Improving spatial and temporal monitoring of harmful algal blooms in aquatic ecosystems using unpiloted aerial vehicle based remote sensing

Cyanobacterial Harmful Algal Blooms (CyanoHABs) are increasing in frequency and intensity in freshwater systems worldwide, particularly in the Midwest, where agricultural nutrient runoff is high. Cyanobacteria can produce toxins, posing environmental, public health, and economic threats. Current monitoring strategies fail to capture the complex spatial, temporal, and toxicological dynamics of CyanoHABs. The advancement of monitoring and detection strategies is imperative for understanding and reducing CyanoHAB impacts. Multi-method approaches have been proposed, capitalizing on the strengths of several monitoring strategies, including grab sample analysis, in-situ probes, satellite, and near-range remote sensing. Grab samples allow toxin analysis at discrete locations, providing limited spatial or temporal information. Satellite systems capture spatial extents of CyanoHABs in large water systems. However, low overpass frequencies, limited spatial resolutions, and atmospheric interference reduce their utility for smaller inland lakes. There is a critical need to develop monitoring methods that provide adequate spatial and temporal resolution data. Near-range remote sensing using Unpiloted Aerial Vehicles (UAVs) offer low costs, on-demand deployments, and high-resolution imagery. UAV platforms collect multiple overlapping images during monitoring, requiring geo-positioning and mosaicking for map generation. Current photogrammetry software achieves mosaicking by matching surface features between images. This is a major limitation for over-water monitoring technologies as surface characteristics may vary substantially between images. We present a novel geometric approach for projecting and mosaicking over-water imagery without directly relying on fixed ground features. Additionally, methods were developed to improve data quality, such as the removal of glint and other noise reduction techniques. In this study, grab samples and multi-spectral imagery with 5 cm x5 cm water surface pixel resolution, collected during a 2021 case study at Lake Darling, are utilized to assess the effectiveness of UAV-based remote sensing in detecting and monitoring Chlorophyll-a concentrations, a common proxy for algal and CyanoHAB biomass.