1 Project Summary. 2 Background

Transcription

1 Student: Lindsay Olinde Course: CE 394K..3 GIS in Water Resources UT Austin Professor: Dr. David Maidment Submitted: December 3, 2010 Investigation into whether differences in channel slope characteristicss exist due to varying lithology in coarse mountain streams, Reynolds Creek watershed, ID

2 1 Project Summary I used this term project to better understand the larger context of the mountain watershed of Reynolds Creek in southwest Idaho because I am currently conducting a field experiment on a 100m section of Reynolds Creek. A common way to examine channels on watershed scales is to first characterize their longitudinal elevation profiles by examining slope characteristics. Due to parts of the Reynolds Creek watershed being underlain with basalt and granite, in this GIS project, I examined whether these two different lithologies force unique profile curvatures in several small mountain streams. Of the streams I analyzed, lithology does not seem to be an obvious control on slope characteristics because the granite and basalt streams have quite a bit of overlap in the results. In Spring 2011, I plan to extend my analyses of these ArcGIS result by performing power spectral analysis on the concavity and curvature signals at different length scales using Matlab and examining the slope area relationships here. In this report, I present background on using slope to gain insight into channels, a description of the mountainous portion of the Reynolds Creek watershed, ArcGIS techniques I applied, discussion of my ArcGIS results, and related conclusions. 2 Background Because the concept of slope as a characteristic in landscape evolution and erosional forcing may be a new perspective for some, in the next few sections, I discuss how slope may be considered from the fluvial geomorphology perspective. 2.1 Channel slope and channel equilibrium According to the concept of river equilibrium, channels adjust to new equilibrium, or grade conditions, following system perturbations such as changes in discharge, geometry, or sediment supply. Adjustments proceed through a series of feedbacks between discharge, width, slope, grain size, and sediment flux such that transport capacity is just balanced with sediment supply [Gilbert, 1914; Leopold, 1980; Mackin, 1948; Strahler, 1952]. Transport capacity refers to the amount of material a channel is able to move downstream [Gilbert, 1914]. A classic fluvial geomorphology image, Lane s Diagram, simplifies this concept of transport capacity and channel equilibrium, showing the adjusting feedbacks that can balance slope, transport capacity, and sediment supply, Figure 1. For example, if a stream has coarse sediment, which is common in mountain streams, the stream will remain steep to be able to transport this coarser sediment resulting in no degradation or aggradation, i.e. equilibrium. A general way evaluate slope is to take the average slope of stream elevations over a longitudinal distance, Equation 1. The convention for stream is to report longitudinal slopes as positive when longitudinal elevations are decreasing as channel flows downstream, so I will follow this convention in my results. Equation 1

3 Figure 1 Lane's Diagram illustrating the dynamic relationship between sediment flux characteristics, slope, and discharge such that transport capacity balances sediment supply [Henson, 2010]. 2.2 Slope and sediment transport Slope is an important stream characteristic because a channel s slope influences the basal shear stress,, particles on the channel bed experience. Simply, basal shear stress is the downstream drag force per area of channel bed that surface particles experience due to gravity driving water discharge, Q, downhill over the bed. For steady, uniform flows in wide streams, reach averaged slope, S, result in average basal shear stresses that can be estimated with Equation 2, where is water density, g is gravity, and h is water depth [Kondolf and Piégay, 2003]. Thee rate of work a stream can do is reflected in stream power, Ω, and can be calculated by stream velocityy times basal shear stress, Equation 3 [Leopold et al., 1964]. Particle entrainment can occur if the basal shear stress is greater than a threshold shear stress, which varies according to particle size and arrangement in the bed, and depending on the degree of transport, a particle will move as bedload, rolling or saltating along the bed, or suspendedd sediment. Particle entrainment conditions are important in understanding the degree and form that sediment is transported through a watershed and, therefore,, are crucial for watershed managers estimating watershed sediment budgets. Equation 2 Equation 3 Ω 2.3 Curvature in the direction of channel slope Having established why slope is an important characteristic inn sediment transport and stream capacity through Equations 2 and 3, one may also want to understandd characteristics of channel slopes. For example, how does a slope change over a longitudinal distance of a channel? Often hillslopes evolve with overall convex surfaces due to overland, sheet flow driven erosion being slope dependent while streams evolve to have concave up longitudinal profiles due to concentrated channel erosion being forced by slope and discharge [Quimpo, 1999]. One may consider that the more concave a stream is, the more erosion it has experienced. ArcGIS can also be used to evaluatee the second derivate of surface elevations, i.e. a surface s curvature. Surface curvature is calculated by ncorporating fourth order polynomials over the surface created from 3 by 3 cell squares [ArcGIS 2010]. Surface curvature is represented by Equation 4 and with parameters visually understood with Figure 2 [ArcGIS, 2010]. I applied the used the ArcGIS profile curvature output as opposed to the general curvature output raster because the profile curvature describes the

4 curvature of the surface in the direction of slope, as shown inn Figure 3 [de Smith et al., 2009]. The degree of a stream s profile curvature influences erosion because it reflects the acceleration or deceleration of discharge [ArcGIS, 2010]. Equation 4 Z = Ax²y² + Bx²y + Cxy² + Dx² + Ey² + Fxy + Gx + Hy + I where coefficients are shown in Figure 2 and are below: A = [(Z1 + Z3 + Z7 + Z9) / 4 (Z2 + Z4 + Z6 + Z8) / 2 + Z5] / L4 B = [(Z1 + Z3 Z7 Z9) /4 (Z2 Z8) /2] / L3 C = [( Z1 + Z3 Z7 + Z9) /4 + (Z4 Z6)] /2] / L33 D = [(Z4 + Z6) /2 Z5] / L2 E = [(Z2 + Z8) /2 Z5] / L2 F = ( Z1 + Z3 + Z7 Z9) / 4L2 G = ( Z4 + Z6) / 2L H = (Z2 Z8) / 2L I = Z5 Figure 2 Curvature calculated from fitted profiles over nine surface elevation cells [ArcGIS, 2010]. Figure 3 Depiction of profile curvature over a hillslope [dee Smith et al., 2009].

5 2.4 Lithology Lithology, the type of rock, on the surface of a watershed can influence how easily a surface can be eroded due to unique rock strengths. Basalt and granite are both igneous rocks with basalt forming from quickly cooling lava (think of what exits volcanoes) while granite forms from slowly cooling magma (consider what remains in volcanoes). Basalt is typically slightly denser than granite, with specific gravities of approximately 3.3 and Site description Reynolds Creek is a coarse alluvial channel located in the Owyhee Mountains and managed by USDA Agricultural Research Service within the Reynolds Creek Experimental Watershed. The contributing mountainous drainage area is approximately 55km 2 and is underlain with mainly basalt and granite. Minimal anthropogenic influences exist in the contributing watershed aside from grazing operations and the highest flows in RCEW occur during spring snowmelt. 3 Methods While granite and basalt are both igneous rocks, I hypothesized that small difference in densities may play a strong control in the slope characteristics of Reynolds Creek tributaries. By only looking in the smaller, mountainous portion of the watershed, the drainage areas are relatively small. The elevation surface I manipulated in ArcGIS was from 2m LiDAR that I obtained directly from the Boise office of the USDA Agricultural Research Service. This LiDAR was obtained outside the snowmelt season, during low flow in November ArcGIS methods The methods that I used in ArcGIS are outlined in the flow chart in Figure 4. I selected five streams in the upper watershed to analyze, Figure 5. I focused on the upper watershed because I am doing field work in the mountains. If a portion of the stream crossed lithological boundaries, I analyzed each lithology separately. I went through each stream manually examine areas where the LiDAR swaths overlapped in these areas, the profile curvature values were incorrect. I tried to use Model Builder for this work but found that Model Builder was a little inefficient because I was selecting only a few streams and could not bring in the 3D analyst profile grapher, which I used to export elevation and profile curvature results into Excel, into Model Builder. However, I found that utilizing batch processing to run many actions with the same tool was extremely efficient and highly recommend one use this if doing similar methods. 3.2 Excel I brought the elevation and profile curvature ArcGIS data along the longitudinal stream profiles into Excel to compare plot the results. The exported elevation data consisted of two columns, one for elevation and the other for distance from the most upstream point both were in exported in meters. I plotted the elevation verses distance from most upstream point for each stream section and found the average slope with this data. Similarly, I plotted the profile curvature verses distance from most upstream point and calculated the average profile curvature value. Additionally, I calculated the cumulative frequency of the profile curvature values for each stream section to evaluate whether unique distribution patterns were apparent.

6 Figure 4 Methods applied in ArcGIS to extract elevation, slope, concavity, and profile curvature data from project streams to compare possible differences due to underlying basalt or granite lithology. Input data: Elevation Surface from 2m LiDAR (USDA ARS source) & Input data: Geologic map of area (USDA ARS source) Delineated stream network using Hydrology tools (Fig 5) Exported individual streams reaches based on lithology for comparison (Fig 6) Dissolved project stream features Created pour points for each reach Delineated upper watershed & subwatersheds of project reaches using Hydrology tools (Fig 6) Converted watershed rasters to polygons to calculate each drainage area Assigned elevation values to each dissolved stream feature by batch processing the Interpolate Shape in3d surface analysis tools Created profile curvature raster in Spatial Analyst Created longitudinal elevation profile graphs and exported the data into Excel for average slope analysis Assigned profile curvature values to each dissolved stream feature by batch processing the Interpolate Shape tool under 3D surface analysis Created longitudinal profile curvature profile graphs and exported the data into Excel for comparison (Fig 7 & 8) 4 Results and Discussion Figures for the results are presented at the end of section 5. Figure 5 shows the stream network delineation and the entire watershed delineation of Reynolds Creek. Figure 6 shows the streams I chose to analyze based on lithology and their respective subwatersheds. Table 1 summarizes the characteristics of the reaches subwatersheds as well as the ArcGIS results of the average stream slopes and profile curvatures. Only Stream 01 resulted in a strong connection to lithology where there was a visible shift in the average profile curvature where the lithology changed, Figure 7. When I presented the promising results of stream 01 to the class in my presentation, I had hoped that the trend may also appear in the other streams I analyzed but this was not the case. Because the mean profile curvatures did not seem to show a clear trend of being the same within each lithology and different between lithologies, I also examined the cumulative frequency of the profile curvature values over the stream to see if the value ranges were similar while their means may have been shifted, Figure 8. Still, no clear trends were observed across the streams to make a definitive statement regarding basalt or granite lithology controlling the profile curvatures for the streams in the upper Reynolds Creek watershed.

7 Table 1 Stream section summary of results. Number Location* Lithology Contributing Drainage Area at Reach Outlet, km 2 Best Fit Slope over Section Average Profile Curvature 1 US Basalt % DS Granite % ALL Granite % N Basalt landslide % S Basalt landslide % DS Basalt % N Basalt % S Basalt % DS Basalt % US Granite % DS Basalt % 3.73 * N North fork S South fork US Upstream sectin DS Downstream section 5 Conclusions Examining the elevation datasets of streams is often used in the first round of landscape analysis that fluvial geomorphologist conduct when examining streams. Based on the analysis of the work completed in this term project, my average cumulative profiles data does not support my hypothesis that underlying lithology controls the slopes and concavity properties of tributaries in the upper Reynolds Creek watershed. However, the spectral analysis of the oscillating signal of the positive concavities may still reveal that the distance wavelengths of the profile curvature fluctuate based on lithology. I will be doing these analysis in Spring Further, it could be possible that instead of lithology, even though all the streams are in a fairly small drainage area, it is the magnitude of the subdrainage areas that is controlling the concavities. I will also be examining this in Spring 2011 by choosing several points within each stream, delineating the drainage areas of each point, then evaluating whether these are linearly correlated on a log log plot of slope verses drainage area of all points. Finally, getting the average slopes of these streams is very valuable to me because I can now calculate theoretical basal shear stresses using Equation 2 and discharge data that the USDA has provided.

8 Figure 5 Delineated stream m network and watershed over Reynolds Creek overlaid with USDA geologic map.

9 Figure 6 Stream sections se elected based on n lithology for profile curvaturee comparison.

10 Figure 7 Shift in profile curvature at lithologic change only observed strongly for one stream, Stream 01.

11 Figure 8 Cumulative frequency of occurrence of profile curvatures over tested streams. All reaches 100% 90% 80% Cumulative Frequency 70% 60% 50% 40% 30% 20% 10% 0% 1 Basalt US 1 Granite DS 2 all granite 3 Landslide N 3 Landslide S 3 Basalt DS 4 Basalt DS 4 Basalt N 4 Basalt S 5 Basalt DS 5 Granite US Profile Curvature Cumulative Frequency 100% 90% 80% 70% 60% 50% 40% 30% All granite reaches Cumulative Frequency All basalt reaches 100% 90% 80% 70% 60% 50% 1 Basalt US 40% 3 Basalt DS 4 Basalt DS 30% 4 Basalt N 20% 4 Basalt S 10% 5 Basalt DS 5 Basalt US 0% Profile Curvature 20% 1 Granite DS 10% 2 all granite 5 Granite US 0% Profile Curvature

12 References ArcGIS, (2010), ArcGIS: How Curvature Works, ESRI Help. de Smith, M., M. Goodchild, and P. Longley (2009), Geospatial Analysis: A comprehensive guide to principles, techniques and software tools, Third., Matador, Leicester, UK. Gilbert, G. (1914), The Transportation of Debris by Running Water, Professional Paper, USGS, Sacramento, CA USA. Henson, P. (2010), Gravel Mining, [online] Available from: Kondolf, G. M., and H. Piégay (2003), Tools in fluvial geomorphology, John Wiley and Sons. Leopold, L. B. (1980), Techniques and interpretation: The sediment studies of G. K. Gilbert, Geological Society of America, (Special Paper 183), Leopold, L. B., M. G. Wolman, and J. P. Miller (1964), Fluvial processes in geomorphology, W.H. Freemand and Co. Mackin, H. (1948), Concept of the graded river, Geological Society of America Bulletin, 59(5), Quimpo, R. G. (1999), GIS modules and distributed models of the watershed: report, ASCE Publications. Roe, G. H., D. R. Montgomery, and B. Hallet (2002), Effects of orographic precipitation variations on the concavity of steady state river profiles, Geology, 30(2), Strahler, A. N. (1952), Dynamic basis of geomorphology, Geological Society of America Bulletin, 63(9),