Drilling lost circulation is an important operational risk, which comprises around 1020 of overall costs of well construction, and does not yet have sufficient operational forecasting with any methodology. The study will present an explainable artificial intelligence model involving the use of Gaussian Process Regression (GPR) and L-BFGS optimization to estimate the rate of drilling fluid loss in real-time in Nigerian vuggy carbonate formations. A combination of physics-based and data-driven workflow combines analytical modelling of mud loss with machine learning based classification.......

KEYWORDS: Lost circulation; explainable AI; Gaussian Process Regression; drilling fluid loss; real-time prediction; Nigerian reservoirs.

Accurate short term photovoltaic power forecasts support grid operation and market decisions because PV output responds rapidly to irradiance and weather variability. This study evaluates machine learning models for one step ahead PV plant power prediction using operational measurements from the Ma'an solar power plant in Ma'an, Jordan. The workflow integrates synchronized electrical monitoring and onsite meteorological measurements recorded at a nominal 15-minute interval from September 1, 2025, to November 2, 2025. After time alignment and quality screening, the merged dataset includes about 5948 records, and listwise deletion yields about 5860 clean records........

KEYWORDS: PV power forecasting, 15-minute ahead, machine learning, persistence baseline, CatBoost, Ma'an Jordan.