Title: Forecasting hazardous waste in Türkiye: a data-driven approach using European panel data

Authors: Esra Ateş; Selim Latif Sanin

Addresses: Department of Environmental Engineering, Faculty of Engineering, Artvin Çoruh University, Artvin 08000, Türkiye ' Department of Environmental Engineering, Faculty of Engineering, Hacettepe University, Ankara 06800, Türkiye

Abstract: Accurate forecasting of hazardous waste (HW) generation is essential for effective environmental management, particularly in rapidly developing countries like Türkiye. This study presents a predictive modelling framework developed using socio-economic and industrial indicators - population, gross domestic product (GDP), industrial production, municipal waste generation, and tertiary education levels - drawn from a panel dataset of ten European countries. Seven models were developed in STATGRAPHICS using regression analyses to estimate Türkiye's HW quantities without national data. Among these, model 6, based on log-transformed variables, exhibited the best performance, yielding the lowest error values (MAE = 0.156, RMSE = 0.199). The model was validated by comparing results with TurkStat data, showing a strong alignment (79%-87%) in years unaffected by external waste imports. The findings highlight population and industrial production as main predictors and suggest considering international waste trade. The approach offers a useful tool for policymaking in Türkiye and similar data-limited contexts.

Keywords: European data; forecasting; hazardous waste; industrial indicators; model; socio-economic indicators; Türkiye.

DOI: 10.1504/IJEWM.2026.155918

International Journal of Environment and Waste Management, 2026 Vol.40 No.4, pp.409 - 428

Received: 17 Feb 2025
Accepted: 08 Jul 2025
Published online: 26 Aug 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article