Application of an improved vegetation index based on the visible spectrum in the diagnosis of degraded pastures: Implications for development

• Autor(a) principal: da Silva Quinaia, Thiago Luiz
• Data de Publicação: 2021
• Outros Autores: do Valle Junior, Renato Farias [UNESP], de Miranda Coelho, Victor Peçanha, da Cunha, Rafael Carvalho, Valera, Carlos Alberto [UNESP], Sanches Fernandes, Luís Filipe [UNESP], Pacheco, Fernando António Leal [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1002/ldr.4071; http://hdl.handle.net/11449/222388
• topic: Google Earth engine; NDVI; pasture degradation; remote sensing; total brightness quotient; unmanned aerial vehicles

Inadequate pasture management causes land degradation through reduction of grass, increased presence of invasive plants or pests, compaction, erosion, and nutrient deficiency. The recognition of pasture degradation is therefore essential. Remote sensing satellite systems allow us to do so at regional-to global scales. A struggle is in progress nowadays is to improve detection accuracy and implement high-resolution surveys at farm scales using low-cost unmanned aerial vehicles (UAVs). The pasture imagery can be translated into maps of degraded pasture using the popular NDVI as diagnostic parameter, but their generation using a UAV requires a high-cost NIR sensor, while the struggle is to use low-cost UAVs equipped with RGB cameras. The first step to recognize degraded pastures using RGB cameras is to define a suitable vegetation index. Thus, the purpose of this study was to present the total brightness quotient of red (TBQR), green (TBQG), and blue (TBQB) bands. The test to the index resorted to LANDSAT-8 satellite images captured over the environmental protection area of Uberaba River basin (Minas Gerais, Brazil) in the 2017–2019 period. The images were not captured by a UAV because the equipment was not then available. The results were promising given the large detection accuracy (88.63%) of the TBQG and the high (0.965) correlation between TBQG and NDVI. Besides, the TBQ-based areas of degraded pasture (17,486.3–25,180.1 hectares) were larger than the NDVI counterparts (12,066.9 hectares). This is an additional reason to oversight degraded pastures based on the TBQs, as they seek for improved environmental compliance and economic development.