Using Multitemporal Imagery to Improve Mapping and Inventory of Forested Roads in Michigan’s Upper Peninsula

Abstract

Although a road network provides transportation for people and goods, several negative ecological impacts stem from a road network. Significant effects of concern for rural roads include an increase in the number of road–stream crossings, which can potentially cause fish passage problems, in-stream habitat impacts, increased stream sedimentation, and an increase in potentially affected wetlands through hydrologic alteration. Although several studies have examined and quantified the effects that roads have on terrestrial and aquatic ecosystems, many studies may have been made on the basis of incomplete road network data sets. An updated and improved road network data set is needed before researchers can draw inferences that quantify the total ecological impact that a road network has on an ecosystem. The purpose of this work was to develop an updated and improved regional road network data layer through interpretation of existing multitemporal aerial imagery, with the resulting geographic information system layer being applied in ecological analyses and planning. Results from two watersheds in Michigan’s Upper Peninsula indicated that each watershed experienced an approximate increase of 200% in total road mileage when compared with the results from the original road network data set. Individual hydrological unit code level-12 subwatersheds within each watershed experienced an increase ranging from 16% to 2,700%. The number of road–stream crossings for each watershed also increased by at least 100%. The updated road network data set also reported the location and total area of wetlands that have potential hydrological impacts, ranging from 70,000 ha to 121,000 ha for each watershed.

Predicitng Failure of Culverts and Associated Impacts in Low Order Streams of Northern Michigan

“…Ground surveys led to the confirmation of 49 potential barriers, which include 20 culverts, 19 bridges, 8 dams, and 2 fords, and exclude removed crossings (Figure 3.1). This is nine less that the MTRI analysis for barriers in the watershed (Banach et al, 2016). Seventy-five percent of inventoried culverts occurred on first-order streams in the North Branch.…”

“…An assessment using satellite and near surface remote imagery analysis determined the North Branch to have a road density of 5.2 miles of road per square mile, with a total of 605 miles of road in the watershed (Banach et al, 2016). These data identified 58 potential fish passage barriers in the North Branch, including dams, culverts, bridges, fords, footbridges, and removed crossings (Banach et al, 2016). This number was greater than the number of crossings detected by the intersection of the total flow line with the State of Michigan's "All Roads"…”

“…where T is the annual average daily traffic, which is estimated at 10 cars for forest roads in the Ottawa National Forest. H is the amount of delay in hours, and we assumed one hour of delay time based on the approximate speed and distance needed to detour around any single failed crossing (Banach et al, 2016) Failed culverts would contribute sediment from the road prism to the stream channel, and this material would need to be replaced to re-open the crossing and would also need to be removed from the stream to return it to its pre-failure condition. The cost of fill replacement (CG) was determined from estimating the total volume of fill that was mobilized from the road prism during failure.…”