Time Dependent Classification of Solar Radiation Sequences Using Best Information Criterion

Large and frequent variations of solar radiation can be observed in tropical climates with amplitudes reaching 800W/m2 and occurring within a short time interval, from few seconds to few minutes, according to the geographical location. Such fluctuations can be due for example to the dynamic of clouds which can be very complex and depend on cloud type, size, speed and spatial distribution and, more generally, due to some specific local meteorological conditions. In this work, we have led a statistical and dynamical analysis of time dependent global solar radiation sequences in Guadeloupe (F.W.I.) located in the East part of the lesser Antilles. It's the first land to receive the transported clouds by the trade winds, there is also cloud generation by convection process. This process is strongly time dependent. Consequently there is direct relation between this process and fluctuations on ground measured solar radiation signals. In this analysis we have used solar radiation data measured at 1 hz. We have classified time conditioned probability density functions (pdf) the clearness index kt. to ensure that the experimental probability density function (pdf) is well defined we have use the best information criterion to find the appropriate number of bins for each pdf. This classification allowed to identify different regimes of solar radiation at specific hours of the day.