Estimativas da radiação global incidente em superfícies inclinadas com base na razáo de insolação


Autoria(s): De Souza, Adilson P.; Escobedo, João Francisco
Contribuinte(s)

Universidade Estadual Paulista (UNESP)

Data(s)

27/05/2014

27/05/2014

07/10/2013

Resumo

Estimating equations of global radiation based on the sunshine duration were proposed for horizontal surface and with inclination of 12.85, 22.85 and 32.85° facing the North in Botucatu, SP, Brazil, in monthly, seasonal and annual groupings of data. Simple linear correlations were applied (for definition of the linear and angular coefficients of Angstrom-Prescott model), in a database measured in all three inclinations in different periods (22.85°: 04/1998 to 07/2001; 12.85°: 08/2011 to 02/2003; and 32.85°: 03/2003 to 12/2007) concomitant with horizontal measures and sunshine duration. The statistical performance of the model was analysed by the means absolute error (MBE), the square root of the mean square error (RMSE) and the index adjustment (d). The minimum global radiation transmissivity varied from 14.35% in August (12.85°) to 27.86% in December (32.85°) and the maximum transmissivity ranged between 62.10% and 78.90%, for June (32.85°) and December (12.85°). Increasing the angle of inclination surface increased the scattering and decreased the index of adjustment and performance. The worst results were found for application of the seasonal and annual models in the months of autumn and winter for 32.85° (RMSE below 42.93% and adjustment superior to 0.4693).

Formato

483-491

Identificador

http://dx.doi.org/10.5039/agraria.v8i3a1894

Revista Brasileirade Ciencias Agrarias, v. 8, n. 3, p. 483-491, 2013.

1981-1160

1981-0997

http://hdl.handle.net/11449/76800

10.5039/agraria.v8i3a1894

2-s2.0-84884837630

2-s2.0-84884837630.pdf

Idioma(s)

por

Relação

Revista Brasileirade Ciencias Agrarias

Direitos

openAccess

Palavras-Chave #Angströn-Prescott model #Atmospheric transmissivity #Photoperiod #Statistical indicators
Tipo

info:eu-repo/semantics/article