Rivers in the sky and floods on land: Is British Columbia’s Lower Mainland in trouble?

26-03-03

Reading time - Temps de lecture: 6 minutes

Doctoral candidate Duane Darren Noel describes his experience working at the Northern Hydrometeorology Group (NHG) under the supervision of Professor Stephen Déry, hydrometeorologist at the University of Northern British Columbia (UNBC), in Prince George, British Columbia. Duane Darren Noel’s research conducted at UNBC focused on examining the relationships among atmospheric rivers, antecedent precipitation and temperature conditions, and rain-on-snow events on recent floods in southwestern British Columbia and northern Washington state.

---

Figure 1. Photo of Duane Darren Noel in-front of I.K. Barber Enhanced Forestry Lab, the building where the Northern Hydrometeorology Group’s (NHG) offices are located at the University of Northern British Columbia (UNBC) campus in Prince George, British Columbia.

Recently, British Columbia (BC) experienced substantial rainfall due to large-scale atmospheric systems. Atmospheric rivers (ARs), also known as “rivers in the sky” are long, narrow bands of water vapour that are formed by warm air temperatures above the ocean. These develop over an extended period and develop relatively warm characteristics. Warm, moist Pacific air masses can then be entrained into ARs, often ahead of cold fronts in mid-latitude cyclones. In mountainous regions such as BC, these large bands of water vapour are pushed upwards over the mountains, leading to heavy precipitation, and potentially flash floods. 

To detect ARs, meteorologists calculate the integrated water vapour transport (IVT), a metric used to describe the horizontal movement of total water vapour content in an area. ARs exceeding the threshold of 250 kg m^-1s^-1 and lasting for a minimum duration of 24 hours are classified as AR events. Recent AR events, well above the 250 kg m^-1s^-1 threshold and minimum 24-hour duration led to major floods that caused significant damage to homes and businesses across southwestern BC (i.e. Abbotsford, BC in November 2021). However, it is unclear if rainfall from recent AR events contributed to floods in October 2024 and December 2025 in BC’s Lower Mainland.

This winter, I completed a seven-week internship at the University of Northern British Columbia (UNBC) under the tutelage of Professor Stephen Déry, principal investigator of the Northern Hydrometeorology Group (NHG) in Prince George, BC, Canada on Lheidli T’enneh First Nation traditional and unceded territory (Figure 1). The NHG’s research primarily focuses on examining the impacts of ARs and other hydrometeorological processes on the surface energy and water budgets in the Nechako Watershed. The Nechako Watershed is one of the largest watersheds in BC, where water flows eastward in the Nechako River towards its confluence with the Fraser River, located in Prince George (Figure 2). The Nechako Watershed serves as a key study site for the NHG due to its cultural, ecological and economic importance to Indigenous, and other communities living in north-central BC. 

Figure 2. The Neckako River near Cottonwood Island Park in Prince George, British Columbia.

I met several hydrometeorologists studying the impacts of drought and land-use changes in the Nechako Watershed. I also received training in extracting climate reanalysis data, managing datasets in depositories, and the integration of artificial intelligence (AI) tools to conduct research in atmospheric sciences. Additionally, I had the opportunity to speak with Professor Tamar S. Richards-Thomas from the University of Regina, a colleague of Professor Déry about her work on ARs in BC.

With this extensive training, I examined the climatological contexts of the November 2021, October 2024 and December 2025 floods in southwestern BC and northern Washington state (Figure 3A). I calculated the IVT’s magnitude and duration to detect ARs making landfall at Vancouver Island (49°N, 125°W) (Figure 3B). Then, I extracted daily streamflow from 12 naturally flowing rivers from the Water Survey of Canada (WSC) and the United States Geological Survey (USGS) databases to compare the timing of peak daily streamflow to the frequency of ROS events (a minimum of 10 mm snow water equivalent and rainfall observed while the air temperature is above 1°C) during AR landfall (Figure 3C-D, Table 1). 

Preliminary results show that ARs influenced the timing of daily peak streamflow and the frequency of ROS events during November 2021 and December 2025 floods (Figure 3B-D). Successive AR events contributed to a 100-year flood for some watersheds across BC’s Lower Mainland. In contrast, the October 2024 floods were caused by a shorter, category 2 AR event on October 13-14, followed by a much longer category 4 AR event, resulting in a 50-year flood event in some rivers in BC’s Lower Mainland. The successive category 2-4 AR events on December 8-9, 10-12 and 13-15, 2025 caused floods, ranging from a 10-year to a 50-year event.

The hydrograph analyses showed that peak daily streamflow in December 2025 occurred during successive AR landfall events (Figure 3B). November 2021 and December 2025 had ROS events exceeding 40% frequency during AR category 3 (November 2021) and AR category 4 (December 2025) landfall events. A maximum frequency of 24% ROS events was seen during AR category 4 event in October 2024, suggesting that October 2024 floods were due to heavy rainfall, rather than ROS events leading to excess surface runoff. These results indicate that rainfall from ARs influence flood generating mechanisms and flood seasonality in southwestern BC-northern Washington state region.

Figure 3. (A) The location of the study site in southwestern British Columbia-northern Washington state region. A rectangular grid was formed between 48.25°N and 50.25°N, and 120.5°W and 125°W for the analyses. The locations and drainage areas of the 12 naturally flowing rivers extracted from the Water Survey of Canada (WSC) and US Geological Survey (USGS) databases are shown in red. (B) Illustration of the 1-hour integrated water vapour transport (IVT) magnitude representing atmospheric rivers (ARs) making landfall at Vancouver Island (49°N, 125°W) at 9:00 UTC (1:00 am local time). Vancouver Island shown as a green diamond for landfall site. Contour lines represent mean sea level pressures. Arrows represent direction of the IVT. The white dot represents the location of maximum IVT. Magnitudes above 250 kg m^-1s^-1 indicate the presence of an AR. (C) Daily hydrographs of 12 naturally flowing river records from the WSC and USGS databases. Shaded regions green. yellow and orange illustrate duration of AR category 2, 3 and 4 events making landfall at Vancouver Island (49°N, 125°W), respectively. (D) The frequency of rain-on-snow (ROS) events in December 2025.

Table 1. List of hydrometric stations extracted from Water Survey of Canada (WSC) and US Geological Survey (USGS) databases in the southwestern British Columbia-northern Washington state study region.

Station Name	Database	Station IDnumber	Map ID	Latitude (°N)	Longitude (°W)	Drainage Area (km2)	Regulation type
Coquihalla River above Alexander Creek	WSC	08MF068	1	49.37	121.38	720	Natural
Chilliwack River at Vedder Crossing	WSC	08MH001	2	49.10	121.97	1230	Natural
Slesse Creek near Vedder Crossing	WSC	08MH056	3	49.07	121.70	160	Natural
Stave River above Stave Lake	WSC	08MH147	4	49.55	122.32	290	Natural
Nicomekl River at 203 Street, Langley	WSC	08MH155	5	49.10	122.66	70	Natural
Kanaka Creek near Webster Corners	WSC	08MH076	6	49.21	122.54	47.7	Natural
Spius Creek near Canford	WSC	08LG008	7	50.14	121.03	775	Natural
Seymour River below Orchid Creek	WSC	08GA077	8	49.52	123.00	63	Natural
Thunder Creek near Hewhalem	USGS	12175500	9	48.67	121.07	269	Natural
Cascade River at Marblemount	USGS	12182500	10	48.53	121.41	440	Natural
Samish River near Burlington	USGS	12201500	11	48.55	122.34	225	Natural
Stehekin River at Stehekin	USGS	12451000	12	48.33	120.69	822	Natural

In conclusion, my results show that ARs play a key role in the water budget of watersheds in this region. Careful monitoring of ARs making landfall will be vital in reducing flood hazards in the Pacific Northwest. During this internship, I learned how to critically assess changes in meteorological conditions. I am grateful for my time working at UNBC, under the mentorship of Professor Stephen Déry, and my colleagues from the NHG. I would like to thank my sponsors NSERC-LEADS and the Quebec Mobility Bursary program for making this internship possible. With this wonderful experience in Prince George, I am excited to apply these new skills on my research on flood-generating processes in eastern Canadian rivers.