Authors: Weitao Nick Rong, Jaime Cathcart
Conference: CDA 2022 Annual Conference
Date: October 17-19, 2022
Water management planning that considers a wide range of climatic conditions is crucial for dam safety. Quantifying the climatic range, however, poses a challenge since site climate records are generally much shorter than what is needed to account for possible climate variability. To reduce the uncertainty of long-term climate estimates for mine sites, regional climate records are typically used to represent mine conditions, often with adjustments according to correlations with short-term site data. Long-term regional datasets, however, may not be available, or they may not exhibit similar hydroclimatic conditions to those of a mine site because differences may result from various topographic factors such as aspect, elevation, and distance to site. Furthermore, regional records may have data gaps, may be of unknown quality, or may not be concurrent with the site data.
State-of-the-art global reanalysis products such as ECMWF ReAnalysis 5 (ERA5) can help reduce the gap between what practicing professionals need (long-term climate estimates) and what they typically have (short-term site\ measurements). Reanalysis products result from combining past observations (collected from various sources such ground stations and earth observation satellites) with model results (weather model hindcasting) to generate consistent time series of multiple climate variables for the period of 1979 to present. Reanalysis products provide complete and consistent data, which can take the place of or supplement regional data. These products come in the form of gridded data that represent average conditions for areas with 8 to 32 km spacing, so similar to conventional regional data, they must be bias-corrected to site observed data to generate reliable long-term estimates of hydroclimatic conditions for a particular site.
Probability-based bias-correction (BC) techniques such as the quantile mapping (QM) method are widely used to bridge the gap between global climate models (GCMs) and local observations in climate change related research, and similarly can be used to bias-correct reanalysis products, though this is not common in the mining industry.
This paper presents a case study of the use of a reanalysis product (ERA5) and a probability-based BC method to develop long-term climate values for a mine site with short-term climate records. The results demonstrate that this state-of-the-art approach can produce reasonable long-term estimates of hydroclimatic conditions for a region where regional observations are scarce. Use of such a state-of-the-art approach can facilitate mine water management planning and increase dam safety by improving estimates of historical hydroclimatic variability.
Download the full paper.