Sane-Mfenyana/SolarFarm-Weather-Impact-SA

Data analysis of how weather variability impacts solar farm performance in South Africa, using long-term irradiance baselines and hourly weather data for actionable insights. # Impact of Weather Variability on Renewable Energy Output in South Africa A comprehensive data analysis project investigating how weather factors drive the performance of solar (PV and CSP) and wind farms in South Africa. Using statistical analysis and machine learning regression models, this project quantifies key relationships to provide actionable insights for grid operators, investors, and energy planners. ## 📊 Project Overview ### 🎯 The Mission To determine the primary weather drivers of solar and wind energy generation and to evaluate the accuracy of weather-based forecasting models. The goal is to translate meteorological data into strategic insights for operational planning, risk assessment, and investment decisions. ### 🔑 Key Questions - How do temperature, solar irradiance, and cloud cover correlate with energy output from PV and CSP technologies? - How accurate are weather-based forecasts for predicting renewable generation? - How does forecast accuracy and energy output vary by season and time of day? - What are the practical implications for grid stability and financial planning? ### 📁 Data Sources - **Open-Meteo Historical API**: Hourly weather data (temperature, solar radiation, cloud cover, wind speed). - **Eskom**: Actual historical energy generation data for PV, CSP, and Wind. ### 🛠️ Tools & Technologies - **Google BigQuery**: Data storage, cleaning, and analysis using SQL. - **BigQuery ML**: Building and evaluating linear regression models. - **Tableau**: Data visualization and dashboard creation. --- ## 1. Laying the Foundation: Data Acquisition & Preparation The foundation of any robust analysis is clean, reliable data. We started by ensuring our dataset was accurate and ready for analysis. - **Data Validation**: Verified the integrity of the hourly weather dataset, ensuring full annual coverage (8,778 records) and consistent formatting. - **Schema Standardization**: Renamed columns in BigQuery for clarity and compatibility (e.g., `te …