Abstract
Mine closure requires designing landforms that can withstand the effects of successive storm events for centuries. This involves using precipitation time series at 15 to 30-minute intervals extended hundreds of years into the future. However, current climate change projections from global circulation models (GCMs) only extend up to 2100 and are computationally demanding.
To address this challenge, a trade-off approach using GCMs from the Coupled Model Intercomparison Project Assessment Report 6 (CMIP6) was proposed. The study generated locally representative time series using meteorological records and climate reanalysis for a nickel mine in South Africa. Over 120 GCMs from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) project were employed, covering all four Shared Socioeconomic Pathways (SSPs) emission scenarios. The Bartlett-Lewis Rectangular Pulse model with a gamma distribution was used to generate high-resolution time series data at 30-minute intervals for a 300-year period.
The historical daily records matched well with locally available data, while the bias-corrected GCMs combined with adjusted hourly records produced projected variability that exceeded historical results. Careful selection of GCMs was done based on their ability to accurately represent local variability. While the methodology has limitations, such as assuming daily precipitation characteristics at the century's conclusion are representative of a 300-year period, it offers a balanced solution to incorporate expected features into landform design models, enhancing their applicability.
Authors:
Victor Muñoz | SRK Consulting Vancouver, Canada
Mehmetcan Ozkadioglu | SRK Consulting Johannesburg, South Africa
Click here to view other presentations at this event
