COMPARATIVE EVALUATION AND DEVELOPMENT OF PAN EVAPORATION MODELS FOR THE DESIGN OF HYDRAULIC STRUCTURE

Abstract

Pan evaporation measurements have been used worldwide as a means of estimating evapotranspiration and free water surface evaporation. It is of great importance for water resource planning and management, and in the design of hydraulic structures. The study was aimed at carrying out comparative evaluation and development of pan evaporation models useful in hydraulic engineering design. Thirty-six year records (1973-2008) for Ibadan; twenty-three year records (1985-2008) for Onne; and nine year records (2000-2009) for Kano of pan evaporation (Epan), solar radiation (SR), windspeed (Ws), temperature (T), were obtained from the database of the International Institute for Tropical Agriculture (lIT A) stations situated in Ibadan, Onne and Kano of Nigeria. Mann-Kendall trend analysis and non-parametric sen's slope estimate were performed on respective meteorological variables to detect monotonic trend direction and magnitude of change over time. The result of Mann-Kendall trend analysis for Ibadan showed that pan evaporation (Epan) and solar radiation (SR) monotonically decrease significantly in all the months (P > 0.01). During one year period in monthly temperature (T) showed positive trends for 7 months (in an undefined order) and the changes were not statistically significant. Windspeed (Ws) decreased in 12months of the year with only 5 months having trends with significant slopes (P < 0.05). The result for Onne, showed that pan evaporation monotonically increased in 11 months with 0,\)5 and 0.01 level of significance in 5 months. Monthly windspeed showed positive trends for 7 months, all changes were not statistically significant. Temperature (T) increases in 11 months with only 3 months having trends with significant slope (P > 0.05), (P < 0.1) and (P < 0.01). The result for Kana, showed that pan evaporation (Epan) increased in 8 months with 0.1 level of significant change in December. Solar radiation (SR) showed positive trend in 8 months which is not statistically significant. Windspeed (Ws) indicated negative trend for 6 months with significant change at 0.1 level in August. Temperature (T) showed increase in 9 months having , 1 level of significance in August. The performance evaluation of commonly used pan evaporation estimation. models were compared for the three regions. The models were Jensen, which depends on solar radiation and temperature as the parameters input; Kohler, which utilizes solar radiation, temperature and xiii windspeed input; and Penman, which depends on temperature, dew-point temperature.windspeed and slope of phychrometric curve as the input nara+ ters. The two non-linear regression models (Jensen-type and Kohler-type) performed b st under Nigeria climatic conditions. The an evaporation values estimated from the non-linear ITl' oeis were compared to the values obtained from the three IITA stations using five standard st?+: .cal parameters. -.) Results of the comparison showed :. ~ best model that provide the closest estimate of pan evaporation relative to the sourced data from the IITA was the Kohler model, followed by Jensen model and that Penman model was not a reliable estimator in Nigeria. Additionally, a number of multilinear regression equations were developed to establish the relationship between pan evaporation with one or more combinations of meteorological parameters. The value of correlation coefficient (R),coefficient of determination (R2) and the value of the standard error of estimate (SEE) were determined for each equation. The equation with the highest value of Rand R2 and least value of SEE is given as: Ep= O.267(Tmax) + O.126(SR) - O.036(Rhmax) + O.004(Ws) - 3.288 Application of these models for estimating inputs for the design of hydraulic structures presented useful case studies.