Estimation of sensible heat flux at a tropical location: a performance evaluation of half-order time derivative method

Abstract

Measurement of surface energy balance (SEB) components such as sensible heat flux is a very fundamental input in weather forecast, hydrological, and air quality models among many other applications. The instrumentation for measuring SEB components is however resource-intensive and highly susceptible to damage under field-experiment conditions. A simple, less equipment-intensive, cost-effective, and relatively accurate method is thus required to overcome these setbacks. The present study, therefore evaluates the performance of a rather simple half-order time derivative (HTD) method in estimating sensible heat flux at a tropical location in West Africa (Ile-Ife, Nigeria 7°33′N, 4°33′E). HTD estimates of sensible heat flux were made from near surface atmospheric parameters measured during Phase I of the Nigerian Micrometeorological Experiment. At the same site, direct turbulence flux measurements from an eddy covariance system comprising a 3-D ultrasonic anemometer (USA-1) and a krypton hygrometer (KH20) were used as standard benchmark values for the HTD performance evaluation. Within a diurnal air temperature range of 18–34 °C, estimated daytime sensible heat flux reached a peak of 318.7 W m−2 and a lowest nighttime value of − 15.8 W m−2, indicating surface cooling. Statistical tests performed; mean bias error (MBE <2 W m−2), root mean square error (RMSE <7.3 W m−2), mean absolute error (MAE< 2 W m−2), and percent mean relative error (PMRE <1%) indicated a good accuracy of the HTD method with an overall significant correlation (R2 = 0.89) with benchmark values. The HTD method is found most suitable for daytime convective conditions than stable nighttime periods at the tropical location.