Akademik Veri Yönetim Sistemi
THEORETICAL AND APPLIED CLIMATOLOGY, cilt.155, sa.7, ss.5727-5743, 2024 (SCI-Expanded)
Sunshine duration (SD) is one of the critical meteorological parameters used in different fields of application such as climate, renewable energy and agriculture. In this respect, determination and/or estimation of the temporal and spatial variability of SD is critical. Meteorological satellite data/products can be used for estimating SD and in constructing their maps due to their frequent observation of large areas at once. In this study, a multilayer perceptron type artificial neural network model was built to estimate the monthly mean SD for Turkiye using the EUMETSAT CM SAF (Satellite Application Facility on Climate Monitoring) CFC (Cloud Fractional Coverage) and CTY (Cloud Type) data, GMTED2010 (Global Multi-resolution Terrain Elevation Data) data, month number and daylength. The datasets of 45 stations, spanning nine years (2005-2013), were used for training the model and 12 stations for testing and validating the simulated values. We have compared the results of our model with the ground-measured values for the whole period under consideration and the root mean square error (RMSE), mean absolute error (MAE), mean bias error (MBE) and the coefficient of determination (R2) were found as 0.7803 h, 0.6206 h, 0.1751 h and 0.9387, respectively. It has been shown that using the new generation cloud products such as CFC and CTY, elevation data such as GMTED2010 and daylength, it is possible to predict the SD for regions under the coverage of the satellite, in case no measurement is possible or may be unreliable, without needing any measured meteorological data.