Tesis
Estimativa do saldo de radiação de uma floresta de transição amazônia-cerrado por sensoriamento remoto
Fecha
2015-02-24Registro en:
MARQUES, Heloisa Oliveira. Estimativa do saldo de radiação de uma floresta de transição amazônia-cerrado por sensoriamento remoto. 2015. xvi, 54 f. Dissertação (Mestrado em Física Ambiental) - Universidade Federal de Mato Grosso, Instituto de Física, Cuiabá, 2015.
Autor
Biudes, Marcelo Sacardi
http://lattes.cnpq.br/7273935697798004
Biudes, Marcelo Sacardi
893.219.011-91
http://lattes.cnpq.br/7273935697798004
Santos, Flávia Maria de Moura
998.134.911-91
http://lattes.cnpq.br/4000857814123856
893.219.011-91
Mello, Geison Jader
283.851.558-64
http://lattes.cnpq.br/9794369843997713
Institución
Resumen
The Amazon has an important role in biodiversity, has the largest expanse of
tropical rainforest in the world. Recently, the use of geotechnologies possible to
identify in real time the changes that occur in the Earth's surface, resulting from
natural and anthropogenic processes several phenomena. Many changes of this level
can be detected by monitoring and determination of radiative exchanges that take
place on the surface. In this sense, the present study aimed to determine the dynamics
of the balance of radiation to the surface by means of images generated by the
satellite Landsat 5 TM sensor in a Transition Forest between Amazon and Cerrado,
in Maracaí Farm, located near Sinop - MT; orbit with 226 and 227, paragraph 68 for
the years 2005 to 2008 were generated letters to vegetation indices, albedo and
surface temperature, instantaneous radiation balance and average daily in 24 hours
using the SEBAL algorithm. The data were validated with measurements carried out
in micrometeorological tower that was installed in the study area. Images for each
year analyzed, to better evaluate the radiative fluxes and estimates that environment
selected. One can see smaller value for the balance of instantaneous radiation during
the dry season, due to haze resulting from burned on site. For Rninst and Rn24h the
surface with respect to the measured data were obtained the following relative and
absolute mean errors with values of 2,4% and 2,0%; 18,2 Wm-2 and 14,3 Wm-2; "r",
and 0,604 to 0,938 and "d" of 0.966 and 0.703 respectively. According to the results
obtained in this work can be said that the methodology for estimation of Rn was
effective.