SPATIAL DOWNSCALING OF DAILY PRECIPITATION d4PDF DATA USING ANN ALGORITHM

This study evaluates the applicability of an artificial neural network (ANN)-based regression model for downscaling d4PDF daily precipitation data from a coarse spatial resolution of 20 km to a finer spatial resolution of 5 km, considering the precipitation as the sole predictor. For this purpose, two sets of d4PDF data have been utilized for downscaling: 20 km to 5 km resolution. The area covered was a square grid of 33 at 20 km resolution, and the downscaling target was the center grid of this square. The results of this study demonstrated that the performance of ANN varies across regions and seasons. While it shows better performance in capturing the long-term mean during winter, it tends to exhibit high biases in the mean, standard deviation (SD), and maximum values, particularly during the summer seasons. Generally, the ANN does not preserve the overall statistical properties of the original data.