Authors: Yang Liu*, University of Alabama, Hongxing Liu, University of Alabama, Xiaofang Wei, Central State University, Krishnakumar V. Nedunuri, Central State University, Ramani Kandiah, Central State University, Subramania I. Sritharan, Central State University
Topics: Remote Sensing, Landscape, Geographic Information Science and Systems
Keywords: remote sensing, shale gas extraction, forest fragmentation, landscape ecology metric, Muskingum Watershed
Session Type: Paper
Start / End Time: 1:10 PM / 2:50 PM
Room: Madison A, Marriott, Mezzanine Level
Presentation File: No File Uploaded
This research investigated the impacts of shale gas extraction on forest structure in the Muskingum Watershed in Appalachian Ohio. Bi-temporal high-resolution natural-color aerial images from 2006 to 2014, gas well point and horizontal pipe data were utilized to determine the landcover changes due to the recent boom of the shale gas extraction. Airborne LiDAR data acquired in 2006 were utilized to construct the initial Canopy Height Model (CHM) before the boom. The Geographic Object-Based Image Analysis (GEOBIA) was employed to extract forest patches (objects) and to characterize the forest structure. A set of landscape ecology metrics were redefined in terms of the spatial adjacency relation of forest objects to indicate and quantify forest structural changes. The analysis of landcover changes indicates that the shale gas extraction and associated forest fragmentation are concentrated in the eastern part of the Muskingum Watershed. Two hot-spots are detected, one is Carroll-Harrison counties, and the other is Belmont-Guernsey-Monroe-Noble counties. The shale gas extraction from 2006 to 2014 have resulted in 224 well pads, 116.52 km access roads, and 880.58 km pipeline right-of-way (ROW) in the Muskingum Watershed. The total area of the well pads, access roads, and their immediate surrounding disturbance zones is 1497.11 ha, the total observable disturbed area by pipeline ROWs is 1417.08 ha. These extraction activities have caused 2204.15 ha of deforestation, leading to the forest volume loss of 1.09×10^10 cubic feet. Moreover, our structural analysis shows that the forest fragmentation is statistically significant in Carroll-Harrison counties, insignificant in the other hot-spot.