Evaluation of the SHAW Model in Simulating the Components of Net All-Wave Radiation

Abstract

Radiation exchange at the surface plays a critical role in the surface energy balance, plant microclimate, and plant growth. The ability to simulate the surface energy balance and the microclimate within the plant canopy is contingent upon accurate simulation of the surface radiation exchange. A validation exercise was conducted of the Simultaneous Heat and Water (SHAW) model for simulating the surface radiation exchange (including downward long-wave and upward short- and long-wave radiation) over a maize canopy surface using data collected at Yucheng in the North China Plain. The model simulated upward short-wave and net all-wave radiation well with model efficiencies (ME) equaling 0.97 and 0.98, respectively. Downward and upward long-wave radiation were overestimated by 12.1 and 8.3 W m-2 with ME equaling 0.68 and 0.89, respectively. Two modifications to the model were implemented and tested to improve the simulated long-wave radiation exchange. In one modification, alternative schemes were tested to simulate cloudy sky long-wave radiation, and the best algorithm was employed in the model. With this modification, both downward and upward long-wave radiation were simulated better, with ME rising to 0.88 and 0.91, respectively. A second modification was implemented to use leaf temperature rather than canopy air temperature to compute emitted long-wave radiation. Although more theoretically correct, this modification did not improve simulations compared to the original model because upward long-wave radiation was already overpredicted and midday leaf temperatures at this site were typically higher than canopy air temperatures. Thus, the modification resulted in even higher overprediction of upward midday long-wave radiation. However, this modification removed some of the bias in nighttime emitted long-wave radiation. While the SHAW model simulates the radiation balance and transfer processes within the canopy reasonably well, results point to areas for model improvement.