Acar, Reşat; Katipoğlu, Okan Mert; Çırağ, Burak

Getting accurate daily estimates of reference evapotranspiration (ET₀) is essential for hydrological modeling, irrigation planning, and climate research. This study evaluated the performance of four machine learning models: Adaptive Boosting (AdaBoost), Categorical Boosting (CatBoost), Deep Neural Network (DNN), and Light Gradient Boosting Machine (LGBM) for daily ET₀ prediction. Feature selection was implemented using Partial Autocorrelation (PACF) and Autocorrelation (ACF) analysis on the basis of an 8-day lag structure, while the CatBoost feature importance framework was implemented to identify the most important inputs. Moreover, the methodologies of decomposition-based preprocessing, Empirical Mode Decomposition (EMD) and Variational Mode Decomposition (VMD) were implemented on the lagged inputs to build the hybrid models augmented with intrinsic constituents. These augmented hybrid models were created to improve the data representation as well as predictive capability. Model performance was evaluated using key statistical metrics, including RMSE, MAE, NSE, KGE, R2, and bias-related indicators. CatBoost produced the lowest RMSE (1.57 mm/day) and MAE (1.13 mm/day), and the highest NSE (0.67), KGE (0.78), and R2 (0.67), along with the lowest AIC score (627.15), indicating strong and reliable predictive performance. Furthermore, the VMD based CatBoost model yielded even greater improvements, highlighting the added value of decomposition enhanced learning. On the contrary, DNN showed the weakest results, with the highest RMSE (2.39 mm/day) and lowest NSE (0.23). These findings suggest that CatBoost and its hybrid variants are highly effective and reliable tools for daily ET₀ prediction, particularly in data limited environments and mountainous regions.

Demirci, Oğuz Kürşat

This study quantifies when battery-electric vehicles (BEVs) reach total cost of ownership (TCO) parity with their internal-com-bustion counterparts and characterises the cradle-to-grave greenhouse-gas (GHG) intensity trajectory to 2035. Our contribution is a transparent, Python-based framework that integrates battery-cost learning, ownership economics and life-cycle impacts within a harmonised scenario set, and links them to policy timing, ISO 15118-20-ready bidirectional charging and power-system carbon intensity. Drawing on a systematic synthesis of 221 peer-reviewed sources (2013–2025), the model runs annually for a representa-tive C-segment BEV across three scenarios (Reference, Fast-Progress, Slow-Progress; 2024–2035). The results indicate that, under the median battery-pack price learning trajectory, BEV TCO falls below the ICE benchmark around 2029 (2028–2032 across sce-narios), while life-cycle GHG intensity declines from approximately 73 to 34 g CO2-eq km-1 by 2035, spanning 29–42 g km-1 depending on grid decarbonisation. Global sensitivity analysis identifies battery price as the principal driver of TCO outcomes and grid carbon intensity as the principal driver of emissions outcomes. Results are reported for three regional aggregates (OECD average, EU-27 and China), and the policy discussion highlights contrasts for the United States to contextualise cross-market dif-ferences. Policy alignment on three fronts—parity-linked purchase-incentive phase-outs, rapid deployment of ISO 15118-20-ready bidirectional charging, and stronger recycled-content targets—shortens time to cost competitiveness and amplifies the climate ben-efits of large-scale electrification.