@MastersThesis{Rodrigues:2019:ReNoPr,
author = "Rodrigues, Danilo Avancini",
title = "Din{\^a}mica dos fatores de degrada{\c{c}}{\~a}o florestal em
fronteira agropecu{\'a}ria da Amaz{\^o}nia: a regi{\~a}o de
Novo Progresso, Par{\'a}",
school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
year = "2019",
address = "S{\~a}o Jos{\'e} dos Campos",
month = "2019-08-19",
keywords = "regress{\~a}o multivariada, degrada{\c{c}}{\~a}o florestal,
fronteira agropecu{\'a}ria, multivariate regression, forest
degradation, logging frontier.",
abstract = "A degrada{\c{c}}{\~a}o florestal na Amaz{\^o}nia, causada pela
explora{\c{c}}{\~a}o seletiva de madeira e pelo fogo florestal,
{\'e} um dos principais impactos decorrentes da
intensifica{\c{c}}{\~a}o da ocupa{\c{c}}{\~a}o dessa
regi{\~a}o, com din{\^a}mica e extens{\~a}o ainda pouco
conhecidas. As altera{\c{c}}{\~o}es na vegeta{\c{c}}{\~a}o,
causadas por degrada{\c{c}}{\~a}o florestal, ocorrem de maneira
gradual e no longo prazo, demandando longos per{\'{\i}}odos de
observa{\c{c}}{\~a}o. Os fatores que influenciam a
degrada{\c{c}}{\~a}o florestal variam ao longo tempo e se
relacionam com os est{\'a}gios de desenvolvimento de uma
regi{\~a}o. Este trabalho identificou os principais fatores, e
suas respectivas contribui{\c{c}}{\~o}es (pesos), relacionados
com a intensidade da degrada{\c{c}}{\~a}o florestal na
regi{\~a}o de Novo Progresso, munic{\'{\i}}pio da Amaz{\^o}nia
paraense, em diferentes per{\'{\i}}odos. Essa regi{\~a}o {\'e}
considerada uma ativa fronteira agropecu{\'a}ria, cuja
ocupa{\c{c}}{\~a}o se intensificou a partir dos anos 2000. O
procedimento metodol{\'o}gico foi realizado em duas etapas: na
primeira, foi realizado o mapeamento da degrada{\c{c}}{\~a}o
florestal com t{\'e}cnicas de processamento de imagens e, na
segunda, foram identificados os fatores explicativos que
influenciam a intensidade da degrada{\c{c}}{\~a}o florestal. Na
etapa de mapeamento foi realizada a classifica{\c{c}}{\~a}o
espectral de imagens anuais de 2012 a 2017 para a
identifica{\c{c}}{\~a}o de fei{\c{c}}{\~o}es relativas {\`a}
degrada{\c{c}}{\~a}o florestal. Em seguida, foi realizada uma
classifica{\c{c}}{\~a}o estrutural, para o mapeamento de
padr{\~o}es de intensidade de degrada{\c{c}}{\~a}o florestal.
Os dados produzidos foram incorporados a um banco de dados,
formando uma s{\'e}rie hist{\'o}rica de intensidade de
degrada{\c{c}}{\~a}o florestal na regi{\~a}o de Novo Progresso
de 34 anos (1984 a 2017). Esta s{\'e}rie hist{\'o}rica foi
dividida em tr{\^e}s per{\'{\i}}odos que refletem diferentes
contextos do desenvolvimento da regi{\~a}o de estudo: 1984 a
1994, 1995 a 2004 e 2005 a 2017. A obten{\c{c}}{\~a}o e
an{\'a}lise dos fatores que influenciam a intensidade de
degrada{\c{c}}{\~a}o florestal foram realizadas por meio de
modelos de regress{\~a}o multivariados. Os modelos demonstraram
que o desmatamento e as estradas, representados pelas
vari{\'a}veis dist{\^a}ncia de {\'a}reas desmatadas,
propor{\c{c}}{\~a}o de {\'a}rea desmatada, dist{\^a}ncia de
estradas vicinais e dist{\^a}ncia da BR-163, influenciaram a
degrada{\c{c}}{\~a}o florestal durante todos os
per{\'{\i}}odos de estudo. A vari{\'a}vel dist{\^a}ncia da
BR-163 foi a de maior contribui{\c{c}}{\~a}o para a
degrada{\c{c}}{\~a}o florestal no primeiro per{\'{\i}}odo,
apresentando menor contribui{\c{c}}{\~a}o no segundo
per{\'{\i}}odo, n{\~a}o sendo significativa no terceiro. Esse
resultado evidencia o esgotamento dos recursos florestais {\`a}s
margens da BR- 163. No segundo per{\'{\i}}odo, a vari{\'a}vel
dist{\^a}ncia de estradas vicinais {\'e} inclu{\'{\i}}da mas
apresenta uma contribui{\c{c}}{\~a}o relativa inferior no
{\'u}ltimo per{\'{\i}}odo. A dist{\^a}ncia de {\'a}rea
desmatada teve sua contribui{\c{c}}{\~a}o ascendente, sendo o
fator de menor contribui{\c{c}}{\~a}o no segundo
per{\'{\i}}odo, e o de maior contribui{\c{c}}{\~a}o no
{\'u}ltimo per{\'{\i}}odo. A vari{\'a}vel dist{\^a}ncia de
assentamentos rurais foi inclu{\'{\i}}da nos dois {\'u}ltimos
per{\'{\i}}odos, enquanto a vari{\'a}vel densidade de focos de
calor apareceu apenas no terceiro per{\'{\i}}odo. Esses
resultados mostram a heterogeneidade de fatores que influenciam a
degrada{\c{c}}{\~a}o florestal ao longo do tempo e sua
identifica{\c{c}}{\~a}o {\'e} fundamental para subsidiar
a{\c{c}}{\~o}es e pol{\'{\i}}ticas para um controle mais
eficaz da degrada{\c{c}}{\~a}o florestal na Amaz{\^o}nia.
ABSTRACT: The forest degradation process in Amazonia, through
selective logging and forest fires, is one of the main impacts
from the intensification of the occupation in Brazilian Amazon,
with dynamics not yet well known. Different from the clear cut,
forest degradation changes occur gradually and in long-term, which
requires long observation period. The factors related to forest
degradation, such as forest fragmentation and logging activity,
vary through time, related to the stages of development where it
occurs. This works purpose is to identify the main factors, and
its weights, related to the forest degradation intensity in Novo
Progresso region, in different time periods. This region is
considered an active logging frontier which occupation was
intensified after 2000. The methodological procedures were
separated in two steps: the first step is mapping forest
degradation through digital image processing from Landsat sensors
and the second step is the identification of the factors related
to forest degradation using multivariate statistical analysis.
Both the steps were sustained by field trips to the study area. On
the mapping step, images were spectrally classified for the years
2012 to 2017 to identify indicators of forest degradation. Still
on the mapping phase, a second step structurally classified the
forest degradation polygons into degraded forest intensity
patterns. The data produced were combined with the database of
forest degradation intensity from Pinheiro (2015), resulting in a
33-years (1984-2017) time-series of forest degradation intensity
in the region of Novo Progresso. This time series were divided in
three time periods that reflects the study area context of
development: 1984 to 1994, 1995 to 2004 and 2005 to 2017. The
acquisition and analysis of the factors influencing forest
degradation intensity were performed through multivariate
regression models. The models showed that deforestation and roads,
represented by the variables distance from deforested areas,
deforested area proportion, distance from roads and distance from
official roads influenced forest degradation during the study
periods. The variable distance from rural settlements was included
in the last two periods of analysis because people living in the
settlements rely on wood as the primary source of money so they
can invest on cattle raising and/or farming. The distance from the
BR-163 was the factor with the highest contribution to forest
degradation in the first period, turning out to be the variable
with the lowest contribution in the second period, and the
variable distance from unofficial roads becomes the most important
variable. On the last period the distance from BR-163 was not
included in the model, while the distance from unofficial roads
presented a lower relative contribution than the previous periods,
exposing the process of forest resources depletion on the margins
of the BR- 163 between the two last periods. The distance from
deforested areas had been ascending throughout the periods. On the
last period, the kernel heat focus was the variable with the
second highest influence on forest degradation intensity. It was
not found evidences that measures taken by the federal government
to control and reduce forest cover removal had any influence in
forest degradation on Novo Progresso, evidencing the lack of
governance in the region. The results can subsidize individual
actions for a more efficient control of forest degradation in
logging frontiers in the Amazon with context and history similar
to Novo Progresso.",
committee = "Kampel, Silvana Amaral (presidente) and Escada, Maria Isabel
Sobral (orientadora) and Monteiro, Antonio Miguel Vieira and
Renn{\'o}, Camilo Daleles and Silva, Edson Jos{\'e} Vidal da",
englishtitle = "Forest degradation factors' dynamics in an Amazonian logging
frontier: the region of Novo Progresso, Par{\'a}",
language = "pt",
pages = "99",
ibi = "8JMKD3MGP3W34R/3TRFQPB",
url = "http://urlib.net/ibi/8JMKD3MGP3W34R/3TRFQPB",
targetfile = "publicacao.pdf",
urlaccessdate = "2024, Apr. 25"
}