@PhDThesis{Brito:2016:EsCeEm,
author = "Brito, Alan de",
title = "Heterogeneidade espa{\c{c}}o-temporal do desmatamento do Cerrado
brasileiro: estimativas e cen{\'a}rios de emiss{\~o}es de
carbono",
school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
year = "2016",
address = "S{\~a}o Jos{\'e} dos Campos",
month = "2016-05-30",
keywords = "savana brasileira, emiss{\~a}o de gases de efeito estufa,
cen{\'a}rios, desmatamento, cerrado, brazilian savanna, carbon
book-keeping, scenarios, deforestation.",
abstract = "O objetivo deste trabalho foi o de promover o entendimento da
rela{\c{c}}{\~a}o entre a din{\^a}mica espa{\c{c}}o-temporal
do processo de desmatamento e a heterogeneidade intrarregional do
Cerrado Brasileiro, a fim de subsidiar a constru{\c{c}}{\~a}o de
um modelo espacialmente expl{\'{\i}}cito de estimativas de
emiss{\~o}es atuais e futuras de gases de efeito estufa (GEE).
Para tanto, foi constru{\'{\i}}do um banco de dados espacial e
multitemporal, onde foram inseridas informa{\c{c}}{\~o}es de
distribui{\c{c}}{\~a}o de desmatamento, de densidade de biomassa
e 36 vari{\'a}veis ambientais, de acessibilidade a mercados,
atratividade econ{\^o}mica, demogr{\'a}ficas, tecnol{\'o}gicas,
de estrutura agr{\'a}ria, sociais e econ{\^o}micas. A
an{\'a}lise do padr{\~a}o do desmatamento ocorrido no bioma foi
realizada por meio da constru{\c{c}}{\~a}o de Mapas de Kernel e
da an{\'a}lise do desmatamento por Estado da
Federa{\c{c}}{\~a}o e tipo de forma{\c{c}}{\~a}o vegetacional
(campestre, sav{\^a}nica e florestal). Foi tamb{\'e}m realizada
a An{\'a}lise de Agrupamentos para promover a
estratifica{\c{c}}{\~a}o do Cerrado de acordo com as
caracter{\'{\i}}sticas ambientais e socioecon{\^o}micas com o
intuito de caracterizar o desmatamento de acordo com as
subregi{\~o}es estratificadas. Adicionalmente, foram estimadas as
emiss{\~o}es de carbono por desmatamento de forma espacialmente
expl{\'{\i}}cita, personalizando o arcabou{\c{c}}o
gen{\'e}rico de modelagem INPE-EM para o bioma: o INPE-EM
BRCerrado. De acordo com os resultados obtidos com a
caracteriza{\c{c}}{\~a}o do desmatamento, o modo n{\~a}o
espacial do modelo (1850-2002) foi parametrizado, analisando o
carbono liberado para cada regi{\~a}o estratificada e avaliando
as incertezas das estimativas associadas. Foram realizadas
estimativas para tr{\^e}s diferentes cen{\'a}rios futuros de
desmatamento at{\'e} 2030, tomando como base
informa{\c{c}}{\~o}es dispon{\'{\i}}veis na literatura
cient{\'{\i}}fica e hip{\'o}teses baseadas na revis{\~a}o do
C{\'o}digo Florestal Brasileiro (CFB) e na
Contribui{\c{c}}{\~a}o Nacionalmente Determinada Pretendida
(iNDC) do pa{\'{\i}}s. Por fim, foram estimadas as emiss{\~o}es
l{\'{\i}}quidas (emiss{\~a}o CO\$_{2}\$ por corte raso -
absor{\c{c}}{\~a}o CO\$_{2}\$ por regenera{\c{c}}{\~a}o)
para os tr{\^e}s cen{\'a}rios, de acordo com cada tipo de
forma{\c{c}}{\~a}o vegetacional. Os resultados da an{\'a}lise
espa{\c{c}}o-temporal do desmatamento do Cerrado mostraram que os
desmatamentos ocorridos at{\'e} o ano de 2002 e entre 2002 e 2010
apresentam padr{\~o}es distintos. At{\'e} 2002, o desmatamento
incidiu mais sobre as forma{\c{c}}{\~o}es campestres (54,5\%),
seguidas pelas forma{\c{c}}{\~o}es sav{\^a}nicas (41,4\%) e
florestais (44,2\%). Entre 2002 e 2010, a maioria das perdas de
vegeta{\c{c}}{\~a}o nativa (26,54\%) ocorreu {\`a} custa da
supress{\~a}o de forma{\c{c}}{\~o}es sav{\^a}nicas. At{\'e}
2010 46,27\%, 69,5\% e 49,54\% das forma{\c{c}}{\~o}es
sav{\^a}nicas, campestres e florestais foram perdidos,
respectivamente. As emiss{\~o}es brutas de gases de efeito estufa
para o Cerrado Brasileiro entre 1850 e 2010 totalizaram 10,8
GtonCO\$_{2}\$ e (2,95 PgC). Estimativas para o per{\'{\i}}odo
2002-2010, obtidas utilizando dados oficiais de desmatamento
espacialmente expl{\'{\i}}citos, foram de 1,37 GtonCO\$_{2}\$
e (0,37 PgC). Em rela{\c{c}}{\~a}o ao total, as
forma{\c{c}}{\~o}es florestais s{\~a}o as maiores fontes de
emiss{\~a}o, seguidas pelas sav{\^a}nicas e por {\'u}ltimo as
campestres. Entre 2002 e 2010, aproximadamente 86\% das
emiss{\~o}es foram concentradas na frente de expans{\~a}o do
desmatamento localizada na regi{\~a}o de MATOBIPA (estados do
Maranh{\~a}o, Tocantins, Piau{\'{\i}} e Bahia), e nos estados
de Goi{\'a}s e Minas Gerais. As an{\'a}lises de sensibilidade
revelaram que as estimativas de {\'a}reas desmatadas foram uma
grande fonte de incertezas nas estimativas de emiss{\~o}es
(46\%). Adicionalmente, foram comparadas as emiss{\~o}es futuras
potenciais de acordo com os tr{\^e}s diferentes cen{\'a}rios:
(a) desmatamento \${''}\$legal \${''}\$ de acordo com a
revis{\~a}o do CFB; (b) continuidade das taxas hist{\'o}ricas de
desmatamento e; (c) meta alternativa de redu{\c{c}}{\~a}o de
emiss{\~a}o para o Cerrado, no contexto da iNDC Brasileira. Os
resultaram indicaram que sem a{\c{c}}{\~o}es pol{\'{\i}}ticas
adicionais, o cumprimento do CFB pode levar {\`a}
libera{\c{c}}{\~a}o de 0,55 PgC (2 GtonCO\$_{2}\$) na
atmosfera. Por outro lado, a meta alternativa de
redu{\c{c}}{\~a}o da emiss{\~a}o associada ao desmatamento do
Cerrado, apresentou potencial de redu{\c{c}}{\~a}o de
aproximadamente 60\% (0,81 GtonCO\$_{2}\$ ou 0,22 PgC) at{\'e}
2030. Finalmente, as emiss{\~o}es l{\'{\i}}quidas m{\'e}dias
dos Cen{\'a}rios A, B e C foram estimadas em 100 MtonCO\$_{2}\$
\∙ano\$^{-1}\$, 49 MtonCO\$_{2}\$
\∙ano\$^{-1}\$, e 41
MtonCO\$_{2}\$\∙ano\$^{-1}\$,, respectivamente,
reconsiderando a recomposi{\c{c}}{\~a}o de 100\% do passivo de
Reserva Legal entre os anos de 2016 e 2030. ABSTRACT: This work
aimed to understand the relationship between the spatial-temporal
dynamics of forest clearing and the intra-regional heterogeneity
in the Brazilian Cerrado, in order to subsidize the development of
a spatial explicit model for actual and future emissions estimates
of greenhouse gases (GHG). For this, a spatial and multitemporal
geodatabase was built with information regarding forest clearing
distribution, biomass density, and 36 variables concerning market
accessibility, economic attractiveness, demography, technology,
agrarian infrastructure, social, and economic data. Forest
clearing pattern analysis in the biome was performed through the
analysis of Kernel maps, the forest clearing data in the Brazilian
States and vegetacional units (grassland, savanna, and forest).
The biome was stratified in regions according to its environmental
and socioeconomic characteristics through a Cluster Analysis in
order to characterize the forest clearing patterns within these
regions. Additionally, carbon emissions from deforestation were
spatially explicit estimated by adapting the INPE-EM modelling
framework to the biome: the INPE EM BRCerrado. The non-spatial
mode of the model (1850-2002) was parameterized according to the
results obtained from the forest clearing characterization. The
carbon released was analyzed within the stratified regions and the
associated uncertainties were evaluated. Carbon emissions
estimates for three different future scenarios of forest clearing
were performed for the biome up to 2030. These scenarios were
based on information available in scientific literature and
hypotheses based on both Brazilian Forest Code revision and in the
\emph{intended Nationally Determined Contribution} (iNDC).
Lastly, Net Emissions (CO\$_{2}\$ emissions from clear cut - CO2
absorptions through regeneration) were estimated for the three
scenarios according to the vegetation unit. The results of the
deforestation spatial-temporal analysis revealed that forest
clearing that had occurred up to 2002 and between 2002 and 2010
had distinct patterns. Up to 2002, deforestation occurred mainly
on grasslands (54.5\%), followed by savannas (41.2\%). Between
2002 and 2010, the most of the native vegetation loss (26,5\%)
occurred on the expenses of savanna formations. By 2010, 46.27\%,
69.5\%, and 49.54\% of savannas, grasslands, and forest were
lost, respectively. Between 1850 and 2010, gross deforestation
emissions for the Brazilian Cerrado totaled 13.44 GtonCO2e (3.66
PgC). Estimates for the period 2002-2010, obtained using official
spatially explicit deforestation estimates were of 1.438
GtonCO\$_{2}\$ e (0.40 PgC). We found that, relative to total
emissions, woodland removal were the main sources of GHG from land
cover change, followed by shrublands and grasslands. Between 2002
and 2010, \$\approx\$86\% of the emissions were concentrated
in the MATOPIBA region (Maranh{\~a}o, Tocantins, Piau{\'{\i}},
and Bahia federal states), an expanding deforestation frontier,
and also in federal states of Goi{\'a}s and Minas Gerais. Our
sensitivity analysis revealed that dissimilar deforestation
estimates were the major source of estimates uncertainty (46\%)
followed by the heterogeneous distribution of biomass across space
(17.5\%). Additionally, we compared potential future emissions
according to three different scenarios: (a) legal deforestation
according to the Brazilian Forest code (b) historical
deforestation rates continuity; (c) alternative emission reduction
target for the Cerrado in the context of the Brazilian Intended
Nationally Determined Contribution (iNDC). Our results indicated
that without further policy actions, the compliance with the
Brazilian Forest Code could lead to the release of 3.7
GtonCO\$_{2}\$e (1.01 PgC) into the atmosphere. On the other
hand, our alternative target for emissions reduction in the
Cerrado, showed the potential to decrease emissions in 60\% (0.81
GtonCO\$_{2}\$ or 0.22 PgC) at{\'e} 2030. Finally, average Net
Emissions of the scenarios A, B, and C were estimated in 100
MtonCO\$_{2}\$ \∙ano\$^{-1}\$, 49 MtonCO\$_{2}\$
\∙ano\$^{-1}\$ e 41 MtonCO2\∙ano\$^{-1}\$,
respectively, considering the recomposition of 100\% of the Legal
Reserve debt between 2016 and 2030.",
committee = "Alval{\'a}, Pl{\'{\i}}nio Carlos (presidente) and Aguiar, Ana
Paula Dutra de (orientadora) and Oliveira, Gilvan Sampaio de and
Bustamante, Mercedes Maria da Cunha and Acerbi J{\'u}nior, Fausto
Weimar",
copyholder = "SID/SCD",
englishtitle = "Space-time heterogeneity of the brazilian cerrado deforestation:
estimates and scenarios of carbon emissions",
language = "pt",
pages = "174",
ibi = "8JMKD3MGP3W34P/3LSQB68",
url = "http://urlib.net/ibi/8JMKD3MGP3W34P/3LSQB68",
targetfile = "publicacao.pdf",
urlaccessdate = "02 maio 2024"
}