Estimating Solar Radiation Using Sunshine Duration.

Solar radiation is an important input variable of crop models. Still, observation data of solar radiation is usually available at a smaller number of weather stations than those for sunshine duration as well as temperature and precipitation. Different approaches have been used to estimate solar radiation using more commonly available variables, e.g., sunshine duration. For example, Angstrom-Prescott model has been used to estimate monthly and daily solar radiation using sunshine duration. Although empirical parameters of Angstrom-Prescott model can be determined using a long-term observation data of solar radiation, it would be challenging to obtain such parameters at a specific site. The objective of this study was to examine if the reliable estimation of solar radiation could be obtained using the Angstrom-Prescott model with a single set of empirical parameters in a large area within a small range of climate conditions, e.g., a temperate climate condition. Long-term observation data of daily solar radiation and sunshine duration were obtained from Korean Meteorology Administration. A simple data client, WiseDownloader, was used to automate the retrieval of daily weather data for 30 years at 22 sites. The statistics of the degree of agreement including concordance correlation coefficient (CCC) were calculated by month, site, and year. Using the parameter values suggested by Frere et al., reasonable estimates of solar radiation were obtained in Korea. For example, the values of CCC ranged from 0.87 to 0.92 by months. The values of CCC by site and years ranged from 0.91 to 0.96 and from 0.88 to 0.95, respectively. Although a single set of parameter value resulted in a reasonable estimation of solar radiation, it appeared that another set of empirical parameters for Angstrom-Prescott model would be useful when solar radiation was low during winter or under cloudy conditions.