Investigating UAS multispectral imagery for total suspended solids and turbidity monitoring in small streams

Authors: PRIOR, ELIZABETH - Auburn University; O'DONNELL, FRANCES - Auburn University; BRODBECK, CHRISTIAN - Auburn University; Runion, George; SHEPHERD, STEPHANIE - Auburn University

Submitted to: International Journal of Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/24/2020
Publication Date: 8/24/2020
Citation: Prior, E.M., O'Donnell, F., Brodbeck, C., Runion, G.B., Shepherd, S.L. 2020. Investigating UAS multispectral imagery for total suspended solids and turbidity monitoring in small streams. International Journal of Remote Sensing. 42(1):39-64. https://doi.org/10.1080/01431161.2020.1798546.
DOI: https://doi.org/10.1080/01431161.2020.1798546

Interpretive Summary: Ease of use, ability to carry sensors and cameras, low cost, and precise maneuverability makes unmanned aerial systems (UAS) a versatile tool for outdoor assessments. This study used UAS imagery to measure turbidity and total suspended solids (TSS) of small streams. Imagery and water samples were collected before and after rain from a recently restored portion of Moores Creek in Lanett, AL. Water TSS and turbidity were related to imagery pixel values to create linear regression models. Models were then tested for Moores Mill Creek (Chewacla State Park, AL). For Lanett, TSS and turbidity regression models were good predictors for low stream flow. Results after rain suggest that additional models may be needed for high stream flow events. When the Lanett models were applied to Chewacla State Park, predicted TSS and turbidity were not comparable to measured values, which indicate that location-specific models may be required. Future research should include stream depth since streambed visibility could have influenced results.

Technical Abstract: Unmanned aerial systems (UAS) are increasingly used for field data collection and remote sensing purposes. Their ease of use, ability to carry sensors and cameras, low cost, precise maneuverability and navigation makes them a versatile tool. The goal of this study is to investigate if multispectral imagery obtained from an UAS can be utilized to measure turbidity and total suspended solids (TSS) of small streams. UAS near infrared imagery and water samples at varying depths were collected before and after rain events from a recently restored reach of Moores Creek in Lanett, Alabama. The water samples were processed for TSS and turbidity and related to pixel values from the multispectral imagery. Linear regression was used to develop models for TSS and turbidity. The models were then tested on Moores Mill Creek in Chewacla State Park, AL, USA. For Lanett, TSS and turbidity regression models for low flow had r2 values of 0.77 and 0.78, respectively. When sampling after rain, different single bands and band ratios were required for sufficient r2 values, suggesting separate models may be needed for high and low flow events. When the Lanett models were applied to Chewacla State Park, predicted TSS and turbidity were not comparable to measured values indicating location-specific models may be required. Future research should incorporate depth as a variable since streambed visibility likely impacts results.