Banca de QUALIFICAÇÃO: ALIBIA DEYSI GUEDES DA SILVA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : ALIBIA DEYSI GUEDES DA SILVA
DATE: 26/03/2024
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:

Caatinga; Random Forest; Google Earth Engine; El Niño; La Niña

PAGES: 86
BIG AREA: Ciências Humanas
AREA: Geografia
SUMMARY:

Land use and coverage represent the materialization of socio-environmental manifestations that raise concerns regarding future scenarios in the current global situation. According to the Intergovernmental Panel on Climate Change, the semi-arid regions of the Brazilian Northeast could be affected by an increase in droughts that will result in socio-environmental disaster events. The environmental and socioeconomic interconnection that is present in the United Nations 2030 Agenda through global action, enacts the protection of the environment based on the Sustainable Development Goals to guarantee quality of life on Earth. However, changes in the landscape due to the activities of societies that replace native forests have caused significant impacts on natural areas. The indiscriminate use of natural cover causes consequences for soil fertility and biodiversity in different degrees of environmental degradation, as well as food and water insecurity. Therefore, mapping land use and cover becomes essential for understanding and applying  it  in  territorial  planning  in  the  semi-arid  region,  since  multi-temporal  changes in caatinga vegetation in semi-arid environments are directly associated with rainfall variability and human action. Thus, the research aims to map the use and coverage of land in the semi-arid region of Rio Grande do Norte for the year 2023 using machine learning and cloud computing algorithms, as well as analyzing the spatio-temporal behavior of vegetation cover in years of occurrence of the El Niño and La Niña phenomena between 1990 and 2023. Due to the different spatial typologies, the study area was subdivided into hydrographic microregions. The Google Earth Engine tool in JavaScript language was used to process orbital data from LANDSAT-5 and LANDSAT-8. The definition of the periods to be analyzed was based on the action of the ENSO phenomenon and the volume of precipitation, collected in CHIRPS (Clima Hazards Group InfraRed Precipitation with Station data), resulting in the years 1990, 1998, 2000, 2014, 2016, 2018, and 2023. The spectral indices NDVI (Normalized Difference Vegetation Index), EVI (Enhanced Vegetation Index), SAVI (Soil-Adjusted Vegetation Index), NDWI (Normalized difference water index), and NDBI (Normalized Density Building Index) were calculated to monitor the conditions of vegetated areas, extracting the metrics. The Random Forest model will be applied for the pixel-based supervised classification. The mapping must identify the classes: 1) dense vegetation, 2) open vegetation, 3) mountain forest, 4) agriculture, 5) irrigated agriculture, 6) pasture, 7) urban area, 8) uncovered area, 9) continental water, and 10 ) coastal water.

COMMITTEE MEMBERS:
Externo à Instituição - WASHINGTON DE JESUS SANT'ANNA DA FRANCA ROCHA
Externo ao Programa - ***.339.107-** - GETULIO FONSECA DOMINGUES - UFMG
Interno - 1818226 - JOAO SANTIAGO REIS
Presidente - 1726169 - SARA FERNANDES FLOR DE SOUZA