Lisímetro de pesagem hidráulica e evapotranspiração de referência por diferentes métodos no agreste pernambucano

Abstract

The aim of this work was to build and test the functioning of an hydraulic weighing lysimeter with a mercury piston arranjement, and to relate the reference evapotranspiration (ETo) determined by the lysimeter with that estimated by agrometeorological models and by evaporimetric pans. The reseach was conducted at the Nossa Senhora do Rosário Farm, in Pesqueira, Pernambuco State, Brazil, where the lysimeter is located under the coordinates 08°10’25” South latitude and 35°11’25” west latitude and at an altitude of 613 m. The local climate is of the BS’s’h type acording to Koppen. The meteorological elements data and the evaporation were obtained from an automatic agrometeorological station (Campbell Scientific ET106), from a Class A pan and from an experimental pan. The following models were used: Blaney-Criddle, Camargo, Hargreaves-Samani, Jensen-Haise, Makkink, Penmam-1963, Penmam-FAO, Penmam-Monteith, Radiação-FAO e Thorntwaite-Camargo. The lysimeter tank had a volume capacity of 2.059 m3 and a surface area of 2.068 m2, planted with pangola grass (Digitaria decumbens Stent.). The preliminary calibrations tests, using up to 50 mm water depth, with progressive witshowed excellent responses throughout the testing. The range ysimeter operated with a final field precision of 0.143 mm, corresponding to a volume of 14 dm3. The lysimeter estability was high, where the greatest positional average error measured at 3.93%. The lysimeter’s performance acording to Willmott’s index, using Penman-Monteith method as a standard, was “Very bad” for measurements on a daily scale and “Bad” with tridial dates, reveling a strong local advective effect upon evapotranspiration. The same effect was evidenced upon the readings of the evaporation pans, which also showed unsactisfactory performance in relationship to the standard method.