Prabha, T, Hoogenboom, G (2010). Evaluation of solar irradiance at the surface-inferences from in situ and satellite observations and a mesoscale model. THEORETICAL AND APPLIED CLIMATOLOGY, 102(4-Mar), 455-469.
Accurate information about the solar irradiance at the soil surface is essential for many agricultural, hydrological and environmental models that take into account the surface energy balance. The main goal of present study was to evaluate the solar irradiance predictions from the Advanced Research Weather Research and Forecasting (ARW) model for both clear sky and cloudy conditions. An extended observational dataset from the Georgia Automated Environmental Monitoring Network (AEMN) provided hourly solar irradiance at the surface and other collocated surface level measurements. The radiation bias (determined from the difference between the ARW predictions and AEMN observations) showed a linear relationship with the cloud optical depth and the cirrus cloud amount from the moderate resolution imaging spectroradiometer (MODIS). For cloud-free days, the ARW model had a positive radiation bias that exceeded 120 W m(-2) over coastal and urban areas of Georgia. The model radiation and air temperature bias increased with increasing aerosol optical depth derived from the MODIS observations during the cloud-free days, attributed to fire events that lasted intermittently throughout the study period. The model biases of temperature, mixing ratio, wind speed, and soil moisture were linearly dependent on the radiation bias.