Rather than listing recent flooding events, as in previous years, this annual summary analyzes and explains targeted hydrological events with an eye toward anticipating future hydrological threats to Yukon’s communities and infrastructure.
Science has made progress on anticipating how larger rivers may respond to climate change, but many local impacts are manifesting in ways that “big picture” models haven’t anticipated. For instance, the minor flood shown in Figure 1 prompted the temporary closure of the Klondike Highway near Dawson while scientific attention was directed to larger watercourses.
Breakup at Dawson – last ice jam flood occurred 46 years ago
Witnessing the power of a river during a dynamic ice cover breakup event is special, and since 1986, Dawsonites have been fortunate to watch ice slabs collide and fracture from the relative comfort and safety of the dike that protects the community against ice-jam floods. The transition from ice-covered to open-water conditions is often a fascinating event that signals the end of winter and the rapid shift into summer temperatures.
The author of this blog post had been wondering whether the ice bridge between Dawson and West Dawson offered more resistance to breakup than the nearby ice cover. The drone footage presented in a former post suggests that it does. Could that stronger, thicker ice sheet intercept a major ice run and cause an ice jam flood in Dawson in the future? This is unlikely, but not impossible. The current breakup regime of the Chu kon’ dëk (Yukon River) near Dawson includes three factors that reduce the risk of ice jam floods for the community:
- Chu kon’ dëk narrows by the Tr’ondëk (Klondike River) delta. This creates a unique geomorphological feature that can delay the initial movement of the ice cover at Dawson, allowing some time for the river system to store energy in preparation for the big push.
- A second narrowing of the channel, this one less obvious, is partly imposed by the dike that was built in 1985-86. At Queen Street, Chu kon’ dëk is less than 300 m wide and has no floodplain. This represents an uncommon combination for this river between… (the author had to search) Coffee Creek and… the Bering Sea (did River Engineers realize this fact back in the 1980s?). This channel constriction also contributes to intercepting small ice runs and to slowing down or delaying the main breakup wave, so that it eventually transits in front of Dawson with more power.
- The Moosehide segment of the Chu kon’ dëk often remains open during winter, allowing the ice run flowing through Dawson to gain momentum. This generally encourages the ice run to keep moving further away from Dawson, and as a result, ice jams, if any, would form where they don’t impact the community (as in 2025).
Normal and climate change-influenced hydrological processes (e.g., snowier winters, advanced snowmelt), in addition to the placer mining legacy in the Tr’ondëk valley, will likely continue to alter the morphology of the Chu kon’ dëk in years to come, with the possible consequence of modifying the post-1986 breakup dynamics of the reach. Following a quiet period from a river ice breakup perspective (since 1980), large ice jams may become more frequent again in the future, just like they were during the Gold Rush years. The 2023 ice jam flood at Forty Mile is a reminder that major ice jams still happen in the Chu kon’ dëk, despite warmer winters.
Ice-jam flood in Old Crow – first major ice jam flood since 1991
The October 2024 freeze-up jam in the Ch’oodeenjìk (Porcupine River) immediately downstream of Old Crow was a source of concern in anticipation of the 2025 spring breakup (location presented in Figure 2). Spring conditions arrived later than usual last May in the Vuntut Gwitchin Traditional Territory, and the hydrological scenario closely matched what Richard Janowicz, respected former hydrologist at the Water Resources Branch, described in his 2017 paper: a compressed snowmelt hydrograph causing a simultaneous breakup and freshet peak flow.
The maximum water level during breakup (May 24) was the highest since the major flood of 1991. The Water Survey of Canada station located in Old Crow was damaged during the event, but the record could be completed using the data provided by an experimental instrument (a GNSS-IR, for Global Navigation Satellite System-Interferometric Reflectometry) tested by the University of Alberta, in collaboration with the Water Resources Branch. The data (Figure 3) confirms that the stage rose rather quickly and dramatically during the night of May 23-24, and observations confirmed that the ice run was momentarily halted by this unusual freeze-up jam, causing moderate flooding within the community. (The water level at a similar flow, but without stationary ice in the river, would have generated a water level that is approximately 3 m lower.)
Factors influencing the river ice breakup sequence in the Ch’oodeenjìk near Old Crow are listed in Chapter 7, Section 4.7 of ArcticNet’s IRIS 5 report. The 2025 breakup event indicates that, whereas past scientific reports emphasized the Sriinjìk (Blue Fish River) aufeis as the potential origin of ice jam floods in Old Crow, ice run obstructions can occur much closer to the community, and their early identifications can support flood response planning. Additional research efforts are currently dedicated at improving river ice breakup forecast tools for the Ch’oodeenjìk at Old Crow, a project supported by ECCC and led through a collaboration between YukonU’s YRC-CCR group, the Government of Yukon, and the University of Alberta.
Fragility of the Yukon River sand and clay banks
Whitehorse residents remember the sudden landslide that occurred on the escarpment above Robert Service Way in 2022. Geoscientists confirmed that the slide was caused by a large volume of spring snowmelt runoff, a consequence of consecutive high-snowpack winters. Another slide occurred along the Takhini River in December 2024, a surprising date, considering the absence of surface runoff that could have promoted the failure of the slope and the assumed frozen state of the ground surface, which one would think increases its resistance to failure. It can be assumed that the riverbank was about to fail at the onset of winter, but the exact trigger of the failure remains mysterious. The occurrence of a snowmelt episode in early December, a groundwater pressure buildup behind a frozen surface, and hydrological freeze-up instabilities in the Takhini River represent possible triggering factors.
More recently, another landslide occurred at one of the high, steep riverbanks made of sand and clay that had not been eroded yet. The event occurred on the Chu Nìikwän (Yukon River) some 25 km downstream of downtown Whitehorse (Kwanlin) between May 14 and May 18, 2025. Satellite imagery (Figure 4) revealed that the Chu Nìikwän channel was significantly narrowed by the deposited debris, to a point where the YRC-CCR hydrology team judged it useful to confirm that it had not influenced water levels as far upstream as the Marwell floodplain. After verification, and with the collaboration of the Yukon Energy Corporation (YEC), who shared flow data from the Whitehorse Rapids Generating Station (WRGS), no backwater (even temporary) could be identified in the record. In fluvial hydraulics, this means that the M1 (backwater) curve reached the normal depth some distance (< 25 km) downstream of Whitehorse.
Readers may wonder what triggered this failure. The Yukon Geological Survey mentions that this riverbank was showing signs of instability before the dynamic failure occurred, therefore one could imagine that the bank had been unstable for some time. Given the period of the year, it seems that spring thaw and snowmelt likely played a role, just like it did for the 2022 landslide upstream of Whitehorse. In terms of event trigger, given the location of the failure, it could not be ignored that changing flow conditions in the Chu Nìikwän could have played a role.
The water level data collected by the Water Survey of Canada and the corresponding flow data shared by YEC reveal that several small hydrological instabilities occurred in the Chu Nìikwän during May 2025. YEC, while adjusting water levels to comply with its Water License, opened and closed a few gates at Lewes Dam during that period (Figure 5). Simply stated, when opening gates (Lewes Dam) or when turning on turbines (WRGS), the flow (and water level) in the Chu Nìikwän at Whitehorse should rise. The largest flow instability occurred on May 14, with an increase from 110 to 140 m3/s in about 15 minutes. It is not possible to confirm that this (or another) flow instability resulting from YEC operations led to the bank failure further downstream. However, it can be alleged that this geological event would have occurred sooner or later and with a comparable magnitude.
Why is this event important enough to be included in the list of 2025 hydrological highlights? Aren’t riverbank failures common after all? Can this hazard even be considered as a hydrological event? Answers lie in the analysis about the possible backwater impact of this partial channel blockage. What if a comparable (or a more significant) mass movement was to occur closer to Whitehorse in the future? It is unlikely that this would cause an immediate and prolonged flooding of the downtown area, but it could represent a concern that compares with that of the 2021 high summer flows or to the 2022 landslide.
In a context where 200-year water levels are being delineated in our flood-prone communities by consulting companies for the Government of Yukon (i.e., flood mapping projects), this event is a reminder that hydrological hazards can take many forms and significantly interact with geological processes. Climate change only stimulates Nature’s creativity in reminding us about the vulnerability of our communities and infrastructure.
Warm fall and delayed winter
Summer conditions were delayed in Southern Yukon, essentially because it was generally cooler than average until mid-July. However, it feels that temperatures remained above historical averages until December 3rd, a reality that seemed to apply to the entire territory. Was this only a perception?
Lakes are sentinels of warm weather conditions because they store heat over long periods and release it gradually when air temperatures drop in the fall. The water temperature of Chapman Lake on the Dempster Highway was still above 10°C on September 10 and the temperature of Atlin River (flowing from Áa Tlein, Atlin Lake) was 8°C on… October 20. This limited sample indicates a significant heat content at the onset of winter in two different regions of the Yukon.
The following graphs (Figure 6) present cumulated freezing degree-days (FDD, essentially a cumulated coldness index; the steeper the curve, the colder it is) for Whitehorse, Dawson, Watson Lake, and Old Crow. They reveal that, in each corner of the Yukon, consistent subzero temperatures arrived later than usual. Moreover, the FDD trend for each community continued to move away from the historical Median until early December. As of Dec. 3rd, winter had been historically warm in Watson Lake, close to record warm at Old Crow, and much warmer than usual near Whitehorse and Dawson.
This trend is now nicely departing from historical warm records because of (good old) cold weather and its associated steep FDD segment. Whitehorse is even back to average winter coldness conditions as of Dec. 15. Several messages can be implicitly obtained from the data presented in Figure 6:
- Permafrost did not appreciate late-summer conditions, especially in areas that received above-average precipitation during the months of August to October (e.g., the Ch’oodeenjìk (Porcupine River) watershed). It is also possible that ground refreezing has been delayed in many areas.
- Travelers should use caution near or on ice-covered water bodies and rivers because of the fast transition from a warm to a cold anomaly.
- Home heating should have costed less this fall and early winter compared to what it was last year. This also means that Yukon Energy will have used less fossil fuel than usual until December 3rd. Unfortunately, the cold month of December will change this tendency quite drastically.
- The November-December 2025 temperature sequence is a great illustration of the expected impact of climate change: 1. Warm conditions, 2. Extreme weather, and 3. Long-lasting temperature anomalies. This reality is unfortunately not disconnected from the general rising cost of living.
Strange freeze-up at Dawson
Every fall, conversations arise about the upcoming freeze-up of the Chu kon’ dëk (Yukon River) at Dawson. Will freeze-up occur in the “normal (suitable) way” with a complete ice cover at the ice bridge (ferry) location, or will there be a persistent gap in the ice cover in front of town? Scientifically speaking, this question is the equivalent of asking whether the ice (frazil) run coming from upriver will initially be arrested at Moosehide or at the Tr’ondëk (Klondike River) delta, two consecutive channel narrows (as described above).
The YRC-CCR was in touch with the Government of Yukon in November 2025, attempting to predict the freeze-up timing (date) and pattern (open water or not). This freeze-up was challenging to forecast because temperatures only dipped slightly below -20°C between November 11 and 15 before rising again. Several contradictory messages and observations were sent by the author of this blog, including the good news that an ice cover had formed at Dawson on November 13, a fact soon contradicted by the release of this freeze-up jam. The longer local freeze-up would take, the more likely the open water (Tr’ondëk delta freeze-up) scenario would be. In the end, that upstream scenario prevailed, and a significant open water area remained in front of Dawson on November 18, which was quite disappointing for several people, including West Dawsonites who had not crossed the river since the last day of the ferry, on October 15. The following graph (Figure 7) shows an interpretation of this yo-yo freeze-up through an analysis of water levels.
The story was not over, and as indicated in the title of this post, Nature was more creative than anticipated. Freeze-up in the lower reach of the Tr’ondëk was not complete yet, another unusual event of 2025, and a small amount of frazil slush was still feeding into Chu kon’ dëk. Each day that followed, a short segment of the open water area between Moosehide and the Tr’ondëk delta was being covered by the intercepted slush (Figure 8) until, on Novembre 24, a thin ice cover was present in front of Dawson. Although this frozen platform was initially too thin to support much weight (it was probably 10% of the normal, post-freeze-up thickness), it offered an unexpected opportunity to build an ice bridge at the usual winter crossing despite a chaotic freeze-up.
Take-home lessons from these events are:
- Hydrological processes never exactly occur as expected, especially in winter,
- Satellites products cannot replace local observations, but they represent useful research tools, especially in remote areas,
- Nature often laughs at us!
The author appreciates the contribution of the YRC-CCR team at YukonU as well as the information and insight shared by YEC, ECCC, University of Alberta (Jennifer Nafziger), Simon Nagano from Klondike H2O, and the Government of Yukon. This work is supported by ArcticNet and is greatly enriched by satellite imagery collected by the European Union’s Space programme.