This study stemmed from the concerns raised regarding the long term trying trends affecting the Peace Athabasca Delta (PAD), one of the world's largest freshwater deltas. A common perception during the 1970's and 1980's was that lower flows on the Peace River due to regulation by the W.A.C. Bennett Dam, minimized the probability of large open water flood events capable of inundating the perched basins of the PAD adjacent to the Peace River. Local traditional knowledge and other anecdotal information suggested that ice jams also played a role in some flood events.
Analysis of hydrometric data in conjunction with various historical and local-knowledge sources confirms that open-water floods have been ineffective in producing high-elevation floods along the Peace River adjacent to the Peace-Athabasca Delta. Even the historically high flow event of
1990 did not produce a flood of sufficient magnitude to flood high-elevation portions of the delta. Over the period of hydrometric record, backwater produced during river-ice breakup has exceeded that of the 1990 open-water event on several occasions. It was breakup backwater, in 1974, that resulted in the last major flooding of the elevated perched basin within the PAD.
In the lower portions of the Peace River, flow regulation seems to have produced minor changes in factors, such as ice thickness and strength, that could significantly affect the severity of breakupandrelatedice-jamflooding. Temporalanalysisofthesefactors,however,alsodetected a weak climate signal suggesting that since approximately the mid-1970's the period of ice cover may have become slightly warmer and the pre-breakup melt period may have become more intense and/or more protracted.
A common perception was that reduced flows due to regulation were responsible for the decline in severe ice jams. Flow contributions from the point of regulation, however, are higher, on average, at the time of breakup near the PAD in the post-regulation period than prior to regulation. The major ice-jam floods that occurred in the 1960's prior to regulation and in the early 1970's after regulation have been associated with large runoff events from downstream tributaries especially the Smoky River. These large tributary flow events also appear to be correlated with large spring snowpacks and associated snowmelt runoff. A preliminary evaluation of temporal trends in the size of the snowpack on the Smoky River suggests that there has been a shift in the mid-1970's to values lower than the long-term average. A similar trend in snowpack accumulation has been identified in British Columbia.
The major effect of regulation on the occurrence of breakup ice jamming near the Peace- Athabasca Delta is related to higher winter flows and increased freeze-up elevations. In general, the higher a freeze-up cover, the greater the flows it can pass without breaking. Two runoff sources combine to generate spring flows that can exceed the freeze-up level:: the upstream flow from above the point of regulation and the downstream tributary flow. Under regulated conditions, a major increase in upstream flows (above the point of regulation) is unlikely at the time of breakup near the Peace River Delta due to the operational transition to lower summer releases. Furthermore, if the amount of regulated flow at the time of breakup is declining, tributary flow will have to account for this "loss" to the main-stem discharge, before having an affect on the ice cover. Under the current regulated regime, production of severe breakups has become more dependent on tributary inflow, particularly from the Smoky River. Large spring runoff from the tributaries have been effective since regulation in producing large breakup floods (e.g., 1972 and 1974) but the apparent decline in spring snowpacks has reduced their subsequent effectiveness. Thus the absence of a large-order event since 1974 seems to be related to a combined effect of flow regulation and the vagaries of climate.