Climate Change and Future River Flooding in the Northwest

CIRC researchers publish a new study on increases in flooding in the Columbia River basin under climate change

Laura Queen

River flooding in the Northwest kills people, destroys property, and both disturbs and sustains ecosystems. From historic examples such as the Heppner flood of 1903 and the Vanport flood of 1948 to flooding of the Willamette River in 2019 and the Umatilla River in 2020, it is clear that flooding remains a potent threat to our region. Furthermore, climate change is likely to exacerbate flooding throughout the Columbia River basin, according to the new study by me and other CIRC-affiliated researchers at Oregon State University and the University of Washington. We found that, by the end of the twenty-first century, floods throughout the region are likely to increase in size and in some catchments flood seasons are likely to lengthen.

Our study, recently published in Hydrology and Earth System Science, compared extreme flood magnitudes associated with 10-, 20-, 50-, and 100-year return periods (floods with a 10%, 5%, 2%, and 1% chance of occurring each year, respectively) between the modeled future (2050-2099) and past (1950-1999). At the extent of the Columbia River basin, we found increasing flood sizes due to climate change at all 396 locations we examined throughout the region. Narrowing our focus to specific rivers, however, we found a more nuanced response of both flood size and timing.

Map of our study area

Our Study Area

Our results for the Willamette River, Snake River, and the main stem of the Columbia River, revealed distinct patterns of increasing flood sizes along each river. Results of our models suggest that along the Willamette and Columbia rivers, the projected increases in flood size grow from downstream to upstream, from a 40% increase in the size of a 100-year flood along the Willamette River at Portland, Oregon, to a 60% increase along the Willamette’s upper tributaries. Similarly, the lower Columbia River near The Dalles, Oregon, is projected to experience only a 5% increase in the size of a 100-year flood, whereas flood sizes may increase by 50% upstream, in the Canadian portion of the basin. However, the Snake River follows a different, less linear pattern, exemplifying how a large tributary can influence flood behavior along a river. Traveling downstream from the headwaters of the Snake to the confluence of the Salmon and Clearwater tributaries, increases in flood size grow from 35% to 55%. The Salmon and Clearwater are larger than upper tributaries, and their flood sizes are projected to increase in flood size less than the upper tributaries. When the Salmon and Clearwater join the Snake, they dilute the influence of the upper tributaries on flooding risk, lowering the increase of flood size to ~30% along the remaining length of the Snake River, to its confluence with the Columbia River.

A key way climate change affects river flow is by shifting precipitation from snow to rain as temperatures rise. Whether precipitation falls as rain or snow has a significant impact on the behavior of rivers in the Northwest. As such, catchments can be characterized by a spectrum of hydrologic regimes, ranging from rain-dominant to snow-dominant. Rain-dominated catchments, such as the Willamette, receive most of their water from rain and usually flood in the wet winter and spring months, from December through April. Snow-dominated catchments, such as the Columbia, receive much of their water from snowmelt and usually flood from May through July. A snow-dominated system is less susceptible to large floods because during spring, melting snow is released into waterways more gradually than rain. 

Hydrologic regime shifts affect the timing of floods, which can influence the severity of a flood’s impact. For example, floods outside the expected season may inundate flood management infrastructure that was designed to accommodate the historic flood seasonality. To examine changes in flood timing, we analyzed changes from the past to future periods on the day of the year on which river flows reached their annual maximum. We found significant expansion of the flood seasons for the Snake and Columbia rivers, but not the Willamette River. Whereas the previously snow-dominated Snake and Columbia rivers historically flooded from May through July, the projected future floods occur anytime from January through June. This extension of the flood season is consistent with a shift from a snow-dominated to a mixed rain-snow system as the climate warms and less precipitation falls as snow. The timing of floods along the Willamette River, which already is rain-dominated, is unlikely to change.

To estimate the impact of climate change on river flooding, we used global climate models and hydrologic models to simulate past and future streamflow in the Columbia River basin. Global climate models simulate the global climate on the basis of potential future greenhouse gas emissions. The models’ outputs are downscaled and fed into hydrologic models, which have a higher resolution than the climate models and accurately capture hydrology in the Northwest. The hydrologic models project daily streamflow at 396 locations throughout the Northwest from 1950 to 2100. 

A strength of our study is the use of a robust, large ensemble dataset previously developed by CIRC- affiliated scientists Bart Nijssen and Oriana Chegwidden. The dataset includes 10 global climate models and 4 hydrologic models for a total of 40 future projections for each of the 396 locations in the study area. Our ensemble-based modeling built on previous studies of future flooding, which considered only a few projections. Surprisingly, our results – ubiquitous increases in flood size throughout the region – contrast with these previous studies, which found both decreases and increases in future flood size throughout the Northwest.

Our modeling did not account for flood prevention and water retention infrastructure, such as dams and reservoirs, which alter the natural hydrology of the Columbia River basin. Therefore, our findings indicated how the natural system will be affected by climate change but not how these changes may be mitigated or exacerbated by existing or future infrastructure. In ongoing research, we are including a reservoir model in our analyses to identify how climate change manifests in the real-world, regulated system. Nevertheless, our research suggested that the effects of climate change on river flooding, such as the considerable increases in flood size along the Willamette River or seasonal shifts along the Snake and Columbia rivers, will likely exceed the ability of current flood management systems to protect communities. 

Laura Queen is a research assistant at the Oregon Climate Change Research institute and is currently pursuing her PhD in physical geography at the Victoria University of Wellington in New Zealand. 

Featured Image: Photography of Sahalie Falls by Dudley Chelton. All rights reserved.

Acknowledgements: This research was funded in part by CIRC, a member of the NOAA Regional Integrated Sciences and Assessments (RISA) program and the National Integrated Drought Information System (NIDIS).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s