Advancing Solar Energy Planning Using Machine Learning and GIS: A Suitability Analysis in Nineveh Governorate Iraq

Solar energy is growing rapidly due to its abundance, availability, cleanliness, cost-effectiveness, and ease of installation. The site selection process plays a crucial role in maximizing the efficiency of solar projects while minimizing the environmental impact. This study focuses on Nineveh Governorate, Iraq, by introducing an advanced solar project site selection methodology that utilizes the Geographic Information System (GIS) and Machine Learning (ML). For this aim, seven machine learning (ML) algorithms, namely Logistic Regression, Averaged Perceptron, Boosted Decision Tree, Decision Forest, Support Vector Machine, Neural Network, and K-Means, are applied on Microsoft Azure ML to identify a suitable site for solar farms. Thirteen factors are identified that influence the solar farm site, including solar radiation, sandstorms, land surface temperature, main road, population, power lines, substation, land ownership, land cover, water resources, slope, and aspect. These factors are categorized into four groups: environmental, climatic, topological, and socio-economic. Moreover, the dataset is analyzed in Azure Machine Learning using unsupervised and supervised ML. As a result, the Boosted Decision Tree model has achieved the highest accuracy at 94.2%. The output results indicate that 27% of the study area is highly suitable for solar farm development. This research underlines the immense potential of the Nineveh Governorate for renewable energy projects. It creates a replicable framework linking GIS and ML to planning energy needs with minimal environmental impacts.