Heavy mental contamination in urban land is a serious environmental issue for large cities. Visible and near-infrared spectroscopy has been rapidly developed as a new method for estimating copper (Cu) levels, which is one of the heavy metals. Spectral pretreatment is essential for reducing noise and enhancing analysis. In the traditional method, the entire spectrum is uniformly pretreated. However, in reality, the influence of pretreatment on the spectrum may vary depending on the wavelengths. Limited research has been conducted on breaking down the entire spectrum into distinct parts for individualized pretreatment, an innovative method called piecewise pretreatment. This study gathered 250 topsoil samples (0–20 cm) in Shenzhen City, southwest China, and obtained their vis-NIR spectra (350–2500 nm) in the laboratory. This study divided the spectrum into three parts, each processed by six commonly used spectral pretreatments. The number of pretreated parts varied from 1 to 3, resulting in 342 PLSR models being built. Compared to the traditional method, piecewise pretreatment showed an increase in mean residual predictive deviation (RPD) from 1.55 to 1.71 and an increase in the percentage of positive outcomes in ∆RPD from 33.33% to 55.56%. Thus, we concluded that piecewise pretreatment generally outperforms the traditional method. Furthermore, piecewise pretreatment aims to choose the most effective pretreatment method for each part to optimize the Cu estimation model.

Amid global sustainability transitions, campus energy systems confront growing pressure to balance operational efficiency, resilience to extreme weather events, and sustainable development education. This study proposes an artificial intelligence-driven framework for smart campus microgrids that synergistically advances environmental sustainability and disaster resilience, while deepening students’ understanding of sustainable development. The framework integrates an enhanced multi-scale gated temporal attention network (MS-GTAN+) to realize end-to-end meteorological hazard-state recognition for adaptive dispatch mode selection. Compared with Transformer and Informer baselines, MS-GTAN+ reduces prediction RMSE by approximately 48.5% for wind speed and 46.0% for precipitation while maintaining a single-sample inference time of only 1.82 ms. For daily operations, a multi-intelligence co-optimization algorithm dynamically balances economic efficiency with carbon reduction objectives. During disaster scenarios, an improved PageRank algorithm incorporating functional necessity and temporal sensitivity enables precise identification of critical loads and adaptive power redistribution, achieving an average critical-load assurance rate of approximately 75%, nearly doubling the performance of the traditional topology-based method. Furthermore, the framework bridges the divide between theoretical knowledge and educational practice via an educational digital twin platform. Simulation results demonstrate that the framework substantially improves carbon footprint reduction, resilience to power disruptions, and student sustainability competency development. By unifying technical innovation with pedagogical advancement, this study offers a holistic model for educational institutions seeking to advance sustainability transitions while preparing the next generation of sustainability leaders.