Salmonids are vulnerable to warmer water temperatures and to floods, both of which are projected to increase with climate change. Water temperature in lower Chico Creek is already teetering on too warm – it consistently exceeds the standard seven-day average daily maximum of 16ºC for Core Summer Salmonid Habitat set by the Washington State Department of Ecology (Natural Systems Design Inc., 2014). Projected air temperature increases and precipitation changes are likely to exacerbate this warming trend in the future, threatening the persistence of salmon.

A warmer climate is also likely to intensify heavy rain events known as atmospheric rivers (Warner et al. 2015; Espinoza et al. 2017), which are a primary driver of flooding in Chico Creek. Floods can directly harm salmonids by washing eggs out of redds (egg nests), causing physical harm to some fish, and by prematurely flushing juveniles downstream.

Prior to this project, no specific assessment of projected changes had been conducted for the Chico Creek watershed, and few if any detailed studies examined these effects on small rain-dominated watersheds more broadly. In addition, all previous studies in the region have used statistical downscaling, and recent research has shown that dynamical downscaling is needed to accurately capture changes in rain intensity (Salathé et al., 2014).

In recent years there have been significant efforts by the Suquamish Tribe, Kitsap County and other partners to implement fish passage and other habitat restoration projects, and to secure conservation protections for riparian corridors to benefit salmon. The data from this study will help these groups understand how climate change may affect salmon; this information will help the Tribe prioritize conservation and restoration actions.

Past hydrologic studies have typically used interpolated estimates of daily weather on model grid cells (statistical downscaling, e.g.: Hamlet et al. 2013). In this project we use dynamically downscaled historical and future weather data (Chen et al., 2018, Lorente-Plazas et al. 2018) as input to the hydrologic model and water temperature simulations. In addition to providing better estimates of changing heavy rain events (Salathé et al. 2014), these new datasets allow us to consider hourly variations in streamflow, whereas previous approaches were limited to daily meteorology.

We used the fine-scale Distributed Hydrology Soil Vegetation Model (DHSVM, Wigmosta et al. 1994) to simulate historical and future flows in Chico Creek. We calibrated the model by adjusting the soil and vegetation properties to improve the match with observed flows, focusing on Dickerson Creek, which is not influenced by a lake upstream. We then processed the hydrologic model results to evaluate changes in monthly flows, along with peak and low flow statistics. For low flows we produced results for two durations (one day, seven day) and two return intervals (two- and 10-year events).

We then used a process-based water temperature model that integrates DHSVM with a vector-based stream temperature model (RBM, Yearsley 2009) and a riparian shading model (Sun et al. 2015). We obtained estimated riparian conditions from our collaborator at the Suquamish Tribe (Steve Todd, personal communication), who provided approximate estimates of tree height, buffer width, overhang and canopy bank distance across the watershed. We then calibrated the model for the six river locations with sufficient data, focusing on the summer season since that is of primary interest to the Tribe. Results are provided as the raw hourly time series, along with monthly averages and changes in “7DADMAX”, defined as the 7-day average of daily maximum water temperatures. Using 7DADMAX, we also calculate the number of days between June 15th and September 15th with a 7DADMAX greater than 16ºC.

Models project near-zero changes for winter and summer precipitation in Chico Creek, and more intense heavy precipitation for all seasons. Our streamflow results are consistent with these precipitation changes. We find very small increases in winter and summer streamflow (approximately one to nine percent), and even smaller increases in extreme low flows (approximately three to four percent). The peak flow projections are also consistent with the precipitation projections, showing average increases of approximately 10-15% for the two-year event. These results are very similar to the findings of Murphy and Rossi (2019) for similar Creeks in the region (Stavis, Seabeck, Big Beef, and Little Anderson). However, Murphy and Rossi (2019) did project larger increases in winter streamflow and larger decreases in summer streamflow, likely due to differences in the climate projections used in their study relative to the current analysis.

The water temperature projections reflect the substantial warming trends for air temperature. The maximum 7DADMAX in each water year is projected to increase from +2.1ºC to +2.9ºC, on average, by the 2080s relative to 1980-2009. Individual model projections for the same quantity range from +0.2ºC to +5.1ºC, again for the end of the century. Another way of viewing these results is in terms of the number of days when 7DADMAX exceeds thermal tolerances for salmonids. Focusing on the warm season (June 15–September 15), our simulations show that Dickerson Creek historically had 12 days each summer when 7DADMAX exceeded 16ºC. By the 2080s, this is projected to rise to 65 day per year, on average, with individual model projections ranging from 22 to 85 days.

An important achievement in the current study was the incorporation of new dynamically downscaled projections in a hydrologic modeling study. This is particularly important in rain-dominant basins such as Chico Creek, where statistical downscaling approaches could not capture future changes in precipitation patterns. However, there are several areas with potential for improvement. Model results downstream of Kitsap and Wildcat lakes are unlikely to be reliable because DHSVM does not account for the effect of the lake on flow, and RBM does not account for the substantial warming effect of each lake. In addition, our modeling lacks an adequate treatment of groundwater and ignores potential changes in land cover; each of these could alter the response of the watershed to climate change.

This work was funded by the Bureau of Indian Affairs, via a sub-contract with the Suquamish Tribe. The authors thank Steve Todd for his help and input throughout the project, and Tom Ostrom and Paul Williams for their review of the final report.

Photo credit: MarcoTjokro3, Unsplash

Climate Change & Flooding in Snohomish County: New Dynamically-downscaled Hydrologic Model Projections. We produced new projections of future streamflow, with a particular emphasis on flooding, for the Snohomish and Stillaguamish Rivers. Funded by Snohomish County, the data will support the County’s efforts to incorporate climate change impacts in floodplain management decisions.

Heavy Precipitation Projections for Use in Stormwater Planning. Through a series of projects, we developed an online tool and online repository of data and other information from dynamically-downscaled projections. Results are designed to support both municipal and rural stormwater planning.

Effect of Climate Change on Flooding in King County Rivers: Using New Regional Climate Model Simulations to Quantify Changes in Flood Risk. We produced a new set of projections of 21st century climate, developed using a regional climate model. A key feature of these projections is that they provide hourly estimates of future weather conditions (temperature, precipitation, humidity, wind, etc.).

Changing Streamflow in Icicle, Peshastin and Mission Creeks. The purpose of this project was to leverage existing hydrologic change datasets to estimate future changes in streamflow in Icicle, Peshastin, and Mission Creeks as well as the seven alpine lakes. These will be used to evaluate proposed alternatives for managing water in Icicle Creek.

Photo credit: Eliud Echevarria/ FEMA