We used a map inundation methodology based on geomorphometric principles (Bolch et al., 2011; Clubb et al., 2017; Sofia et al., 2014). The basic premise is to let topography dictate how water will fill a particular landscape during a flood event. It allows high‐speed computation and accurate identification of locations that are likely to be inundated after heavy precipitation. The objective is to estimate the inundation depths and spatially capture the locations likely to be flooded due to any known amount of rainfall. For this purpose, a variable of x inches of rain is placed in each of the digital elevation model cells to obtain the first estimate of water surface elevation. Next, the water surface elevations are distributed/balanced over the entire terrain using geomorphological principles such that the point with lower elevation would have greater flood depth and vice versa. Once the steady state is reached, the water surface elevations are determined by the corresponding digital elevations to obtain the inundation depths. The model was initially validated for Hurricane Harvey when it hit the Houston, TX shores in 2018.

In addition to the selection of flooded and unflooded water bodies, we also obtained the swine farm locations from the North Carolina Department of Environmental Quality (NCDEQ) Animal Feeding Operations Program (NC DEQ, 2019). The 2,283 permitted farm locations were overlaid onto the flood map described above, then queried by flood depth. All swine farms with flood depths of 5 ft or greater were selected as potential sampling sites because there was high certainty about flooding at the location, resulting in the selection of 40 swine farms. Based on our analysis of actual flood report by NCDEQ, all 40 swine farms (100%) were indeed flooded. Twenty‐four swine facilities reported lagoon discharging, an additional eight reported inundation (surface water surrounding and flowing into lagoon), and eight more reported lagoons at full capacity and likely to overtop (NC DEQ, 2018). Similarly, unflooded farms were identified. Twenty‐three farm locations were selected from the list of unflooded farms by visually inspecting the flood map. Those unflooded farms met two criteria. First, the unflooded location must be relatively close to flooded farms of interest, making travel and collection less burdensome. Second, the flood maps must show the location as free of flooding to a high certainty (i.e., not located on or near the boundary of the flood extent). None of the selected unflooded farms were reported flooded in the post flood report. From these modeling and post flood analysis, 25 sites were selected and sampled in North Carolina on October 7, 2018 (3 weeks posthurricane), ranging geographically from the coast to about 100 miles inland. This included 16 flooded and nine unflooded locations (Table S1). While in the field, water samples were taken at publicly accessible locations downstream and as near to the selected farms as possible.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.