Using the particle swarm optimization algorithm to calibrate the parameters relating to the turbulent flux in the surface layer in the source region of the Yellow River

- Particle swarm optimization algorithm was introduced to the MOST for calibrating the parameters.
- The differences between calculations used the calibrated parameters and observations were rather small.
- PSO algorithm is a stable and efficient approach which can be applied in MOST parameter estimation.

Accurately determining the fluxes of mass and energy between land and the atmosphere is important for understanding regional climates and hydrological cycles. In numerical modeling, the parameterization of a turbulent flux is usually based on Monin-Obukhov similarity theory (MOST). According to this theory, it is necessary to simultaneously calculate the empirical similarity parameters βm, βh, γm, and γh, the aerodynamic roughness (z0m) and the thermal roughness (zT). However, it is difficult to solve a simultaneous set of nonlinear equations for these six parameters. In this study, a new method was introduced to solving this problem. Using measurements from Maqu Station in the source region of the Yellow River, this study employed the artificial intelligence particle swarm optimization (PSO) algorithm to calibrate the parameters relating to the turbulent flux in the surface layer. We concluded that the differences in the sensible heat and momentum fluxes between the calculations that used the calibrated parameters and the measurements were rather small and that their correlation coefficients were relatively high. The results suggested that PSO algorithm is a feasible approach which can be applied in MOST parameter estimation.