Monitoring temporal chlorophyll-a using Sentinel-2 imagery in urban retention ponds receiving a biological-chemical treatment

Abstract

Eutrophic conditions occur when waterbodies accumulate excessive nutrients, which leads to algae growth, depleted dissolved oxygen, and increased turbidity. To improve water quality, a biological-chemical treatment was implemented using floating treatment wetlands (FTWs) and slow-release lanthanum delivered through airlift pumps. FTWs remove nutrients via plant uptake and denitrification while lanthanum binds and precipitates phosphate from the water column. The study sites, Densmore and Wilderness Ridge Ponds, had two FTWs and two lanthanum airlift pumps installed in 2020 and 2021. Our hypothesis was that if the biological-chemical treatment reduced nitrate and phosphate, algae growth would be limited, water clarity would improve, and benthic vegetation would increase. Chlorophyll-a, the photosynthetic pigment found in plants and algae, was analyzed to create a direct correlation to algae content within waterbodies. Chlorophyll-a samples were collected twice a month at strategically selected, georeferenced locations in both ponds between May and October in 2022. Water samples were collected when the European Union's Sentinel-2 multispectral satellite passed over the study sites. To estimate chlorophyll-a concentrations prior to the biological-chemical treatment, a correlation was developed between the collected chlorophyll-a concentrations and Sentinel-2 band algorithms. Sentinel-2 imagery provided a means to analyze how chlorophyll-a fluctuated before and during treatment. We found the novel band algorithm, 3BR1, had the best correlation (R2 = 0.72). Applying this band ratio for Sentinel-2 images from 2019 to 2022, a Tukey's HSD test analyzed monthly differences between years. The FTW and lanthanum treatment reduced chlorophyll-a concentrations in Densmore Pond, but not in Wilderness Ridge Pond. One likely reason was Wilderness Ridge Pond had a complex irrigation system and was continuously replenished with nitrate-rich groundwater. Our results show Sentinel-2 can be used to predict historical chlorophyll-a concentrations by using the novel 3BR1 band algorithm and that FTWs and lanthanum can effectively reduce algal growth.