How can the uncertainty in the natural inflow regime propagate into the assessment of water resource systems

The Canadian Rocky Mountain headwaters support the water resource systems of the Canadian Prairies. Significant variations in natural headwater contributions have been observed due to warming climate. Projecting future natural headwater flows under climate change effects, however, has large uncertainty. First, there are difficulties in climate modeling and downscaling in alpine regions. Second, streamflow modeling in mountainous areas is extremely challenging. There is therefore a need to understand the effects of uncertainty in the natural inflow regime, and in particular how this translates into uncertainty in representing the state and the outflow of water resource systems. Considering the Oldman River basin in Alberta, Canada, we synthesized different inflow regimes based on site/inter-site properties of the historical inflow regime. The water resources system was then conditioned on the synthesized inflow regimes to identify the mechanisms of error propagation from the headwater streamflows to the water allocations. The results show that the response of the water resource system to the uncertainty in the generated inflow regime depends on the system state, flow condition and the component of interest. Generally, the response of the reservoirs to the uncertainty in the estimated inflow regime is more significant in dry years, in particular during low flow conditions. The response at the system outlet is rather different, as the propagation of the headwater uncertainty is more significant during high flow conditions. Also, similar inflow estimates in terms of error and uncertainty may result in different error and uncertainty estimates in the simulated outflows; therefore, lower bias and uncertainty in estimating the regional inflow regime does not necessarily mean lower bias and uncertainty in simulating the streamflow at the outlet of the system. Our results provide improved understanding of uncertainty propagation through complex water resource systems, but also portray the need for better climate and hydrological modeling in the Rocky Mountains for improved water management in the Canadian Prairies, particularly in the face of uncertain climate futures. This will be crucial if the natural headwater inflows decline and/or the system faces drought conditions.