UAV-based hyperspectral imaging for river algae pigment estimation and development of a low-cost multispectral imager

Abstract

Harmful and nuisance algal blooms are becoming a greater concern to public health, riparian ecosystems, and recreational uses of inland waterways. Algal bloom proliferation has increased in the Upper Clark Fork River in western Montana, USA, due to a combination of warming water temperatures, naturally high phosphorus levels, and an influx of contaminants through anthropogenic nitrogen enrichment along its banks. To improve understanding of bloom dynamics, such as algal biomass, a UAV-based hyperspectral imaging system was deployed to monitor several locations along the Upper Clark Fork River. Image data were collected across the spectral range of 400 - 1000 nm with 2.1 nm spectral resolution during two field sampling campaigns in 2021. Included are methods to estimate chlorophyll a standing crops using regression analysis of salient wavelength bands, before and after separating the pigments according to growth form. Estimates of total chlorophyll a standing crops generated through a brute-force analysis are compared to in-situ data, resulting in a maximum rsquared of 0.62 for estimating filamentous plus epiphytic chlorophyll a. Estimates of total and epilithic pigment standing crops are also included. The salient wavelengths bands used to estimate these pigments were then used as the basis for creating a low-cost imaging system for identifying algal blooms.

Developing a low-cost multispectral imager for detecting algal blooms in rivers

“…Our initial hyperspectral algorithms for estimating chl a and phycocyanin content from this imaging system use band ratios in which systematic errors tend to cancel. 44 In the future, these band ratios will be used with the multispectral imaging system as well to estimate chl a and phycocyanin content while minimizing systematic errors.…”

“…Further discussion of the deployment of this drone-based hyperspectral imaging system on the Upper Clark Fork River for river algae pigment estimation was presented elsewhere. 44 Each spectral band was implemented in the multispectral imager design with a physical bandpass filter that can be easily replaced if different wavelengths are determined to be useful (Table 1). Each bandpass filter was placed over one of four separate Raspberry Pi V2 NoIR cameras, where the NoIR specification indicates that the cameras do not have an infrared-blocking filter, extending the spectral response of the cameras to 400 -1000 nm.…”