Comparative use of remote sensing and in-situ data for chlorophyll-a estimation in urban lakes: a case study of Chitgar Lake, Tehran, Iran

Abstract

Chlorophyll-a (Chl-a) is a key indicator of phytoplankton biomass and freshwater quality and is strongly influenced by nutrient availability, seasonal variability and climatic conditions. This study investigates the spatiotemporal dynamics of Chl-a in Chitgar Lake, an artificial urban lake in Tehran, Iran, by integrating long-term in-situ observations (2013–2025) with satellite-derived estimates from Sentinel-2 and Landsat-8. Seasonal analysis revealed that Chl-a concentrations were highest during summer (mean: 2.93 mg/m³), peaking in August (3.53 mg/m³), coinciding with elevated water temperatures (up to 30.8°C) and reduced dissolved oxygen (mean: 8.35 mg/L), indicative of eutrophic conditions. In contrast, winter and spring exhibited lower Chl-a levels associated with cooler temperatures and more stable oxygen conditions. The mean total phosphorus concentration (0.03 mg/L) suggests phosphorus limitation, consistent with international eutrophication thresholds. Four satellite-based algorithms – Two-Band Ratio, Normalized Difference Chlorophyll Index (NDCI), Three-Band Difference Algorithm (3BDA), and Maximum Chlorophyll Index (MCI) – were evaluated using S2 and L8 imagery. Sentinel-2 consistently outperformed Landsat-8, with NDCI achieving RMSE of 1.23 mg/m³, MAE of 0.93 mg/m³, and R² of 0.68. Seasonal validation identified fall as the most reliable period for Chl-a estimation (RMSE: 0.88–1.03 mg/m³), while winter showed the lowest errors but limited explanatory power (R²: 0.23–0.28). Landsat-8 yielded acceptable performance, particularly in spring (R²: 0.64–0.84), despite its coarser spatial resolution. Spatial Chl-a maps derived from NDCI revealed consistent distribution patterns between satellites and strong agreement with in-situ measurements. However, Sentinel-2’s finer spatial resolution (10–20 m) enabled improved detection of small-scale variability. Overall, the results highlight the value of Sentinel-2 red-edge bands for Chl-a retrieval in optically complex inland waters, while confirming the continued importance of Landsat-8 for long-term trend analysis. The findings emphasize the need for local calibration and support integrated satellite – in-situ monitoring frameworks for improved eutrophication management in urban lakes.