Banca de DEFESA: ALIBIA DEYSI GUEDES DA SILVA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : ALIBIA DEYSI GUEDES DA SILVA
DATE: 14/02/2025
TIME: 14:00
LOCAL: Google Meet
TITLE:

MAPPING LAND USE AND VEGETATION COVER IN THE SEMI-ARID OF RIO GRANDE DO NORTE: A STRATEGY BASED ON MACHINE LEARNING AND CLOUD COMPUTING

KEY WORDS:

Semiarid; Classification; Google Earth Engine; Algorithms.

PAGES: 149
BIG AREA: Ciências Humanas
AREA: Geografia
SUMMARY:

Landscape transformations resulting from the replacement of native forests by human activities have generated significant impacts on natural areas. Mapping land use and cover is essential to support territorial planning, since changes in Caatinga vegetation in semi-arid environments are associated with rainfall variability and anthropogenic action. In view of this, this research aims to map land use and land cover in the Caatinga biome of the state of Rio Grande do Norte for the year 2023, using machine learning algorithms and cloud computing. In addition, to analyze the relevance of environmental variables, especially climatic components, in the spatial behavior of vegetation patterns. The LANDSAT-8 collection was processed on the Google Earth Engine (GEE) platform. The methodological approach involved two main stages: spatial prediction of precipitation and supervised classification, both using the Random Forest (RF) algorithm. The modeling used environmental variables such as estimated precipitation, surface temperature, topographic factors and biophysical indices. Feature Importance was used to identify the relevance of the 153 covariates. The results show an average overall accuracy of 0.83 and a Kappa index of 0.80, producing a mapping of “very good” quality. The spatialization of forest vegetation classes and water bodies was largely consistent. Imprecisions were detected between Urban Areas and Other Uses, and between Savannah Vegetation, Agriculture and Pasture. The most important variables were spatial position, altitude and climate. The NIR and SWIR bands showed a strong influence, as did the NDTI vegetation index, transformed bands and fraction images. However, the NDVI, GNDVI, EVI and SAVI indices were of low relevance. This research explored advanced satellite data processing and analysis techniques in order to obtain a classification model with good accuracy and high reliability for spatializing the biophysical heterogeneity of the Caatinga. The approach was successful, resulting in positive statistical evaluations and satisfactory spatialization of the 12 grouped classes.

COMMITTEE MEMBERS:
Externa à Instituição - BETÂNIA QUEIROZ DA SILVA
Externo à Instituição - GETULIO FONSECA DOMINGUES - UFMG
Interno - 1818226 - JOAO SANTIAGO REIS
Presidente - 1726169 - SARA FERNANDES FLOR DE SOUZA
Externo à Instituição - WASHINGTON DE JESUS SANT'ANNA DA FRANCA ROCHA