STREAM POWER MAPPING OF GLACIALLY CONDITIONED RIVER CATCHMENTS IN SOUTHERN ONTARIO INTRODUCTION AND EXECUTIVE SUMMARY Roger T.J. Phillips and Joseph R. Desloges Department of Geography, University of Toronto, 100 St. George Street, Toronto, ON M5S 3G3 1 1 1 [email protected] Lawrence Ave. 1) STUDY AREA AND STREAM POWER 0 50 km N 2 5 2) DISCHARGE AND WIDTH REGIME MODELS Source Region Discharge, Q a b R 2 n This study Southern Ontario 2 Year 0.2483 0.9095 0.86 210 This study - South Georgian Bay (I) 2 Year 0.2042 0.8634 0.84 25 This study - Toronto and Region (II) 2 Year 0.2830 0.8633 0.86 32 This study - Niagara Peninsula (III ) 2 Year 0.1560 0.9454 0.82 24 This study - Grand River (IV) 2 Year 0.2439 0.9367 0.92 38 This study - Lake Erie North Shore ( V) 2 Year 0.0944 1.0461 0.79 22 This study - Thames St. Clair (VI) 2 Year 0.4002 0.8758 0.93 34 This study - South Lake Huron ( VII) 2 Year 0.4028 0.9013 0.97 22 This study - Saugeen River (VIII) 2 Year 0.1515 1.0023 0.87 23 Gingras et al. (1994) Ontario and Quebec 2 Year 0.6397 0.734 0.70 183 Gingras et al. (1994) - Lake Ontario, Erie, St. Clair (Region 7) 2 Year 0.2393 0.946 0.86 49 Gingras et al. (1994) - Lake Huron (Region 8) 2 Year 0.2438 0.883 0.88 13 Annable (1996) Southern Ontario (All Rosgen Stream Types) Bankfull 0.52 0.74 0.64 47 Annable (2011) Southern Ontario (Urban and Rural) Bankfull 0.6 0.73 - 49 Mulvihill et al. (2009) New York State Bankfull 1.067 0.810 0.89 82 Mulvihill et al. (2009) - Lake Ontario (New York State) Bankfull 0.730 0.765 0.94 10 Mulvihill et al. (2009) - Lake Erie (New York State) Bankfull 0.911 0.842 0.90 14 Richol et al. (2009) Southern Michigan State Bankfull 0.073 0.950 0.60 28 Sherwood et al. (2005) Ohio – Lake Erie Bankfull 1.951 0.637 0.82 37 Source Region Width, W a b R 2 n This study Southern Ontario (All Data) Bankfull 1.1595 0.5082 0.87 542 This study Southern Ontario (Censored) Bankfull 1.2012 0.4992 0.88 493 This study - Nottawasaga River (1) Bankfull 1.1142 0.4589 0.79 69 This study - Toronto and Region (2, 3) Bankfull 1.8464 0.4113 0.88 47 This study - Welland River, Niagara (4) Bankfull 1.2110 0.5797 0.81 35 This study - Grand River (5) Bankfull 0.8924 0.5584 0.93 54 This study - Lake Erie North Shore (6) Bankfull 2.0514 0.3984 0.78 48 This study - Thames St. Clair River (7, 8) Bankfull 1.4496 0.4749 0.93 25 This study - Ausable, Bayfield Rivers (9) Bankfull 0.9910 0.5193 0.89 129 This study - Maitland River (10) Bankfull 0.7316 0.5941 0.94 46 This study - Saugeen River (11) Bankfull 1.3173 0.4904 0.89 110 Annable (1996) Southern Ontario (All Rosgen Stream Types) Bankfull 3.71 0.49 0.61 47 Mulvihill et al. (2009) New York State Bankfull 4.256 0.401 0.84 281 Mulvihill et al. (2009) - Lake Ontario (New York State) Bankfull 2.647 0.458 0.89 33 Mulvihill et al. (2009) - Lake Erie (New York State) Bankfull 4.220 0.419 0.79 50 Richol et al. (2009) Southern Michigan State Bankfull 2.025 0.440 0.69 28 Sherwood et al. (2005) Ohio – Lake Erie Bankfull 4.632 0.356 - 45 Faustini et al. (2009) Conterminous USA Bankfull 2.81 0.24 0.42 1558 Faustini et al. (2009) - Eastern Highlands, USA Bankfull 2.68 0.38 0.75 275 Faustini et al. (2009) - Great Lakes, USA Bankfull 2.45 0.33 0.62 53 Faustini et al. (2009) - Northern Appalachians, USA Bankfull 2.55 0.39 0.72 87 Faustini et al. (2009) - Upper Midwest, USA Bankfull 1.74 0.39 0.68 70 Table 1: Discharge regime models Table 2: Channel width regime models 3) DEM Channel Slope Extraction Figure 3: Statistical regression models for (left) 2 year flood discharge (Q ) and (right) channel width (w ) versus drainage area (A ). Short dash lines are 95% confidence intervals and long dash lines are 95% prediction intervals. Censored red data points in width model were excluded as outliers in a second test based on effects of vegetation in headwaters, urbanization, and lake backwater. 2 bf d Figure 1: Specific stream power mapping for select major rivers and tributaries in southern Ontario, in the context of glacial landform mapping by Chapman and Putnam (1984, 2007). Continuous mapping of discharge and channel width along each river profile was derived from statistical regime models predicted by drainage area (A - calculated from the provincial 10 m digital elevation model). Discharge (Q) models are based on flood frequency analysis of the annual maximum mean daily discharges for 210 stream flow gauges in southern Ontario (WSC, 2011). Channel width (w) models are based on rapid surveys of bankfull width from ~500 sites. Locations of stream flow gauges and rapid width surveys are presented in . Statistical regression of log-transformed data were used to produce power-law regime models of the form: where statistical results for southern Ontario yield 0.25 and 0.91 for flood discharge frequencies of a 2 year return period (Q ); and a 1.2 and b 0.5 for bankfull channel width (w ). Statistical regression models and data, including 95% confidence and prediction intervals, are plotted in . The detailed sub-region results for the model coefficients and exponents are presented in . d 2 bf Figure 2 Figure 3 Tables 1 and 2 Q= A (Eq. 2) w=aA (Eq. 3) a d d b b a b ≈ ≈ ≈ ≈ Peninsular southern Ontario is surrounded by the Great Lakes of Huron, Erie, and Ontario, which are perched within a low-relief topography of Paleozoic sedimentary bedrock and Pleistocene glacial sediments. River systems in the study area drain to base-levels of these three lower Great Lakes and have relatively small drainage areas of less than 10 km . The purpose of this study was to map the spatial properties of stream power in southern Ontario, for eleven (11) major watersheds ( ), within the context of glacial landforms and sediments (Chapman and Putnam, 1984, 2007). As these fluvial systems are carved into the complex architecture of glacial overburden and have evolved over relatively short geologic timescales of about ten thousand years, it is expected that the longitudinal profiles and boundary materials may still be strongly conditioned by the glacial landscape. Specific stream power ( ), or power per unit bed area (units of Watts per m ), is a measure of the potential energy available in fluvial systems to perform geomorphic work, namely sediment transport (Bagnold, 1966; Parker 2011). In practical terms, mapping the spatial distribution of specific stream power only requires models of gross channel properties, such as channel slope (s) and width (w), as well as estimates of discharge (Q) conveyed through the drainage network: where is the specific weight of water (9792.3 kg m m s at 20°C). 4 2 2 -3 -2 Figures 1 and 2 w g et al. w g = Q s w (Eq. 1) -1 No Yes DEM Processing - ArcGIS Hydrology Spatial Analysis Tools - Flow Direction & Accumulation (D8 Model) - Export stream profile data as 3D point file 2.5 m Vertical Slice X Window Longitudinal Multi-Pass Moving Window Averaging X = 1 > 1.5 > 2 km window Assess Profile Slope Variations - Real Variations? - Data artifacts? - Imposed Variations? e.g., dams Accept Profile? Yes or No? X = 2 > > 5 km window Expand Moving Window in 0.5 km Increments Calculate potential specific stream power ( ) from slope (s), discharge (Q), and width (w) models = Q s w w w g -1 Import 3D stream profile data into GIS with attributes of slope (s) and stream power ( ) < Map Specific Stream Power > w GIS Longitudinal Profile Processing and Slope Generalization GIS Figure 4: Flowchart of GIS and profile processing for mapping of specific stream power. Channel slope (s) was evaluated based on longitudinal profiles derived from Ontario's provincial 10 m Digital Elevation Model (DEM) Version 2.0.0 (OMNR, 2008). Longitudinal profiles of selected major river drainages in southern Ontario were extracted as 3D points from the DEM using GIS, and channel gradient was subsequently generalized as outlined in . The scaling of the vertical slice was based on the reported absolute vertical accuracy of the DEM (2.5 m), and the moving window average was intended to represent reach scale patterns in channel slope (1 - 2 km). The generalized channel slopes were used to calculate and map specific stream power as presented in . Figures 4 and 5 Figure 1 Figure 5: Example profile for theAusable River. Profile Generalized Channel Slope Vertical Slice DEM Gradient Raw DEM Gradient 4) STUDY RESULTS AND CONCLUSIONS REFERENCES AND ACKNOWLEDGEMENTS Figure 6: A) B) Eq. 4 C) Slope-area analysis; Glacial landform classifications (Chapman and Putnam, 1984) and specific stream power curves; Bed material grain size classifications compared to Hack (1957) ; and Channel planform classifications compared to threshold slope for anabranching (s*, Eq. 5) with increasing bank strengths ( ’) based on dimensionless discharge (Q*) versus channel slope (Eaton 2010). m et al. A B C I II III V VI VII VIII IV 2 3 1 4 5 6 7 8 9 10 11 Lake Huron Lake Erie Lake Ontario Georgian Bay Legend N 0 50 km Stream Flow Gauges Rapid Width Surveys Figure 2: Study area map for southern Ontario showing select major river drainages (1 - 11) and sub-regions for discharge regime models ( ). I - VIII Thanks to NSERC for financial support through a PhD CGS Scholarship to R.T.J. Phillips. Generous academic and financial support by co-author (Joe Desloges) has also been vital to this PhD thesis research. Honourable mention to many field assistants over period of 2010-2011, namely Stephanie Mah, Joyce Arabian, Beata Opalinska, James Thayer, and Jennifer Henshaw. Advice and insights from first author’s PhD Supervisory Committee are greatly appreciate; thanks Sarah Finkelstein, Bill Gough, Brian Branfireun, and Nick Eyles. Thanks to the many agency staff from OMNR and Conservation Authorities who responded to our information requests, including Brynn Upsdell, Kari Jean, and Ross Willson (ABCA); Peter Dragunas and Tony Difazio (CCCA); Shannon Wood and David Pybus (SVCA); Glenn Switzer (NVCA); Muriel Andreae and Rick Batterson (SCRCA); Joe Gordon and Brian Widner (KCCA); and Kent Todd (OMNR). Annable, W.K. (1996). . Toronto, Ontario Ministry of Natural Resources, 92 p. Annable, W.K., Lounder, V.G., and Watson, C.C. (2011). Estimating channel-forming discharge in urban watercourses. , 738-753. Bagnold, R.A. (1966). . U.S. Geological Survey Professional Paper, , pp. 1–37. Chapman, L.J. and Putnam, D.F. (1984, 2007). ; Ontario Geological Survey, ISBN 978-1-4249-5158-1. Eaton, B.C., Millar, R.G., and Davidson, S. (2010). Channel patterns: Braided, anabranching, and single-thread. , 353-364. Faustini, J.M., Kaufmann, P.R., and Herlihy,A.T. (2009). Downstream variation in bankfull width of wadeable streams across conterminous United States. , 292-311. Gingras, D., Adamowski, K., and Pilon, P.J. (1994). Regional flood equations for the provinces of Ontario and Quebec. , 55-67. Hack, J. T. (1957). , U.S. Geological Survey Professional Paper, pp. 45–97. Leopold, L.B. and Wolman, M.G. (1957) . U.S. Geological Survey Professional Paper, 39–85. Mulvihill, C.I., Baldigo, B.P., Miller, S.J., DeKoskie, D., and DuBois, J. (2009) : U.S. Geological Survey Scientific Investigations Report 51 p. Nanson, G.C. and Croke, J.C. (1992). A genetic classification of floodplains. , 459-486. OMNR. (2008). . Water Resources Information Program, Geographic Information Branch, Ontario Ministry of Natural Resources. Parker, C., Clifford, N.J., and Thorne, C.R. (2011). Understanding the influence of slope on the threshold of coarse grain motion: Revisiting critical stream power. , 51-65. Rachol, C.M., and Boley-Morse, K. (2009). : U.S. Geological Survey Scientific Investigations Report 300 p. Sherwood, J.M. and Huitger, C.A. (2005). : U.S. Geological Survey Scientific Investigations Report 38 p. WSC. (2011). [Digital Files]. Accessed online: [May, 2011]. Morphologic relationships of rural watercourses in southern Ontario and selected field methods in fluvial geomorphology River Research and Applications An approach to the sediment transport problem from general physics Physiography of Southern Ontario Geomorphology Geomorphology Journal of the American Water Resources Association Studies of longitudinal profiles in Virginia and Maryland River channel patterns: braided, meandering and straight Bankfull discharge and channel characteristics of streams in New York State Geomorphology Provincial Digital Elevation Model (DEM) Version 2.0.0. Product Specification – Edition 1.1 Geomorphology Estimated bankfull discharge for selected Michigan rivers and regional hydraulic geometry curves for estimating bankfull characteristics in southern Michigan rivers Bankfull characteristics of Ohio streams and their relation to peak stream-flows Water Survey of Canada hydrometric monitoring program - HYDAT Database, Environment Canada 27: 422 I Miscellaneous Release--Data 228 120: 108: 30: 294 B: 282B: 2009–5144: 4: 126: 2009–5133: 2005–5153: ‐ ‐ http://www.ec.gc.ca/rhc-wsc Stream networks of southern Ontario are carved into a complex architecture of glacial landforms and sediments, draining to base-levels of the lower Great Lakes (Huron, Erie, Ontario). These fluvial systems have evolved over relatively short geologic timescales of about ten thousand years, with longitudinal profiles and boundary materials still strongly conditioned by the glacial landscape. Based on longitudinal profiles derived from the 10 metre provincial digital elevation model (DEM version 2.0.0; OMNR, 2008), potential stream power is mapped for eleven (11) major river catchments across southern Ontario . Due to multi-scale variance in raw DEM stream gradients from both real and artifact sources, vertical slice and multi-pass moving window approaches are used to generalize channel slope to the reach scale (typically 1 – 2 km, as a representative spatial scale for fluvial processes operating over long time periods). Results of a slope-area analysis demonstrate that fluvial channels may be over- steepened or under-steepened in association with glacial landform classifications. Specific stream power ( ) results generally fall within the theoretical domain for single-channel meandering rivers, with an “average” condition of = 30 Wm . Despite higher stream powers in some reaches, single-channel meandering fluvial planforms still dominate the landscape of southern Ontario due to the moderating effects of: bed material sediment calibre from inherited coarse-grained valley fills limiting fluvial bedload transport; and/or bank strength controlled by sediment cohesion and vegetation. Understanding the spatial properties of stream power can provide insights into regional patterns of sediment transport and alluvial floodplain development, relevant to applied aspects of aquatic ecology, archeology, erosion assessment, stream restoration, and urban storm water management. (Figure 1) 1) 2) ω ω -2 From mapping of specific stream power ( ), a sample of 117 reaches were selected and classified based on glacial landforms by Chapman and Putnam (1984, 2007). Results of a slope-area analysis demonstrate that fluvial channels may be over-steepened (e.g, moraines) or under-steepened (e.g., till plains, glaciolacustrine plains) in association with glacial landform classifications ( ). Specific stream power ( ) results generally fall within the theoretical domain for single-channel meandering rivers (10 - 60 Wm ; Nanson and Croke, 1992), with an “average” condition of = 30 Wm . Constant stream power curves for southern Ontario (based on Eq.’s 1, 2, and 3) plot on a slope-area graph with a slope exponent of -0.4 (Leopold and Wolman, 1957). Despite higher stream powers in some reaches which are often associated with multiple-channel planforms, single-channel meandering fluvial planforms still dominate the landscape of southern Ontario due to the moderating effects of bed material sediment calibre and bank strength. Systematic decreases in bed material size downstream are frequently interrupted by glacial landforms and sediments, in some cases with inherited coarse-grained valley fills limiting fluvial bedload transport. Expected downstream variations in grain size for a theoretical “graded” profile have been previously described by Hack (1957), as Figure 1 Figure 6A ω ω -2 -2 summarized in and presented in . Limitations on bedload transport, particularly in cobble dominated channels, will suppress bar development and thus multiple-channel planforms. where M is the median particle size in mm (Hack, 1957). Also, high relative bank strength, controlled by sediment cohesion and vegetation, will limit formation of multiple channels by restricting bank erosion and channel widening. As demonstrated in , many single-channel reaches fall above the multiple-channel anabranching threshold (s*, ) for low relative bank strength ( ’ = 1); however, they tend to be consistent with the single-channel meandering domain for higher bank strengths ( ’ 2) (Eaton 2010). where s* is the critical threshold slope for multiple-channel anabranching, ’ is dimensionless relative bank strength, and Q* is dimensionless discharge (Q*) (Eaton 2010). The spatial properties of stream power can contribute to understanding regional patterns of sediment transport and alluvial floodplain development, with application to many other disciplines. Eq. 4 Figure 6B Figure 6C Eq. 5 s = 0.006 (M/A ) (Eq. 4) s* = 0.40 ’ 1.41 Q* (Eq. 5) d 0.6 -0.43 m m m ≥ et al. et al. m