fundation it has been involved in research and capacity building in remote sensing and natural resources areas. I INPE has played an important role in creating a critical mass, in space area, mainly in remote sensing, in Brasil and Latin America. INPE is the only insitution in the region that have developed a software dedicated to GIS and image processing. At the end of 60's and beginning of 70's INPE created the Remote Sensing Project-SERE, the embryo of the Remote Sensing Division. In June 1973 the Brazilian LANDSAT antena started its operation. In order to create its own researcher team INPE has created a master degree course in remote sensing in 1968 and the doctor degree was created in 1998. In the end of the 70's and begining of the 80's INPE created the SITIM system, the first Brazilian image processing software. In 1986 INPE has created the SGI system, the first Brazilian GIS software. These two softwares were developed in DOS and in Portugues; In the 80's INPE started to developd the Amazonia Deforestation Project-PRODES, the largest forest monitoring project in the world. In 1985, in order to dissemiante and to increase the remote sensing and GIS community in Brazil INPE started the RS Regional Labs Program. Totaday it is possible to find more than 100 labs in all country regions. In 1992 INPE started to developed the SPRING Project; a new geo-processing software;
  • SPRING - (Georeferenced Information Processing System) is a state-of-the-art GIS and remote sensing image processing system with an object-oriented data model which provides for the integration of raster and vector data representations in a single environment.
  • INPE have invested more than 140 men/year on the development of SPRING which has been used for important projects in Brazil such as:
    • Multi-temporal evaluation of deforestation in the Amazonia rain forest;
    • Ecological-Economical Zoning for Brazil;
    • The National Soils Database.
  • SPRING Main Features are:
    • An integrated GIS for environmental, socio-economical and urban planning applications;
    • A Multi-platform system, including support for Windows95/98/NT, Linux and Solaris;
    • A widely accessible freeware for the GIS and remote sensing community, on http://www.dpi.inpe.br 
Monitoring the Brazilian Amazon Forest - PRODES (Deforestation Project)- is the largest forest monitoring project in the world

Its objectives are:
  • to estimate the extent of gross deforesation and annual rate of gross deforestation;
  • to update the digital database
  • to distribute the increments of deforestation :
    • into major forest types
    • into classes of size
The Brazilian Amazon comprises the states of Acre, Amapá, Amazonas, Mato Grosso, Pará, Rondônia, Roraima, Tocantins and portions of Maranhão and Goiás, totaling an area of approximately 5 million square kilometers, large enough to accommodate the entire Western Europe. Of this, approximately 4 million is covered by forest formations.

The concern with a good territorial management has motivated the Government to better understand the human activities in the region, aiming to better guide its policies, including its sustainability, the rational use of the natural resources, and the well-being of the local population.

A fundamental aspect of the Government actions concerns the generation of periodic estimates of the extent and rate of gross deforestation in the Amazon, a task which could never be conducted without the use of space technology.

Brazil carries out these assessments based on the analysis of Landsat images acquired and processed by INPE since 1974. The first complete assessment conducted by the Institute was carried out for years 1974 and 1978. Annual assessments have been conducted by the Institute since 1988, under a project known as PRODES (Estimate of Amazon Gross Deforestation).

Deforestation, here, refers to the conversion of areas of primary forest physiognomy by anthropogenic activities, for the development of agriculture and cattle raising, detected by orbital platforms. Gross deforestation indicates that areas in process of secondary succession or forest recovery are not subtracted in the calculation of the extent and the rate. The above definition does not include areas of forest affected by selective logging activities or wild fires, which are object of specific reports by INPE.

Color composite satellite images at a scale of 1:250,000 allow the identification of changes in forest cover of 6.25 ha and larger. The Brazilian Amazon is covered by 229 of such images, which are interpreted for the boundaries between primary forest and other types of non-forest vegetation.

At each assessment, the areas of new deforestation are delimited. The deforested areas are outlined in overlays, which are subject to a rigorous quality control. When approved, the overlays are scanned and integrated into a Geographical Information System to provide the localization and extent of the deforested areas.

The digital data are converted to a format that allows the presentation of deforestation in any geographical format, for instance, large vegetation classes, municipalities etc.

Each assessment involves approximately 50,000 person x hours, 10,000 machine x hours hired by Fundação de Ciência, Aplicações e Tecnologia Espaciais - FUNCATE. In addition to the 70 specialists involved in the project, a team of 15 researchers from INPE accompanies all phases of PRODES, ensuring its high quality control.
  • AMAZONIA Information System- it is a GIS that integrates deforestation areas in several years, county area and vegetation types.;
  • DATA ANALYSIS- there are three main problems investigated by analysis of the datasets:
    • the estimation of the rate of deforestation per year and among several years (cloud cover, errors that difficult integration of the two datasets, errors of image analysis, digitization and georeferencing due to the poor quality of some maps in the Amazon region)
    • the assessment of how much deforestation has been occuring inside the different types of vegetation (deforestation may have different impacts depending on the type of vegetation in wihch it occurs);
    • the assessment of deforestion by county (the amount of deforestation inside each county may be important for different socio-economical studies);
PRODES, in addition to the extent and rate of gross deforestation, provides the spatial distribution of critical areas in the Amazon. For instance, in 1998, more than 78% of the deforestation was concentrated in 44 of the satellite images analysed. PRODES data are overlaid on the Vegetation Map from the RADAMBRASIL Project-FIBEG to identify the forest physiognomies affected by antropogenic activities.

The distribution of the deforested areas by classes of size is used by governmental and non-governmental agencies as an indicator of the possible causes of deforestation in the Amazon. The annual results presents the distribution of the mean rate of deforestation (% / year) by classes of size. and the distribution of the mean rate of gross deforestation for years for the following thematic classes:
  1. early primary succession communities;
  2. Amazon white-sand woodland;
  3. dense savanna woodland;
  4. contact zone;
  5. seasonally deciduous tropical forest;
  6. open tropical rain forest and dense tropical rain forest, respectively.
The annual rate of gross deforestation in the Brazilian Amazonia tends to concentrate in a relatively small fraction of the 229 TM-Landsat scenes that cover the entire region. For example, for the period 1996/1997, more than 78% of the mean rate of gross deforestation in Amazonia concentrated in 44 scenes, representing approximately 19% of the total of 229 images.

The sample used to generate the provisional estimates of the mean rate of gross deforestation are selected so as to include scenes with the largest rates. This procedure was adopted in 1997, which involved 47 scenes. The estimated rate was later found to be in good agreement with that based on the analysis of all scenes, within 1.5%. The second estimate, for the period 1997/1998, based on 46 images, was in error of 3,1% in relation of the final estimate.

Ecological-Economical Zoning for Brazil
  • This project is carried on by the Brazilian Government:
    • National Institue for Space Research-INPE
    • Brazilian Institute for Statistic and Geography-IBGE
    • Stategic Affairs Office-SAE
  • In 1989 the Brazilian Government has created this program which main objective is to develop a territorial zoning to help the government planning policy, designed to regional development and improvement of social, economical and environmental aspects
  • The two main approches are: agroecological (land use and the human pressure over the environment) and ecological-economical (territory reorganization based on the society/environment relation)
Ecological - Economical Zonning for Brazil
  • Legal Amazonia (5 millon square kilometers) was the priority area in the begining: Acre, Amapá, Amazonas, Mato Grosso, Pará, Rondônia, Roraima, Tocantins and portions of Maranhão and Goiás states
  • This program was also developed in Pernanbuco (northeast region) Mato Grasso do Sul (central-west region) states and Curitba (south region) and São José dos Campos (southeast region) cities.
  • In Amazonia Region were trained 120 people, in 9 multidisciplinar training courses (geologist, geographer, agronomist, forest engineer)
  • All methodology was developed by INPE and IBGE and LANDSAT-TM imagery, SPRING, maps and social-economical datas were used
Ecological - Economical Zoning for Brazil
It were developed maps in the following scales:
  • 1:2.500.000 - 1:1.000.000 - environmental diagnoses about Amazonia;
  • 1:250.000 - ecological-economical zoning from selected area according to economical and environmental criterias;
  • 1:100.000 - studies in critical areas in order to introduce some recovery actions and/or to avoid environmental impacts;
First Phase - Diagnosis
Inventory Disciplinary Analysis
  1. ENVIRONMENT: Datas: Soil, Geology, Biology, Geomorphology, Climate, Hydrology

    PRODUCTS: Maps: soils, geology, geomorphology, climatic agressivity, vegetation, water resources, biodiversity
  2. SOCIAL ASPECTS: Datas: Demography, Health, Economical Activities, Rural, Urban and Economical Infrastructure

    PRODUCTS: Maps: population dinamics, agriculture expansion, landuse, agriculture and cattle production, industrial activities, infrastructure, health
Thematic Analysis - Multi to Interdisciplinary
  1. ECOLOGICAL: Landscape system map (dynamic and vulnerability); Vulnerability matrix
  2. POLITICAL INSTITUTIONAL: Maps of Governmental actions; Land ownership map
RESULTS: TERRITORIAL ORGANIZATION: Population Dynamic map; urban map, Sustainability matrix

INTEGRATION ANALYSIS-INTER AND TRANSDISCIPLINARY

RESULTS - ENVIRONMENTAL QUALITY: Conflict matrices; Environmental quality map

ECOLOGICAL-ECONOMICAL ZONING FOR BRAZIL

SECOND PHASE - PREDICTION

Final Product :
  • Map with guidelines for strategy development
  • Report with guidelines for strategy development
Main Problems to establish a National Geo-Spatial Infrastructure :
  • COMPUTATIONAL INFRASTRUCTURE (computers and softwares); the number of computers available is not enough for the establsihment of a NGDI or the software are quite expensives. The SPRING is the only free software available in the country . Up to now 5000 license were delivery to the community
  • LACK OF BASIC MAPPING IN DIFFERENT SCALES (mainly 1:25.000 and 1:50.000) - sometimes there are no maps available in any scales;
  1. 1:250.000 - 90% of the country, lack of maps in Amazonia region
  2. 1:100.000 - 50% mainly in the South, Southeast and Northeast region
  3. 1:50.000 - less than 40%, mainly in Souht and Southeast reagion, all Brazilian state capitals (26)
  4. 1:25.000 - around 5% just in specific regions like Brasilia, Rio de Janeiro, etc.
  • OUTDATE MAPS (1972, 1973)
  • LARGE TERRITORIAL AREA; 8.5000.0000 Km2
  • INACCESSIBLE AREAS (jungle, deserts, swamp areas)
  • LACK OF ROAD INFRASTRUCTURE;
  • CLOUD COVER (mainly in Amazonia and Northeast region);
  • w LACK OF DATAS (satellite imagery, aerial photografy; environmental, social and economical datas)
  • NO FINANCIAL SUPPORT
  • Lack of Government Policy Dedicated to Establish a NGDI
The most important probelm to establish a NGDI is lack of qualified professionals to establish, to analyse the datas and mainly to get all information available ina a NGDI (sometimes less than 20% of information are really used)

Usually the professionals are extremely able to manage the softwares, the computers, to create and to set up a database, but not rare, they do not know what to do with all information available in the databases.

Usually they are more concerned in setting up the database than to analyse the datas, or to get all information that is possible to have from the databases.

To set up a database it is not necessary to be a professional with very high qualification. It is necessary to have a very good manager, well qualified and which know a lot about database and to have a very good professional in computational area. In order to introduce the datas in the database a very good technician is able to do a very good job. The manager and the professional in computer area know how to set up and to control the system, the technician role is to know how to introduce the datas in the database.

But to analyse, to get all information from the databases, it is necessary to have a very qualified professional, that have all knowledge about the datas, how to integrate these datas, how to get all information available, and how to use them for different purposes, like urban and regional planning, road infrastructure, environmental impacts, etc. These kind of professionals are, geographers, urbanist, agronomists, geologists, etc, that are concerned with the data information, not with the computer and software operation

"Having a very good database without professionals qualified to get all information from the datas is the same of having a wonderful and red FERRARI car, in a bad road with a very bad driver".

How to Solve the Problem :
Capacity Building, Education, Training not only in operating the softwares and computer, not only in seting up the databases, but mainly to take advantages from the datas and to solve the social, economical and environmental problems.

© GISdevelopment.net. All rights reserved.