J. Laksanaboonsong

Determination of Angstrom’s turbidity coefficient over Thailand

Values of the Angstrom’s turbidity coefficient, β, at 53 meteorological stations covering Thailand were determined by using three different methods. A selection of the methods was based on input data available at each station. It was started with the calculation of β at Nakhon Pathom (13.81 °N and 100.4 °E) using narrow-band spectral irradiance data obtained from a multi-filter rotating shadow band radiometer. Langley’s method was employed to calculate β from the spectral data. The values of β derived from this method were used as references to validate a method for computing β from broad-band direct irradiance proposed by Louche et al. (Solar Energy 38(2)89). It was found that this method was valid for a tropical climate. Then Louche et al.’s method was used to calculate β at meteorological stations situated at four main cities, namely Chiang Mai (18.78 °N, 98.98 °E) located in the north, Ubon Rachathani (15.25 °N, 104.87 °E) in the northeast, Songkhla (7.20 °N, 100.60 °E) in the south and Bangkok in the central region. Based on values of β of these cities, a new model relating β to visibility, suitable for the tropical climate was developed. This model was used to estimate β at the other 48 meteorological stations where the visibility was routinely observed. Finally, seasonal variations of β were investigated. It was found that for the stations in the north, the northeast and the central region, the values of β are relatively high in the dry season (November–April). They decrease in the wet season (May–October). For most stations in the south, β was relatively low and remained nearly constant all year round. It was also inferred that the northeast monsoon and the southwest monsoon had a strong influence on the seasonal variations of β.

A model for calculating global illuminance from satellite data

A model for calculating global illuminance on horizontal surfaces from meteorological satellite data was developed. The data used for developing the model are global illuminance measured at four solar monitoring stations situated in different parts of Thailand and 8-bit digital data from visible channel of GMS-5 satellite covering the whole country for the period of 1–2 years. Values of normalized global illuminance defined as a ratio of global illuminance to clear sky global illuminance were calculated. These values were used to correlate with those of cloud index derived from the satellite data. From the correlation, a model relating the normalized global illuminance to cloud index was established. The performance of this model was investigated using an independent illuminance data set. It was found that the global illuminance calculated from the model agreed well with that obtained from the measurement, with a root mean square difference of 5.38 klux or 7.0% of the mean values.