Prediction of the effect of climate covariates on wind potential in Ethiopia

Abstract

This study explores the interaction of climate covariates and spatial elements with wind speed in Ethiopia. It intends to extrapolate the potential spots of wind at unobserved spatial points using a meteorological dataset. We applied a combined dynamic spatial panel autoregressive random effects model with a spatial weight of inverse quartile separation distances of locations. This spatial weight outperforms the other spatial weights to capture spatial dependence and gain efficient estimates. The result describes that mean wind speed varies over the longitude range and latitude span, is influenced by climate covariates and fluctuates over the months of a year. Wind speed intensity is high along the central, eastern and northeastern parts of the region. It is also high in February, March, June, and July relative to September and October months. The evidence shows that wind speed is higher in summer and spring but relatively lower in winter and fall seasons. This implies that wind speed is high mainly after the rainy season ends and before it starts. The model estimates also show that mean wind speed is spatially correlated across neighboring stations and over temporal points. Particularly, the mean wind speed increases with altitude and temperature but decreases as precipitation increases. Sunshine fraction and relative humidity have negative effects, but their influence is not statistically significant with p=0.2496 and p=0.4484 respectively. In conclusion, the methods are recommended for the prediction of data that exhibits a stochastic process.