Classificadores de Imagens por Técnicas de Sensoriamento Remoto e Geoestatística

Abstract

I II I S Si im mp pó ós si io o d de e G Ge eo oe es st ta at tí ís st ti ic ca a A Ap pl li ic ca ad da a e em m C Ci iê ên nc ci ia as s A Ag gr rá ár ri ia as s 1 19 9 e e 2 20 0 d de e m ma ai io o d de e 2 20 01 11 1 B Bo ot tu uc ca at tu u--

November 2023

“…The calibration of parameters "a" and "b" of the algorithm is fundamental for local estimation, as it affects the inclination of the response surface of the NDVI variables, surface temperature and albedo in the estimation of ETr/ETo. Estimates with parameters 1 and -0.008 ("a" and "b") respectively, point to a maximum ETr/ETo of 0.55, a value much lower than the expected for the crop, therefore, an underestimate (Feitosa et al, 2016), while 1.8 and -0.008 the values are close to the expected maximum and in some cases inferring a higher coefficient for phases of higher demand (Silva et al, 2019).…”

“…The algorithm presented an adjustment for the evapotranspiration estimation for both rainy and dry periods, as long as the plant is established (Teixeira et al, 2016). While the impact of the estimation of the variable "surface temperature" has a relation to the pixel size, causing disuniformity of the estimates when considered long cycle and passive bedding cultures, or due to radiometric difference of the imaging sensors (Silva et al, 2019). The calculated statistical indices demonstrate the temporal dispersion of ETr/ETo and ETr estimates compared to the FAO standard of culture coefficient and potential evapotranspiration by Penman-Monteith (Table 1).…”

Identificação de padrões de uso do solo urbano em São Paulo/SP utilizando parâmetros de variogramas.

“…A geoestatística considera os números digitais de uma imagem digital como variável regionalizada que apresenta características como aleatoriedade e correlação espacial (Silva et al 2011). …”