Projecting Changes to Future Extreme Precipitation Events for the Design of Tailings Facilities

Tailings storage facilities (TSFs) in British Columbia (BC) are designed to withstand a range of large and
statistically infrequent rainfall and snowmelt events. Estimating the rainfall and subsequent runoff from
these events generally relies on limited periods of on-site rainfall records that are supplemented by regional
data. The result is the development of intensity-duration-frequency (IDF) curves that describe the
relationship between rainfall intensity, rainfall duration, and return period. These curves are then used in
the design of water management elements associated with the TSF.

While TSFs have a limited operational life, their design life extends beyond this into closure. From
published data and studies, a warming climate in BC will result in an increase in intensity and frequency of
future extreme rainfall events. This will affect IDF curves, which in turn may affect the required size of
water management structures such as diversions and spillways. The closure plan/design for each TSF is
unique, but in all cases there are water management elements. Development of a closure plan for TSFs
should therefore take into consideration the impact of future climate scenarios.

In this paper, projected changes in extreme rainfall are estimated for several mine sites in BC using
four different methods. These methods include reviewing trends in historical data; reviewing trends in data
from downscaled (i.e., modelled) future climate scenarios; using the IDF_CC Tool developed by the
University of Waterloo; and scaling based on estimated increases in atmospheric moisture-holding capacity
(i.e., Clausius-Clapeyron scaling). The expected change and estimated confidence limits from each method
are compared to evaluate their relative merits.

Results of the comparison indicate that there is little or no benefit in using site-specific rainfall data
that spans only a short period of record. Most historical datasets, particularly those with a period of record
of less than 50 years, are only marginally adequate for estimating 1,000-year return period events, and
simply do not contain enough data to project future changes of such events. Based on the observed level of
uncertainty involved in the climate change projections, more emphasis should be placed on obtaining
consistent and reliable estimates than site-specific ones.

Clausius-Clapeyron scaling was found to be the most consistent and reliable method for projecting
future changes to maximum precipitation. For an increase in average annual temperatures of 3.1°C in BC
from historical levels to the end of the 21st century, Clausius-Clapeyron scaling estimates a 25% increase
in extreme rainfall magnitudes. A downscaled future climate scenario is appropriate for estimating future
changes to secondary flood-affecting factors such as snow accumulation, evaporation, and soil moisture.

Clark, S. and M-J. Piggott. 2019. “Projecting Changes to Future Extreme Precipitation Events for the Design of Tailings Facilities,” in Proceedings of Tailings and Mine Waste 2019, 17-20 November 2019. Vancouver, BC : University of British Columbia.