Can a global predictive geospatial machine learning framework model the spatially explicit stocks and transaction coefficients of waste producing processes in Copper, Nickel, and Zinc mining operations?
The escalating demand for critical minerals such as copper, nickel, and zinc, driven by the global en- ergy transition, underscores the necessity for efficient and scalable methodologies to assess mining-related waste generation. This study presents a global predictive geospatial machine learning framework designed to model spatially explicit stocks and transaction coefficients of mine wastes, including waste rock and tailings. By leveraging geospatial covariates and advanced machine learning algorithms, this research ad- dresses significant challenges related to data availability, heterogeneity, and methodological consistency. The framework integrates process modeling approaches with geospatial data analysis, machine learning techniques, and uncertainty analysis to establish a robust predictive system compatible with Life Cycle Inventory (LCI) and Input-Output (IO) methodologies. The results of this study are expected to in- crease the accuracy of both models and enables environmental decision making at unrepresented scale and resolution.
- D : data gathering
- every data soure gets a subfolder
- E : data exploration
- M : modelling
- R : result - Analysis
- ZZ: obseoltet scripts
- No units
- No empty spaces
- first Letter large
- example: Production_mass
- Purpose, Parameter, Functions, main functions structure