AMORIM, V. P.; http://lattes.cnpq.br/9246169297596139; AMORIM, Victória Pereira.
Resumo:
This work was developed from data from a field experiment with tomatoes carried out at the Experimental Agrometeorological Station of the Department of Atmospheric Sciences (DCA) of the Federal University of Campina Grande (UFCG), in the municipality of Campina Grande - Paraíba, Northeastern region of Brazil. The experiment was carried out in the years 2016 and 2017, with the planting of two tomato cycles and five different irrigation treatments (100, 80, 60, 40 and 20% of water replenishment) were used. The present work aimed to develop empirical equations from the parameters to estimate the actual evapotranspiration of the tomato crop, using the dimensional analysis tool, based on the Buckingham π theorem. The variables of real evapotranspiration, precipitation, soil moisture, wind at 2 meters, irrigation, reference evapotranspiration, solar radiation, height and plant diameter were used, and from that, the dimensionless parameters were developed, where were considered the basic dimensions of mass [M], length [L], and time [T]. Then, three dimensionless models were formulated. The statistical analyzes used were the regression (r2), to evaluate the results, and for the validation of the three models the root mean square error (RMSE), root mean square root normalized error (NRMSE), efficiency coefficient of the Nash-Sutcliffe model (EF) and the Willmott index (d). The results presented a satisfactory performance for the dimensionless model that involves the variables of irrigation, height, plant diameter and ETo. In addition, the dimensionless model with only the phenometric variables (height and diameter) presented the worst performance for all water treatments. Thus, it was concluded that the dimensional analysis is a satisfactory tool for the estimation of crop evapotranspiration when cultivated without water restriction, but with controlled irrigation.