Métodos para determinar a evapotranspiração real da cana-de-açúcar

Abstract

An essential factor for sugarcane development is the optimal water distribution along its cycle. In rainfed crop there is a need of water demand knowledge by actual evapotranspiration (ETR) in various development crop stages, for better planning of agricultural operations (tillage, fertilization, planting, weeds, diseases, pests, harvest and/or irrigation) to increase productivity. When no measure ETR this is calculated by various methods which varies according with physical soil, physiology crop and atmospheric ambient conditions. Thus the objective was to evaluate the actual evapotranspiration in sugarcane crop by five methods and evaluate soil water storage (ARM) by three methods. This study was performed based on field research data at Centro de CiênciasAgrárias at Universidade Federal de Alagoas in Rio Largo-Alagoas region from September 2005 to November 2006. A RB92579 sugarcane variety was used. In ETR estimation in quinquidial time scale (five days) made use of: i) Penman-Monteith FAO “dual Kc” method (ETR 1); ii) available water method in its original form (ETR 2,a – NULLET; GIAMBELLUCA, 1988) and its modifications using root growth by crop coefficient and leaf area index (ETR 2,c and ETR 2,c); iii) crop water balance (ETR 3 – THORNTHWAITE; MATHER, 1955); iv) Penman-Monteith FAO “single Kc” method with reduction coefficient (Ks = ETR/ETc) derived from the available water method proposed by NulletandGiambelluca (ETR 4,a) and crop water balance proposed by Thonthwaiteand Mather (ETR 4,b); v) Penman-Monteith with canopy surface resistance (ETR 5). We Calculated the ARM based methodologies ETR 1; ETR 2, a; ETR 2,b; ETR 2,c and ETR 3 classified then as ARM 1; ARM 2,a; ARM 2,b; ARM 2,c and ARM 3. Initially this methods were compared with ARM 1 and during rainy season were compared all methods with real ARM measured in situ with TDR (Time Domain Reflectometry). All results were statistically compared through Willmont index agreement (d), Pearson correlation coefficient (r), percentage deviation (DP) and root mean squared error (RMSE). The ETR 2,a and ETR 4,a methods showed better statistical indexes when compared with ETR 1, followed by ETR 3 and ETR 2,c. The ETR 5 method showed better adjustment during the rainy season with correlation coefficient (r) of 0,95. It is observed the ETR ranged from 754 mm (ETR 2,c) to 1052 mm (ETR 3). The ARM 2,b; ARM 2,c and ARM3* showed better adjustment compared to ARM 1. When comparing the ARM methods with real ARM, the best statistical indexes (d = 0,99 and RMSE = 5,57 mm) for ARM 2,c.