Modeling the effects of climate change on streamflow and stream temperature in the South Fork of the Stillaguamish River

Thesis Proposal for the Master of Science Degree, Department of Geology, Western Washington University, Bellingham, Washington

Kate Clarke May 2018

Approved by Advisory Committee Members:

______Dr. Robert Mitchell, Thesis Committee Chair

______Dr. John Yearsley, Thesis Committee Advisor

______Dr. Douglas Clark, Thesis Committee Advisor

Problem Statement My objective is to model the effects of forecast climate change on streamflow and stream temperatures in the South Fork Basin of the Stillaguamish River in northwest Washington State. I will use gridded historical meteorological data to calibrate the Distributed Hydrology Soil Vegetation Model (DHSVM; Wigmosta et al., 1994) to simulate hydrology and the River Basin Model (RBM; Sun et al., 2014; Yearsley 2009, 2012) to simulate stream temperature. I will then apply gridded forecast climate data that have been downscaled to the region to predict future changes in streamflow and stream temperature through the 21st century throughout the South Fork Basin. I predict that spring and summer streamflow will decrease, and stream temperatures will increase. The Stillaguamish River (Figure 1) is an important regional water resource and serves as critical habitat for several species of salmonids (Washington State Department of Ecology (WSDOE), 2012). The results of my study will help river managers to determine where to focus salmon habitat remediation efforts. Introduction The South Fork Basin encompasses approximately 38% (660 km2) of the Stillaguamish River Basin and serves as an important resource for local agriculture and industry and for habitat for fish (Figure 1). The river is currently subject to a temperature total maximum daily load (TMDL; WSDOE, 2004; SCSWM, 2015), which means that according to the U.S. Clean Water act, it does not meet water quality standards in terms of temperature and must be mitigated. The Stillaguamish Indian Tribe relies on the river for both traditional and economic salmon fishing, so there is concern about the effects of forecasted warming climates on stream temperatures and salmon habitats. In the Pacific Northwest (PNW), global climate models project that the mean air temperature will increase between 3°C to 7°C from late 20th century historic mean temperatures through 2099 (Abatzoglou and Brown, 2012; Mote and Salathé, 2010). Previous studies of similar Puget Sound river basins predict that the forecast increases in average temperature will change precipitation patterns and result in less overall precipitation in the summers and less precipitation falling as snow in the winters (e.g., Cao et al., 2016; Dickerson-Lange and Mitchell, 2014; Murphy, 2016). Earlier snowmelt in the spring and lower streamflow in the summer have been shown to cause higher stream temperatures in Puget Sound rivers (Cao et al., 2016). Cao et al. (2016) modeled the entire Stillaguamish River Basin and found that runoff will increase in the winter and decrease in the summer in response to increased air temperature, increased evapotranspiration, earlier snowmelt, and warming stream temperatures. Field observations and a higher model resolution will allow me to refine and improve on work of Cao et al. (2016). I will use the calibrated DHSVM and RBM to project future streamflows and stream temperatures in response to 20 projected climate change scenarios that have been deemed appropriate for the PNW (Rupp et al., 2013). Modeling results will be used to evaluate threats to fish habitat and will help river managers to assess where to focus habitat remediation efforts. I will also compare my results with the modeling results of Kyra Freeman, who is conducting a similar study in the North Fork of the Stillaguamish River, which is geologically and topographically different than the South Fork Basin. Background Basin characteristics The Stillaguamish River Basin is located mainly in Snohomish County, in northwestern Washington State, and is in Water Resource Inventory Area (WRIA) 5. The Washington Watershed Management Act of 1998 divided watersheds in the state into WRIAs to define a water resource planning framework and management jurisdictions throughout the state (WSDOE, 2000). The bedrock geology of the South Fork Stillaguamish River Basin includes Jurassic metamorphic rocks in the west and Tertiary sedimentary rocks in the east (Benda, 1992). The basin has been substantially shaped by repeated episodes of continental glaciation during the Quaternary Period. The last major advance of the Cordilleran Ice Sheet (ca. 18,000 yr B.P.; Porter and Swanson, 1998) dammed the South Fork, overran much of the basin, then dammed it again as the ice sheet retreated approximately 14,000 years ago. This sequence of events formed thick, extensive glaciolacustrine silts and clays and glaciofluvial sands and gravels along the valley floors and walls throughout much of the basin (Benda, 1992; Booth et al., 2013). These thick sequences of poorly consolidated sediments, combined with a high water table, make the valley susceptible to landslides. The North Fork Stillaguamish River was once the outlet for the nearby upper Skagit River, Suiattle River, and Sauk River during the most recent glaciation of western Washington (Booth et al., 2003). As a result, the main river valley of the North Fork is much wider than the main valley of the South Fork. Due to differences in groundwater contribution to the two forks, the South Fork may respond more drastically to climate change than the North Fork. The National Oceanic and Atmospheric Administration (NOAA) classifies land use within the South Fork of the Stillaguamish River Basin as predominantly forested land with some wetlands and developed land. Timber harvesting has declined in the basin since the 1990s, but the river is still greatly influenced by historic practices, particularly logging of riparian zones, which buffer the stream temperature (Stillaguamish Implementation Review Committee (SIRC), 2005). The climate in the Stillaguamish River Basin is considered maritime, with warm, dry summers and cool, wet winters. This maritime climate classifies the watershed as a rain-snow transitional basin, which is sensitive to climate change (Vano, 2015). Small temperature changes influence whether precipitation will fall as snow or as rain. Surface elevation ranges from about 13 meters near the South Fork mouth to just over 2000 meters at the headwaters near Del Campo Peak. This relatively low elevation range makes the basin particularly sensitive to small changes in winter temperatures, which can change whether precipitation falls as snow or as rain. The position of the basin in the western foothills of the North Cascades results in a steep orographic precipitation gradient. The 30-year normal precipitation means vary between 1.17 meters at low elevations near the South Fork River mouth to about 4.56 m near the high elevation peaks (PRISM Climate Group, 2014). Rainfall runoff contributes to streamflow rapidly, whereas snow is stored and later melts and contributes to streamflow while mitigating stream temperature throughout the spring as air temperatures and day lengths increase (USGS, 2016; WSDOE, 1981). Mean annual discharge of the South Fork at the Washington State Department of Ecology (WSDOE) gauge 05A105 near Granite Falls, WA (Figure 1) is approximately 2440 cubic feet per second (WSDOE, 2018). The highest discharges occur in the fall and winter, while the lowest occur in the dry season between July and September.

Climate change in the PNW A general climate warming trend in western Washington has been reported by many studies (e.g., Mote et al., 2014; Mote and Salathé, 2010; Vano et al., 2015). Annual mean temperatures have increased by 0.6°C to 0.8°C from 1901 to 2012. Global climate models project warming of air temperatures by 3°C to 7°C through 2099 in the PNW (Abatzoglou and Brown, 2012). Future trends are expected to increase both the frequency and the intensity of precipitation events in western Washington (Mauger et al., 2016). A 2% to 5% increase per decade of spring precipitation has been observed from 1901 to 2012 (Abatzoglou and Brown, 2012). Climate models used in the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report predict increases in extreme high precipitation in western Washington and reductions in snowpack in the Cascades (Snover et al, 2013). The University of Washington Climate Impacts Group (UW-CIG) predicts that average spring snowpack in Washington will decrease by 38% to 46% by the 2040s and by 56% to 70% by the 2080s. As a result, seasonal streamflow peaks and patterns will change significantly (Snover et al., 2013).

The Stillaguamish River as a fish habitat The Stillaguamish River provides critical habitat for eight salmonid species, three of which have been classified as threatened by the Endangered Species Act since 1999 (SIRC, 2005). The Stillaguamish Tribe depends on the threatened Chinook salmon (Oncorhynchus tshawytscha) as the fish are of high cultural and economic importance. Chinook salmon runs occur once in the summer and once in the fall. The summer runs occur May to September, and the fall runs occur September to December (Kip Killebrew, personal communication, 15 March 2018). Increasing stream temperatures, projected by the UW-CIG, are likely to cause stress and migration barriers for anadromous salmon species (Littell et al., 2009). Chinook salmon that use the Stillaguamish River for summer runs are at a particularly high risk. Higher stream temperatures decrease the total dissolved oxygen content, which threatens developing salmon embryos (Wade et al., 2013). High stream temperature is also linked to loss of salmon migration capabilities, which affect how and where salmon will spawn (Wade et al., 2013). The maximum temperature threshold for safe salmonid spawning, rearing, and migration is 17.5°C (WSDOE, 2015). The WSDOE has set temperature standards throughout the basin to protect salmonid migration, with the maximum allowable temperature for headwaters set at 12°C, middle reaches at 16°C, and 17.5°C toward the mouth of the river (Figure 2; WSDOE, 2015). Preliminary future climate scenarios modeled for the Stillaguamish River Basin by Cao et al. (2016) predict over 50 days a year in which the maximum daily temperature at the outlet stream exceeds 20°C. These days that exceed 20°C are likely to occur during the warmest summer months, which correspond to the lowest streamflows and the Chinook salmon summer runs.

Stream temperature Stream temperature is influenced by a number of factors including stream depth, air temperature, solar radiation, riparian vegetation, groundwater, and snowmelt (WSDOE, 2004). Riparian cover can mitigate stream temperature by providing shade, particularly in the summer. Daily maximum temperatures are greatly affected by riparian vegetation loss. Influx of groundwater can supply cold water to the stream in some sites. Snowmelt mitigates stream temperature by supplying cold water to the stream in the spring and early summer as air temperatures rise. From 2008 to 2011, tributaries of the Stillaguamish River exceeded WSDOE temperature standards 30% of the time, and the main stem exceeded temperature standards 70% of the time (SCPW, 2015). Temperature projections for the Stillaguamish River Basin estimate increased stream temperatures of 1.8°C to 2.2°C for the 2050s and 3.1°C to 3.7°C for the 2080s (Krosby et al., 2016). Cao et al. (2016) predict substantial increases in summer stream temperatures of 0.7°C to 5.3°C over the next century and 0.2°C to 2.8°C in the winter in the PNW.

Numerical modeling The DHSVM was developed at the University of Washington and the Pacific Northwest National Lab (PNNL; Wigmosta et al., 1994) and has been applied extensively to mountainous watersheds throughout the PNW (e.g., Cao et al., 2016; Dickerson-Lange and Mitchell, 2014; Cuo et al., 2008; Sun et al., 2014; Murphy, 2016; Truitt, 2018). Gridded spatial characteristics in the model are defined by a digital elevation model, soil type and depth, vegetation and landcover, and a stream network. Hydrology is forced by meteorological data, including temperature, precipitation, wind speed, shortwave radiation, and longwave radiation at a user-defined time step (e.g., 1 hour to 1 day). The model uses physical and empirical relationships based on the gridded spatial inputs and the meteorological data to estimate an energy and water budget at the grid-cell level and estimates, for example, stream flow and snow accumulation and melt. The RBM, developed by Yearsley (2009, 2012), is a time-dependent, one-dimensional, semi-Lagrangian model that tracks individual parcels of water along the river and solves heat and energy budget equations for the conservation of thermal energy between the air-water interface. The DHSVM was modified by Sun et al. (2014) to produce energy and flow outputs for the RBM. The model estimates initial headwater temperatures based on air temperatures using a relationship developed by Mohseni et al. (1998). The RBM was modified by Truitt (2018) to account for snowmelt on initial head-water stream temperatures. Down-stream segment heat exchanges include net solar radiation, net longwave radiation, sensible heat flux, latent heat flux, groundwater, and advected heat from adjacent tributary segments. The RBM has been applied to watersheds in the PNW (Yearsley, 2009, 2012; Truitt, 2018) and is currently being used to estimate stream temperatures in the North Fork of the Stillaguamish River (Freeman et al., 2017).

Methods I will use the DHSVM and RBM to predict the timing and magnitude of streamflow and stream temperatures in the South Fork of the Stillaguamish River into the 21st century. The models will be calibrated with measured stream data and historical meteorological data. The calibrated models will be used in conjunction with projected climate scenarios to identify reaches of the South Fork of the Stillaguamish River that are particularly at risk for changing streamflow and high stream temperatures.

Scope of work 1. Use ArcGIS software to create 50-meter gridded digital basin spatial characteristics using publicly available data from government agencies 2. Calibrate and validate the DHSVM using gridded historical meteorological data and historical WSDOE gauge streamflow data 3. Conduct field work to collect data for estimating streamflow parameters for the RBM 4. Calibrate and validate the RBM using gridded historical meteorological data and historical temperature data from WSDOE and the Stillaguamish Tribe 5. Perform simulations of the DHSVM and the RBM using downscaled forecast meteorological data to estimate projected streamflow and stream temperatures 6. Statistically analyze results and identify reaches that are most at risk for temperature increases and streamflow changes

DHSVM Basin Setup The DHSVM requires digital grids of spatially variable watershed characteristics, including a digital elevation model, soil type and depth, vegetation and landcover, and stream networks. I will use a 50-meter grid resolution. Light detection and ranging (LiDAR) data and digital elevation model (DEM) data from the Washington State Department of Natural Resources were resampled to a 50-meter resolution for the Stillaguamish River Basin (Freeman et al., 2017). I will use Hydrology tools in ArcGIS to determine the watershed boundary. I will also create a landcover layer (from NOAA), a soil type layer (from Penn State University Environmental Modeling and Ecosystem Management), and a soil thickness layer using a Python script developed at the PNNL (Ning Sun, personal correspondence). Riparian cover will be estimated using first and last return LiDAR data (Akay, 2012). Details for the basin setup can be found in Appendix A of Murphy (2016).

Gridded Meteorological Data Daily historical gridded meteorological data are available at 1/16th degree latitude/longitude resolution from the Livneh et al. (2013) daily dataset. The data that have been bias corrected and statistically disaggregated to the three-hour time steps required for the DHSVM are available from the UW-CIG. Future meteorological data are available at 1/16th degree latitude/longitude resolution using multivariate adaptive constructed analogs (MACA) downscaling, a statistical method for downscaling general circulation models (GCMs) to a regional scale (Abatzoglou and Brown, 2012). MACA downscaling uses twenty different GCMs from the Coupled Model Intercomparison Project Phase 5 (CMIP5) and two representative concentration pathways (RCPs) that represent projected emissions scenarios. RCP 4.5 is a moderate emissions scenario that represents moderate warming. RCP 8.5 is a high emissions scenario that represents extreme warming. I will use the ten GCMs deemed suitable for the PNW by Rupp et al. (2013) with RCP 4.5 and RCP 8.5 emissions scenarios that have been disaggregated to three-hour time steps for the DHSVM simulations.

DHSVM Calibration I will calibrate the DHSVM to historical stream discharge measured at a WSDOE stream gauge (Figure 1). Calibration and validation will be completed by forcing the model with historical gridded meteorological data and adjusting model parameters. Identifying which meteorological grid cells yield the most reasonable precipitation magnitudes and snow coverages in the basin will be explored. Successful streamflow calibration of the model will be determined based on the Nash-Sutcliffe efficiency (NSE) coefficient method, which determines the accuracy of the model (Nash and Sutcliffe, 1970). The NSE coefficient method compares the recorded streamflow from the WSDOE stream gauge to the streamflow simulated by the historical meteorological data. The accuracy of the model is given by the NSE coefficient, where values range from negative infinity to one, with one representing a perfectly fit model. An NSE value of 0.5 is considered acceptable for hydrologic models (Moriasi et al., 2007). I will estimate the daily and monthly NSE streamflow values, as well as the daily NSE over the spring and summer flows when stream temperatures are highest. I will also calculate the root mean square error (RMSE) values between the simulated and observed daily flows, as well as the cumulative flow over the calibration period. The calibration and validation period will be from 2004 to 2013. The WSDOE gauge started recording data in August 2004 and the historical gridded data extends to December 2013.

RBM Calibration Streamflow, energy, and meteorology outputs from the DHSVM are used as inputs for the RBM. Other inputs required are the Mohseni parameters, the Leopold parameters, and an assessment of the riparian vegetation. I will calculate the Mohseni parameters, which are used to estimate headwater temperatures, from measured headwater air temperatures (Mohseni et al., 1998). I will complete field work to measure headwater and air temperatures to calculate the Mohseni parameters. I will use these parameters to estimate initial conditions of the stream temperature throughout the watershed. The Leopold parameters use observed channel morphology relationships to estimate channel velocity and depth throughout the basin (Leopold and Maddock, 1953). I will also measure channel velocity, depth, and discharge at various sites throughout the basin. Calibration and validation of the RBM will be achieved with the WSDOE gauge historical stream temperatures and other available historical data (2004 to 2013). I will use first and last return LiDAR data and landcover data to assess characteristics of the riparian vegetation along the main stem and major tributaries (Sun et al., 2014). The NSE coefficient method will be used to compare the recorded stream temperature to stream temperature outputs simulated by the RBM. The Mohseni and Leopold parameters can be adjusted to improve the fit of the model. Riparian vegetation can also be adjusted to account for more or less vegetation along the stream.

Field Work The Mohseni and Leopold parameters for the RBM inputs will be estimated partly from stream measurements and from algorithms provided by Dr. John Yearsley. Employees of the Department of Natural Resources of the Stillaguamish Tribe have placed TidbiT water temperature data loggers in the river at seven sites throughout the South Fork Basin. These temperature loggers continually record water temperature, and I will download the data from the loggers during visits to each site. I will also construct discharge profiles across the stream at each of the TidbiT sites and record stream velocity, depth, and discharge twice during the project: once in the summer to capture high flows, and once in the fall to capture low flows. I will use a Flo-Mate 2000 flow meter to measure stream velocity and the USGS midsection method (Turnipseed and Sauer, 2010) to calculate stream discharge.

Simulations and Analysis I will use the calibrated DHSVM and RBM to project future streamflow and stream temperatures of the South Fork of the Stillaguamish River using 20 MACA forecast climate scenarios (10 GCMs with RCP 4.5 and RCP 8.5 emissions scenarios). Simulations will run at three-hour time steps for 30-year intervals centered on the years 1996 (hindcast), 2025, 2050, and 2075 and will be analyzed using R, an open-source, statistical computing package. To assess threats to fish habitat, I will analyze long-term streamflow and stream temperature trends, including daily means for each of the twenty climate scenarios and the mean of all 10 GCM scenarios for each RCP. I will examine the 7-day average of the daily maximum temperature, which is an average of the maximum daily temperature over seven consecutive days. Modeling results for the South Fork of the Stillaguamish River will be compared to results from similar studies of Puget Sound basins, including the North Fork of the Stillaguamish River (Freeman et al., 2017; Cao et al., 2016; Truitt, 2018).

Expected Results Previous studies of Puget Sound rivers (e.g., Cao et al., 2016; Cuo et al., 2011; Murphy, 2016; Truitt, 2018) predict that warming in the next century will cause changes in precipitation patterns with less overall summer precipitation and less winter precipitation falling as snow. Cao et al. (2016) predict increased winter streamflow and decreased summer streamflow in Puget Sound rivers. Lower summer streamflows and earlier snowmelt would result in an overall increase in stream temperatures and earlier peak streamflow that corresponds with decreased snow-water equivalent. The South Fork of the Stillaguamish River Basin is similar in relief and characteristics as the South Fork of the Nooksack River Basin. In general, Truitt (2018) found that stream temperatures in the South Fork Nooksack increased with warmer air temperatures, a reduction in snowpack, and lower stream discharges. I anticipate a greater warming trend in the South Fork Stillaguamish than in the North Fork because of differences in channel morphology. Because the North Fork valley is wider than the South Fork valley, more groundwater helps to mitigate the stream temperature in the North Fork than in the South Fork. As a result, the South Fork may respond more drastically to climate change than the North Fork. By modeling the South Fork of the Stillaguamish River at a 50- meter resolution and calibrating that with recent meteorological data and riparian measurements, I will be able to determine regions where habitat remediation will be most crucial. The Stillaguamish Tribe will use the results from my study to determine where to focus their salmon habitat remediation efforts. My ultimate goal is to develop my thesis into a manuscript for publication in a peer-reviewed journal, such as Hydrological Processes or Climate Change.

Timeline

TASK Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Lit. review X X X X X X X X X X X X X X X

Basin setup X

Calibration X X X X X X

Field work X X X X

Data X X X X X X X X X X X X X collection Writing, drafting, X X X X X X X X X X X X X X reporting Thesis tabling and X X defense

Figure 1. Location map of the South Fork of the Stillaguamish River Basin and the Washington State Department of Ecology stream gauge to be used for model calibration.

Figure 2. Regions of maximum temperature thresholds throughout the Stillaguamish River Basin. Red markers indicate sites where stream temperature exceeds thresholds based on measured 7DAD-Max values, and blue markers indicate where measured stream temperature has not exceeded thresholds. Stream temperatures were measured each summer from 2008 to 2012 (WSDOE, 2015).

References Abatzoglou, J.T., and Brown, T.J., 2012, A comparison of statistical downscaling methods suited for wildfire applications: International Journal of Climatology, v. 32, p. 772–780.

Akay, A.E., Michael G. Wing & John Sessions (2012) Estimating structural properties of riparian forests with airborne lidar data, International Journal of Remote Sensing, 33:22, 7010-7023, DOI: 10.1080/01431161.2012.697206

Benda, L., T. J. Beechie, R. C. Wissmar, and A. johnson. 1992. Morphology and evolution of salmonid habitats in a recently deglaciated river basin, Washington State, USA. Can. J. Fish. Aquat. Sci. 49: 1246-1 256

Booth, D. B., Haugerud, R. A., Troost, K. G., 2003. The geology of Puget Sound lowland river, Chapter 2 in: Montgomery, D. R., Bolton, S., Booth, D. B., Wall, L., Restoration of Puget Sound River, University of Washington Press, Seattle, pp. 15-45.

Cao, Q., Sun, N., Yearsley, J., Nijssen, B., Lettenmaier, D.P., 2016, Climate and land cover effects on the temperature of Puget Sound streams: Hydrological Processes, v. 30, p. 2286-2304.

Cuo, L, D.P. Lettenmaier, B. V. Mattheussen, P.Storck and M. Wiley, 2008: Hydrological prediction for urban watersheds with the Distributed Hydrology-Soil-Vegetation Model, Hydrological Processes, 22(21) 4205-4213 DOI: 10.1002/hyp.7023.

Dickerson-Lange, S.E., and Mitchell, R., 2014, Modeling the effects of climate change projections on streamflow in the Nooksack River basin, Northwest Washington: Hydrological Processes, v. 28, p. 5236-5250.

Freeman, K.M., R.J. Mitchell, and J.R. Yearsley, 2017. Calibration of a hydrologic and stream temperature model to the North Fork of the Stillaguamish River for Climate Change Modeling. Geological Society of America Abstracts with Programs. Vol. 29, No. 6.

Krosby, M., Morgan, H., Case, M., Whitely Binder, L. 2016. Stillaguamish Tribe Natural Resources Climate Change Vulnerability Assessment. Climate Impacts Group, University of Washington.

Leopold, L.B., and Maddock, T., 1953, The Hydraulic Geometry of Stream Channels and Some Physiographic Implications: U.S. Government Printing Office, 68 p.

Littell, J.S., Elsner, M.M., Whitely Binder, L.C., and Snover, A. (Eds.), 2009, The Washington Climate Change Impacts Assessment: Evaluating Washington’s Future in a Changing Climate - Executive Summary, in The Washington Climate Change Impacts Assessment: Evaluating Washington’s Future in a Changing Climate, Seattle, Washington, Climate Impacts Group, University of Washington, http://www.cses.washington.edu/db/pdf/wacciaexecsummary638.pdf

Livneh, B., Rosenberg, E.A., Lin, C., Nijssen, B., Mishra, V., Andreadis, K.M., Maurer, E., and Lettenmaier, D.P., 2013, A Long-Term Hydrologically Based Dataset of Land Surface Fluxes and States for the Conterminous United States: Update and Extensions: Journal of Climate, v. 26, p. 9384–9392, doi: 10.1175/JCLI-D-12-00508.1. Mauger, G., Lee, S., Bandaragoda, C., Serra, Y., and Won, J., 2016, Refined Estimates of Climate Change Affected Hydrology in the Chehalis Basin: Report prepared for Anchor QEA, LLC. Climate Impacts Group, University of Washington, Seattle. doi, v. 10.

Mohseni, O., Stefan, H.G., and Erickson, T.R., 1998, A nonlinear regression model for weekly stream temperatures: Water Resources Research, v. 34, p. 2685–2692, doi: 10.1029/98WR01877.

Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., Veith, T. L.; “Model Evaluation Guidelines for Systematic Quantification of Accuracy in Watershed Simulations.” Transactions of the ASABE, vol. 50, no. 3, 2007, pp. 885-900., dio: 10.13031/2013.23153.

Mote, P.W., and Salathé, E.P., 2010, Future climate in the Pacific Northwest: Climatic Change, v. 102, p. 29–50, doi: 10.1007/s10584-010-9848-z.

Mote, P., Snover, A., Capalbo, S., Eigenbrode, S., Glick, P., Littell, J., Raymondi, R., and Reeder, S., 2014, Ch. 21: Northwest: Climate Change Impacts in the United States: The Third National Climate Assessment, JM Melillo, Terese (TC) Richmond, and GW Yohe, Eds., US Global Change Research Program, p. 487–513

Murphy, R., 2016, Modeling the Effects of Forecasted Climate Change and Glacier Recession on Late Summer Streamflow in the Upper Nooksack River Basin: WWU Masters Thesis Collection, http://cedar.wwu.edu/wwuet/461.

Nash, J.E., and Sutcliffe, J.V., 1970, River flow forecasting through conceptual models part I — A discussion of principles: Journal of Hydrology, v. 10, p. 282–290, doi: 10.1016/0022- 1694(70)90255-6.

Porter, S.C., and Swanson, T.W. (1998). Advance and Retreat rate of the Cordilleran Ice Sheet in southeastern Puget Sound Region. Quaternary Research 50, 205-213.

PRISM Climate Group, 2014, Average Annual Precipitation (1981-2010), Washington: Oregon State University, http://prism.oregonstate.edu/projects/gallery_view.php?state=WA.

Rupp, D.E., Abatzoglou, J.T., Hegewisch, K.C., and Mote, P.W., 2013, Evaluation of CMIP5 20th century climate simulations for the Pacific Northwest USA: Journal of Geophysical Research: Atmospheres, v. 118, p. 2013JD020085, doi: 10.1002/jgrd.50843.

Snohomish County Public Works (SCPW), 2015. Stillaguamish Temperature TMDL Adaptive Assessment and Implementation Project: Assessment Synthesis and Project Identification Report.

Snover, A.K, G.S. Mauger, L.C. Whitely Binder, M. Krosby, and I. Tohver. 2013. Climate Change Impacts and Adaptation in Washington State: Technical Summaries for Decision Makers. State of Knowledge Report prepared for the Washington State Department of Ecology. Climate Impacts Group, University of Washington, Seattle. Stillaguamish Implementation Review Committee (SIRC). 2005. Stillaguamish Watershed Chinook Salmon Recovery Plan. Published by Snohomish County Department of Public Works, Surface Water Management Division. Everett, WA.

Sun, N., Yearsley, J., Voisin, N., and Lettenmaier, D.P., 2015, A spatially distributed model for the assessment of land use impacts on stream temperature in small urban watersheds: Hydrological Processes, v. 29, p. 2331-2345.

Truitt, Stephanie E., 2018, Modeling the Effects of Climate Change on Stream Temperature in the Nooksack River Basin. WWU Graduate School Collection. 642. https://cedar.wwu.edu/wwuet/642

Turnipseed, D.P., and Sauer, V.B., 2010, Discharge measurements at gaging stations: U.S. Geological Survey Techniques and Methods USGS Numbered Series 3-A8, http://pubs.er.usgs.gov/publication/tm3A8 (accessed May 2017).

United States Geological Survey (USGS). “Snowmelt - The Water Cycle.” The Water Cycle, Dec. 2016, water.usgs.gov/edu/watercyclesnowmelt.html.

Vano, J. A., B. Nijssen, and D. P. Lettenmaier (2015), Seasonal hydrologic responses to climate change in the Pacific Northwest, Water Resour. Res., 51, 1959–1976, doi:10.1002/2014WR015909.

Wade, A.A., Beechie, T.J., Fleishman, E., Mantua, N.J., Wu, H., Kimball, J.S., Stoms, D.M., and Stanford, J.A., 2013, Steelhead vulnerability to climate change in the Pacific Northwest (A. Punt, Ed.): Journal of Applied Ecology, p. n/a-n/a, doi: 10.1111/1365-2664.12137.

Wigmosta, M. S., Nijssen, B., Storck, P., 2002: The Distributed Hydrology Soil Vegetation Model, In Mathematical Models of Small Watershed Hydrology and Applications, V. P. Singh, D. K. Frevert, eds., Water Resource Publications, Littleton, CO., p. 7-42.

WSDOE, Washington State Department of Ecology, 1981, Stillaguamish River Basin Instream Resources Protection Program, https://fortress.wa.gov/ecy/publications/publications/8111004.pdf

WSDOE, Washington State Department of Ecology, 2000, WAECY – Water Resource Inventory Areas (WRIA), geo.wa.gov/datasets/waecy::waecy-water-resource-inventory-areas-wria.

WSDOE, Washington State Department of Ecology, 2004, Stillaguamish River Watershed Temperature Total Maximum Daily Load Study, Washington State Department of Ecology

WSDOE, 2012, Stillaguamish River Basin TMDL Project Summary Page, Washington State Department of Ecology, http://www.ecy.wa.gov/programs/wq/tmdl/StillaguamishTMDL.html.

WSDOE, 2015, Stillaguamish Temperature TMDL Adaptive Assessment and Implementation Report, Washington State Department of Ecology.

WSDOE, Washington State Department of Ecology, 2018, S.F. Stillaguamish R. at Jordan Rd. Bridge, https://fortress.wa.gov/ecy/eap/flows/station.asp?sta=05A105. Yearsley, J.R., 2009, A semi-Lagrangian water temperature model for advection-dominated river systems: Water Resources Research, v. 45, p. 1–19.