American Journal for Young Scientists · ISSN 3068-9015
Can low-cost autonomous drones detect early algal blooms as accurately as satellites?
Authors: Srujan Pininty
Published: October 2, 2026
Category: Environmental Science
Abstract
Harmful algal blooms can make drinking water and recreational water unsafe for people and wildlife. Satellites can monitor large areas, but their images may miss small blooms near shore. This project asks whether a low-cost autonomous Crazyflie drone equipped with an AI Deck can detect early algal blooms as accurately as satellites. The drone was built as a proof of concept for autonomous, close-range image collection and processing. To create a satellite benchmark, California bloom records and satellite products were combined with field measurements from the Water Quality Portal. The field data included 5,160 chlorophyll-a and 2,864 phycocyanin measurements which indicate algae growth, and 1,630 microcystin results, which is a dangerous toxin produced by some blooms from California lakes and reservoirs. After carefully matching measurements taken from the same lake on the same date, the satel- lite data showed a moderate correlations with field chlorophyll (ρ = 0.53) and phycocyanin (ρ = 0.65), but a weaker correlation with microcystin (ρ = 0.22). These findings indicate that satellite and drone images may help detect where algae are growing but cannot deter- mine whether the water contains dangerous toxin levels. Most matched observations came from Clear Lake, so more testing is needed in other California lakes. The results support using drones for detailed local imaging, and laboratory samples for toxin confirmation. Code availability: https://github.com/s1010101111101/algae-drone.