Browsing by Subject "Erosional hotspots"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Identifying Erosional Hotspots in Duluth-Area Streams after the 2012 Flood Using High-Resolution Aerial Lidar Data
    (2015-08) Manopkawee, Pichawut
    During June 19th to 20th, 2012, northeastern Minnesota experienced a high record rainfall event, causing soil saturation and inundation, slope failures, flash flooding, and damage to public infrastructures. Geomorphic effects of the flood included severe streambank and bluff erosion and landslides. The purposes of this study are to investigate erosional hotspots and channel reaches in Duluth-area streams that experienced significant geomorphic changes as a result of the 2012 flood and to determine the roles of human modifications to the stream networks on erosional hotspots at a reach scale. Erosion occurred when driving forces, controlled by topography and precipitation, overcame resisting forces, controlled by shear strength of materials at the bed and banks. One-meter resolution lidar data, which were collected before and after the flood, were used to extract driving forces, predict erosional hotspots, and map out locations of significant geomorphic change in a GIS framework. Lidar data were first filtered in GeoNet to remove noise in low-gradient areas and enhance geomorphic features. The lidar-derived DEMs were then used to calculate a stream power-based erosion index (SP) and angle of impingement (AOI), to identify stream reaches with high bend curvature, and to identify stream reaches proximal to high bluffs. Bedrock exposure locations, which could significantly limit erodibility, were obtained from Minnesota Geological Survey maps (Hobbs, 2009a, b, c). These parameters were used to predict preliminary erosional hotspot locations. Field observations were done to verify the results of predicted erosional hotspots from the GIS-based predictive model and develop a new threshold model. The refined predicted erosional hotspots were classified into different types and were compared with valley types and channel-reach types to determine how susceptible different channel-reach types were to change. The preliminary GIS-based predictive model had low accuracy of prediction due to misidentification of many erosional sites along the streams, compared to erosional sites from field observations. The refined threshold model improved the percent of accuracy for all points and for FEI >= 2 to greater than 80%, with less than 10% of points over- and under-predicted on three sample streams. The Minnesota Geological Survey (MGS) bedrock exposure maps were verified and improved by field maps, resulting in 90-95% accuracy in new bedrock exposure locations. In terms of types of erosional hotspots, more than 70% of erosional hotspots in the northern and the southern ends of Duluth are located in pool-riffle reaches within entrenched valleys; these hotspots are primarily classified as topographic erosional hotspots, which likely had significant geomorphic changes due to topography, substrate materials, and planform geometry. In the central area of Duluth, more than 70% of erosional hotspots are located in pool-riffle reaches within confined valleys; these hotspots are classified as topographic/anthropogenic erosional hotspots due to being located in the areas of significant anthropogenic influences. Pool-riffle reaches are identified as the most susceptible channel-reach type to change from the 2012 flood. The GIS-based predictive model could locate erosional hotspots and susceptible reaches which were strongly affected by the 2012 flood. It also incorporates anthropogenic effects to describe the influences of human construction causing erosion in a particular area. The project offers the City of Duluth useful data on causes of the channel changes in the particular reaches, potentially helping with future stream restoration efforts.