@TechReport{XimenesAmarVale:2012:ImCeGe,
author = "Ximenes, Arimat{\'e}a de Carvalho and Amaral, Silvana and
Valeriano, Dalton de Morrisson",
title = "Os impactos dos cen{\'a}rios gerais de mudan{\c{c}}as
clim{\'a}ticas sobre as ecorregi{\~o}es da amaz{\^o}nia legal
brasileira",
institution = "Instituto Nacional de Pesquisas Espaciais",
year = "2012",
type = "RPQ",
number = "INPE-17104-RPQ/888",
address = "S{\~a}o Jos{\'e} dos Campos",
keywords = "ecoeregi{\~o}es, amaz{\^o}nia, mudan{\c{c}}as
clim{\'a}ticas.",
abstract = "Este relat{\'o}rio apresenta a metodologia adotada e os
resultados obtidos pelo projeto BIOMA, referentes ao subprojeto
que buscou estudar os impactos dos cen{\'a}rios gerais de
mudan{\c{c}}as clim{\'a}ticas sobre as ecorregi{\~o}es da
Amaz{\^o}nia Legal Brasileira. Para subsidiar a discuss{\~a}o
destes impactos, foi necess{\'a}rio construir uma base de dados,
e gerar um mapa de ecorregi{\~o}es com crit{\'e}rios objetivos.
Uma abordagem quantitativa, semiautom{\'a}tica, e baseada em
dados de sensoriamento remoto e redes neurais permitiu gerar o
mapa de ecorregi{\~o}es a partir de m{\'u}ltiplas camadas de
informa{\c{c}}{\~o}es. As vari{\'a}veis altitude, declividade,
densidade de drenagem, porcentagem de cobertura de dossel,
dist{\^a}ncia aos principais rios amaz{\^o}nicos, longitude e
latitude foram selecionadas por possuem significado biol{\'o}gico
associado a defini{\c{c}}{\~a}o de ecorregi{\~o}es. Ap{\'o}s a
sistematiza{\c{c}}{\~a}o em um banco de dados geogr{\'a}fico em
espa{\c{c}}os celulares, estas vari{\'a}veis foram submetidas ao
algoritmo de rede neural do tipo Self-Organizing Map (SOM) para
que finalmente as regi{\~o}es homog{\^e}neas fossem agrupadas a
partir de k-m{\'e}dias. Ap{\'o}s an{\'a}lise e
edi{\c{c}}{\~a}o, 60 ecorregi{\~o}es foram mapeadas para a
Amaz{\^o}nia Legal, excluindo-se as {\'a}reas mapeadas como
desmatamento recente pelo projeto PRODES. Para discutir o efeito
das mudan{\c{c}}as clim{\'a}ticas obteve-se um mapa
s{\'{\i}}ntese das diferen{\c{c}}as de temperatura e
precipita{\c{c}}{\~a}o fornecida pelo cen{\'a}rio A2 de altas
emiss{\~o}es do Modelo de Proje{\c{c}}{\~a}o Regional
ETA\$_{-}\$CCS (50 km) para o clima presente (1961 a 1990) e
futuro (2070 a 2100). Os locais de maiores diferen{\c{c}}as de
temperatura e precipita{\c{c}}{\~a}o foram confrontados com o
mapa de ecorregi{\~o}es de modo a identificar as {\'a}reas de
maior impacto potencial sobre a biodiversidade. A
distribui{\c{c}}{\~a}o das ecorregi{\~o}es frente ao mosaico
atual das unidades de conserva{\c{c}}{\~a}o permitiu identificar
{\'a}reas priorit{\'a}rias para a conserva{\c{c}}{\~a}o. As
ecorregi{\~o}es terrestres identificadas a partir de m{\'e}todos
quantitativos neste trabalho podem auxiliar e complementar
informa{\c{c}}{\~o}es para o gerenciamento dos recursos
florestais e no planejamento sistem{\'a}tico de
conserva{\c{c}}{\~a}o frente {\`a}s mudan{\c{c}}as
clim{\'a}ticas previstas. ABSTRACT: This report presents the
methodology adopted and the results achieved by the sub- project
\textquotedblleft{Impacts of climate change scenarios over the
Brazilian Amazon Ecoregions}\textquotedblright that was part of
BIOME - INPE/CENPES-Petrobr{\'a}s project. To support the
discussion of these impacts, it was initially necessary to build a
database, and to generate a map of ecoregions. A quantitative and
semi-automatic approach, based on remote sensing data and neural
networks procedure allowed the creation of a map of ecoregions
from multiple layers of information. The variables altitude,
slope, drainage density, percent of canopy coverage, distance to
major Amazonian rivers, longitude and latitude were selected
because they have biological significance for ecoregions
definition. After the systematization in a geographic database in
cellular spaces, a neural network algorithm, the Self-Organizing
Map (SOM) processed these variables and then, homogeneous regions
were grouped from k-means. After analysis and edition, 60
ecoregions were identified in the Amazon region. Areas mapped as
recent deforestation by PRODES project were excluded from the
ecoregion mapping. To discuss the effect of climate change, we
computed the differences in temperature and rainfall provided by
high emissions (A2) scenario of the regional projection model
ETA\$_{-}\$CCS (50 km) for the present (1961 to 1990) and future
(2070 to 2100) climate. Sites with the highest differences in
temperature and precipitation were compared to the ecoregions
mapping to point out the areas of greatest potential impact over
biodiversity. Ecoregions distribution over the current mosaic of
conservation units also suggested priority areas for conservation.
The terrestrial ecoregions identified in this work by quantitative
methods can assist and provide complementary information for the
management of forest resources and systematic planning of
conservation as mitigation action for climate change.",
affiliation = "{} and {Instituto Nacional de Pesquisas Espaciais (INPE)}",
copyholder = "SID/SCD",
language = "pt",
pages = "35",
ibi = "8JMKD3MGP7W/3CKR4SB",
url = "http://urlib.net/ibi/8JMKD3MGP7W/3CKR4SB",
targetfile = "publicacao.pdf",
urlaccessdate = "17 maio 2024"
}