Abstract:
Four evapotranspiration models (Mc Guiness-Bordne, Jensen-Haise, Hargreaves-Samani and Romanenko) have been estimated and tested against reference evapotranspiration data computed by the FAO-Penman-Monteith model to assess the accuracy of each model in estimating reference evapotranspiration (ETo) in Nigeria. Models were compared at fourteen synoptic stations of Nigeria using meteorological data from 1975 to 1994. Comparisons were also made using four climatic regions: Coastal, Guinea Savannah, Midland and Sahel. The results show that in the Sahel and Midland regions, the Mc Guiness-Bordne model was the least accurate daily model with a R2 of 0.02. Jensen-Haise and Hargreaves-Samani performed best with a R2 of 0.88 and 0.93 respectively. Though, the Romanenko model also has good coefficient of Determination (R2 = 0.88) but its behavior is limited by negative intercept (Intercept = -4.06). These results will assist agricultural and environmental planners in assessing the available water resources in Nigeria. The performance of the original and recalibrated evapotranspiration models and Penmann Monteith models across Nigeria in terms of Standard Error of Estimate (SEE) and Coefficient of Determination (R2) values, the JENSEN HAISE’S and HARGREAVE SAMANI models showed the lowest SEE values at 0.22 and 0.15 mm/day with the values of R2 being 0.92 and 0.93 for the original Jensen Haise and Hargreave Samani evapotranspiration respectively and for the recalibrated Jensen Haise, we have the SEE values alongside the R2 as 0.24 and 0.88 respectively. In examining ETo, the trends in the model performance are fairly clear. HARGREAVE SAMANI is almost always underestimating FAO-PM ETo, although it follows the annual trends fairly well.
The most suitable among these is that of the Jensen-Haise alongside that of the Hargreave-Samani. Therefore; owing to the large data input requirement of the reference PM ET, it becomes worthy
to recommend Jensen-Haise and Hargreave-Samani for the calculation of evapotranspiration across any region in Nigeria in place of the reference PM ET.
