Skip to main content
NCBI home page
Springer logoLink to Springer
. 2026 Oct 7;198(11):1149. doi: 10.1007/s10661-026-16007-4

Tampa Bay observing network: a continuous data acquisition system in support of science and environmental management

Bo Yang 1,2, Yonggang Liu 1,✉, Jason A Law 1, Robert H Weisberg 1, Samantha E D’Angelo 1, Kaili Qiao 1, Marcus W Beck 3, Edward T Sherwood 3, Thomas K Frazer 1
PMCID: PMC13645953  PMID: 42842094

Abstract

The Tampa Bay estuary is a vital ecological and economic hub, whose productive, shallow-water habitats are highly vulnerable to both chronic anthropogenic stress and acute environmental disturbance events (e.g., extreme weather, human incidents). Historically, the bay’s health was monitored primarily through low-frequency, monthly field sampling that has provided a foundational baseline of water quality for the last fifty years. However, discrete sampling is inadequate for capturing the rapid, high-frequency processes that govern the bay’s physics and biogeochemistry. Recent perturbations to the Tampa Bay system have exposed this shortcoming and illustrated clearly the need for an autonomous, continuous data acquisition system. This paper introduces the Tampa Bay observing network (TBON), a system designed for high-frequency environmental data acquisition. In addition to detailing its design and operation, case studies are reported to demonstrate its capacity to: (1) supplement traditional low-frequency field sampling, (2) bridge data gaps, and (3) provide continuous monitoring during acute disturbance events when field sampling is impractical. This work is also aimed at informing resource managers and stakeholders about TBON's capabilities and data availability, thereby facilitating its utilization for collaborative research, environmental protection, and effective policy formulation.

Keywords: Water quality, Autonomous sensor, Continuous monitoring, Real-time monitoring

Introduction

The Tampa Bay estuary, encompassing approximately 400 square miles, is Florida’s largest open-water estuary and a vital component of the state’s coastal economy. The bay's characteristic shallow, well-mixed waters (e.g., Weisberg & Zheng, 2006; Zhu et al., 2015a, b; Chen et al., 2023b) and extensive habitats, including tidal mud flats, mangrove-dominated wetlands, and seagrass meadows, support a biologically diverse and productive coastal ecosystem, providing critical nursery grounds for numerous invertebrates, fishes, manatees, dolphins, and other wildlife (Sherwood et al., 2017; Yates et al., 2011). Despite decades of intensive conservation and restoration efforts that have significantly improved water quality (e.g., Beck et al., 2019; Greening et al., 2014), the bay remains highly vulnerable to a broad suite of anthropogenic stressors that originate in a highly populated and agriculturally intensive watershed (e.g., Xian et al., 2007). The biggest challenges include excessive nutrient loading, which results in excessive algal growth and subsequent loss of critical seagrass habitat (e.g., Greening & Janicki, 2006), an increasing occurrence and intensity of harmful algal blooms (HABs, e.g., Lopez et al., 2023; Yentsch et al., 2008), as well as physical disturbances due to dredging (historical and ongoing) and shoreline modification (e.g., Goodwin, 1987).

For many years, the water quality of Tampa Bay has been monitored through monthly field sampling at fixed locations (e.g., Karlen et al., 2023). This low-frequency sampling regime, while essential for building long-term environmental records, fails to capture the more rapid, higher-frequency events and associated processes that govern the estuary's physics and biogeochemistry. These processes include tides, regular wind-forcing, episodic tropical storms, and pulsed runoff from extreme rainfall or anthropogenic incidents within the watershed. The vulnerability of the Tampa Bay ecosystem to such events has been starkly demonstrated by several recent perturbations to the system. A structural failure at a former phosphate processing facility at Piney Point in March 2021 resulted in the discharge of 215 million gallons of nutrient-rich wastewater into the bay, delivering a massive load of inorganic nitrogen and other contaminants to the system. This wastewater release was strongly implicated in the maintenance of an intense Karenia brevis bloom that caused widespread fish kills (Beck et al., 2022, 2023; Morrison et al., 2023; Liu et al., 2024). In this instance, a multi-site monitoring network providing continuous observations of nutrient concentration dynamics before, during, and after the incident would have been invaluable for tracking, managing, and mitigating the crisis—insights that low-frequency routine sampling or post-incident emergency field sampling simply cannot provide. Likewise, the bay is also threatened by extreme weather events such as tropical storms and hurricanes. Following Hurricane Irma (2017), the negative storm surge (Chen et al., 2018) and a large influx of freshwater significantly lowered bay salinity and increased pollutant loads, contributing to poor water quality in subsequent months (Karlen et al., 2023). Real-time, sensor-based observing networks provide the continuous data stream necessary to evaluate environmental responses during such severe weather, filling critical gaps left by shipboard sampling.

The need for continuous, high-resolution environmental data is not unique to Tampa Bay. In fact, moored and buoy-based monitoring networks have been deployed across the U.S. to provide real-time observations since the early 2000s. While these systems initially focused on fundamental physical variables like temperature and salinity, they have since expanded to include meteorological and biogeochemical variables as due to advancements in sensor technologies. Notable examples of such networks include the NOAA Chesapeake Bay interpretive buoy system (CBIBS, https://buoybay.noaa.gov/), the National Estuarine Research Reserve System’s system-wide monitoring program (NERRS SWMP, https://coast.noaa.gov/nerrs/), the coastal margin observation & prediction (CMOP) network for Columbia River estuary (https://cmop.critfc.org/datamart/observation-network/), and the San Francisco Estuary Institute’s (SFEI) moored sensor monitoring program (https://www.sfei.org/projects/moored-sensor-monitoring-program). In Florida, similar systems have been deployed in the Caloosahatchee River estuary (https://recon.sccf.org/) and the Indian River Lagoon (https://irlon.org/). Estuarine observing networks have also been established outside the U.S., such as in the Gironde Estuary, France (water quality; Etcheber et al., 2011), the Guadalquivir Estuary, Spain (weather, ocean currents, and water quality; Navarro et al., 2011), and the Venice Lagoon, Italy (tides and weather; Ferla et al., 2007). Data from these observing networks have been used in various applications, including routine water quality monitoring (e.g., Hanisak & Davis, 2018), nutrient load assessment (e.g., Gilbert et al., 2013), hypoxia monitoring (e.g., Tango et al., 2025), current and wave measurements (e.g., Wilson & Siegel, 2011), light attenuation studies (e.g., Milbrandt et al., 2016), and coastal acidification monitoring (e.g., Rosenau et al., 2021). Early versions of these systems stored data locally and relied on manual data acquisition during routine maintenance visits. Later, with advancements in telemetry technology, real-time data transmission became available via satellite or mobile networks. Meanwhile, efforts have been made to fully understand and utilize the large datasets from these networks, including the application of mathematical/statistical methods (e.g., Murphy et al., 2010), the development of software packages for QA/QC and data analysis (e.g., Beck et al., 2016), the creation of new data frameworks for visualization (e.g., Zhang et al., 2025), and integration with numerical modeling (e.g., Cerco & Noel, 2013) and machine learning approaches (e.g., Ahmed et al., 2019).

Building upon decades of experience in deploying, maintaining, and utilizing autonomous real-time sensor networks, the Ocean circulation lab (OCL) at the University of South Florida (USF) established the Tampa Bay observing network (TBON) as a direct response to the 2021 Piney Point incident. TBON is an autonomous monitoring system designed to bridge the temporal gaps in data collection inherent in discrete sampling programs. Distinct from observing systems in other estuaries, TBON is characterized by three primary innovations: (1) it represents the first high-frequency, real-time, interdisciplinary monitoring infrastructure dedicated specifically to Tampa Bay—a unique environment shaped by the intersection of urbanism, industry, agriculture, and extreme weather events including major hurricanes; (2) for most sites, a comprehensive sensor package is installed on existing US Coast Guard (USCG) range towers to reduce construction costs and avoid ship traffic hazards. Featuring a top-mounted weather station, surface-mounted water quality sensor, and bottom-mounted current profiler, this setup delivers a complete suite of co-located metrological, water quality, and physical oceanography data via cellular telemetry—a combination rarely offered by other monitoring networks; and (3) it is fully integrated with the operational Tampa Bay coastal ocean model (TBCOM; Chen et al., 2023b), enabling the real-time synthesis of observational data (currently covering meteorological and physical oceanographic data) and numerical forecasting.

Herein, we outline TBON’s technical design, sensor suite, and data architecture, accompanied by practical case studies. While this paper aims to inform local researchers, stakeholders, managers, educators, and the general public about data availability and application, we also present it as an adaptable model for global efforts. Rather than comparing TBON to other networks or claiming superiority, we offer these commonplace remarks to initiate open dialogue and provide a useful reference for others building their own coastal monitoring systems and data portals.

Methods

Since 2023, eight TBON stations have been established throughout Tampa Bay (Fig. 1, Table 1), with one (TB08) in Hillsborough Bay (HB), three (TB01, TB02, and TB05) in middle Tampa Bay (MTB), and four (TB03, TB04, TB06, and TB07) in lower Tampa Bay (LTB). An additional station (TB09) is scheduled for installation in Old Tampa Bay (OTB) in 2026. The typical TBON station setup (Fig. 2) utilizes existing USCG navigational range towers, which is cost-effective and avoids issues with marine traffic. Meteorological instruments (MET) and data transmission systems are installed on top of each tower, water quality sensors (WQ) are attached to the underwater tower section near the surface (~ 1–2 m depth), and acoustic Doppler current profilers (ADCP) are installed on the seafloor adjacent to the tower. Table 1 provides station coordinates, lists of instrumentation deployed, and the period of operation. In terms of instrumentation, the water quality sensor (YSI EXO2) measures temperature, conductivity (salinity), pressure (depth), dissolved oxygen, pH, turbidity, and chlorophyll-a, with a frequency of every 15 min. Fluorescent dissolved organic matter (FDOM) sensors were added beginning in 2024, providing proxies for dissolved organic carbon. The meteorological instrument (Gill GMX500) measures wind speed and direction, air temperature, humidity, air pressure, and precipitation (only available at TB05 and TB08), with a sampling interval of 30 min. The ADCP (Nortek Signature 1000) provides non-real time current velocity. Two departures from this standard setup are TB05 and TB06. TB05 is installed on the seawall along Bayboro Harbor in MTB, and TB06 only has an ADCP installed on the seafloor. Six stations are currently in service. Stations TB02 and TB03 were discontinued following hurricane damage to the USCG range towers, though their sensors were redeployed to establish new sites.

Fig. 1.

Fig. 1

Open in a new tab

Map of the Tampa Bay Observing Network (TBON) showing active, planned, and discontinued stations. TB06 consists solely of a bottom-mounted ADCP and is not surface-visible. The bottom-right inset shows the geographic position of Tampa Bay relative to the USA main land

Table 1.

Tampa Bay observing network (TBON) station information

Station Latitude Longitude Instruments Operational Period
TB01 27°45.042' N 82°34.326' W WQ, MET, ADCP 05/15/2023—Present
TB02 27°42.978' N 82°31.341' W WQ, MET, ADCP 05/31/2023—01/30/2025
TB03 27°39.795' N 82°37.182' W WQ, MET, ADCP 05/22/2023—11/12/2024
TB04 27°38.863' N 82°35.803' W WQ, MET, ADCP 01/13/2024—Present
TB05 27°45.600' N 82°37.980' W WQ, MET, ADCP 10/19/2023—Present
TB06 27°35.502' N 82°41.202' W ADCP 06/28/2023—Present
TB07 27°36.234' N 82°33.384' W WQ 07/10/2025—Present
TB08 27°53.856' N 82°26.634' W WQ, MET 06/20/2025—Present
TB09 27°56.700' N 82°42.000' W WQ, MET, ADCP Planned for 2026
Open in a new tab

Fig. 2.

Fig. 2

Open in a new tab

Schematic of a typical TBON station setup (except for TB05 & TB06). MET: autonomous weather station, WQ: water quality sensor, ADCP: Acoustic Doppler Current Profiler

For real-time quality assurance/quality control (QA/QC), TBON follows guidelines established as part of the NOAA Integrated Ocean Observing System (IOOS) quality assurance of real time oceanographic data (QARTOD) project (https://ioos.noaa.gov/project/qartod/). The real-time data are subject to automated tests for gross range, spikes/outliers, and rate of change. Based on these test results, quality flags are assigned to each data point.

Results

A dedicated web portal designed for TBON provides background information, displays real-time data, and allows access to archived data. Users may choose a specific TBON station at the main page (https://tbon.marine.usf.edu, Fig. 3a) to view location coordinates, background information, and a series of snapshots of the water quality or meteorology variables. For example, Fig. 3a shows water quality data from station TB08, as selected under the "Water Quality" panel. Subsequently, the station location is highlighted on the map, a table with the latest values of all available water quality variables appears to the right of the map, and a photo of the station is displayed on the far right. Below the map, several time-series plots for each water quality variable are shown. All the plots are interactive, featuring tooltips that display corresponding values when the user hovers over specific data points. Options under the "Report Period" button allow users to display data from the past 3 days, 1 week, 1 month, or 1 year. The "Quality Filter" button is used to choose whether to display all raw data or data that have been processed through the real-time automatic QA/QC (see the Methods section for details). Finally, the "Units" button allows users to choose between imperial and metric units. Similarly, meteorological data for each station can be selected under the "Weather Stations" panel to display the corresponding time series.

Fig. 3.

Fig. 3

Open in a new tab

Web portal for Tampa Bay observing network (TBON). (a) Main page: https://tbon.marine.usf.edu; (b) Data query builder: https://tbon.marine.usf.edu/graph.php

The data query builder (https://tbon.marine.usf.edu/graph.php, Fig. 3b) serves as the main interface for accessing the TBON data. A navigation bar allows users to set several functions: choosing a specific time period ("Date Range"), selecting the time display format (local time or UTC, under "Time Zone"), setting the units (metric or imperial, under "Units"), and determining if the automatic QA/QC should be applied ("Quality Filter"). Then, using the drop-down menu under the "Available Parameters" panel, users can select all the desired water quality or meteorological variables for each TBON station. Once all options are selected, clicking the blue "Go" button (located to the far right of the drop-down menu) will display an interactive plot containing all chosen variables for the selected period at the top of the page. Links are provided on top of the plot to download the results as an image (PNG or PDF format) or as a dataset (TXT, CSV, or JSON format). After downloading the image/data, the "Reset" button (located next to the drop-down menu) may be clicked to start a new data query.

The user-friendly portal with its interactive data visualization capacity provides an effective way to deliver high-frequency, real-time Tampa Bay environmental data to the general public. By graphically presenting the data rather than simply dumping numbers, the portal is especially useful for education and outreach. Simultaneously, the query builder enables professionals, including resource managers and emergency response personnel, to efficiently extract data needed to support informed decision-making and incident response.

Discussion

Supplementing low-frequency field sampling

Autonomous sensor-based high-frequency (sub-hourly) data streams from TBON supplement a long-term, monthly field sampling effort by providing greater temporal resolution at strategic locations. A case study is presented using salinity data from TBON station TB01 and Hillsborough County’s Environmental Protection Commission (EPC) field sampling station 82 at the same location. The box and whisker plot (Fig. 4a) shows the monthly distribution of salinity data at this location from January to September 2024. For each month, the box represents the middle 50% of the data (interquartile range, IQR), the line inside indicates the median value, and the upper/lower whiskers shows the highest/lowest non-outlier values. The red dots show the salinity values from the monthly field sampling event, with the corresponding sample dates denoted at the bottom (red text). This figure illustrates that discrete data derived from a monthly sampling regime often fall at the edge of the IQR (February, March, April, June, and July), or fall outside the IQR entirely (May, September), occasionally even representing outliers (January). In fact, Weisberg and Zheng (2006) found it necessary to average discrete samples over three months to arrive at a representative salinity field for comparison with their Tampa Bay numerical model simulations. In contrast, the enhanced temporal resolution provided by TBON offers a more comprehensive representation of salinity in the Bay. Additionally, due to missing timestamps for some discrete samples, comparisons were limited to TB01 daily means (Fig. 4b). Agreement between datasets occurred only when field collection happened to coincide with the daily average (Fig. 4c). Deviations, on the other hand, are common and Fig. 4d illustrates the bias introduced by sampling during diurnal minima. This example demonstrates clearly the inherent limitations of a low-frequency field sampling regime in a highly dynamic estuarine environment. It further demonstrates the value of real-time autonomous measurements that more fully capture temporal variation at continuous and diurnal time scales.

Fig. 4.

Fig. 4

Open in a new tab

(a) Box and whisker plots of station TB01 salinity measurements from January to September of 2024. The red dots indicate the results from monthly field sampling, with the sampling dates of each month displayed in red text. (b) Comparison of salinity from TB01 and field samples. (c, d) Blow-out views showing diurnal variations of salinity measured by the TB01 mooring (blue lines) on February 20, 2024, and July 16, 2024. The red dashed lines in panels (c) and (d) indicate the salinity values obtained from field sampling (with no specific sampling time of the day recorded)

Bridging the gaps in existing monitoring approaches

Due to its extended temporal and spatial coverage, satellite remote sensing has been widely used as a complementary approach to low-frequency field sampling, especially for variables like turbidity and water clarity (Chen et al., 2007). Chlorophyll-a concentration (Chl-a) is often used as a proxy for phytoplankton density or biomass, making it a critical water quality variable for assessing primary production, detecting the onset of HABs, and serving as an indicator of eutrophication. However, the application of satellite-based Chl-a monitoring is limited in coastal and estuarine environments like Tampa Bay because of the optical complexity of these waters and bottom reflectance caused by shallow depths (e.g., Le et al., 2013).

As a second example of TBON's utility, we examine Chl-a variability from January to September 2024 at station TB01. We compared TB01 autonomous sensor data with data generated from field samples at the co-located EPC sampling station and satellite-derived data (Fig. 5). The NASA Moderate Resolution Imaging Spectroradiometer (MODIS) Chl-a product is a state-of-the-art, widely used satellite dataset for marine environmental monitoring. We retrieved the standard 8-day, 4-km resolution Chl-a product directly from the NASA Ocean Color data repository without additional processing (https://oceandata.sci.gsfc.nasa.gov/l3/). Full technical details regarding the MODIS Chl-a product algorithms and validation are available at https://modis.gsfc.nasa.gov/data/dataprod/chlor_a.php. Field samples were chemically extracted for chlorophyll-a, and its concentration was then measured using a lab-based fluorometer. Similarly, the TBON sensor estimates Chl-a by measuring in vivo fluorescence and converting it to concentration using sensor-specific algorithms. In contrast, satellite remote sensing estimates Chl-a by calculating the ratio of blue-to-green light reflected from the surface to determine light absorption by phytoplankton (Hu et al., 2012, 2019). Since all three approaches provide results in the same concentration units, and both satellite and sensor estimates are ultimately validated against field samples, their values are considered comparable.

Fig. 5.

Fig. 5

Open in a new tab

Comparison of chlorophyll-a concentrations (Chl-a, µg L−1) during TB01 2024 deployment. Red circle: measurements from field samples taken by the Environmental Protection Commission (EPC) of Hillsborough County monthly field sampling station 82 (location overlapped with TBON station TB01). Yellow line: estimates from NASA MODIS 8-day satellite product. Blue line: sensor data from station TB01 (shown as daily mean)

Chl-a from monthly field sampling (red circle in Fig. 5) shows little variation from January 30 to May 22, with all values below 2 μg L−1. Chl-a, then increased from June 25 (2.6 μg L−1) to July 16 (6.2 μg L−1), before peaking (20.5 μg L−1) on August 21. Chl-a estimates from satellite remote sensing (yellow line with cross mark in Fig. 5) fail to reproduce either the magnitude or the general temporal trend of Chl-a revealed by field samples, which demonstrates a limitation of the satellite remote sensing approach in estuarine and coastal waters. In contrast, a daily mean Chl-a (blue line with dots in Fig. 5) time series derived from TBON sensor data shows good agreement with the monthly field samples, while offering much better temporal resolution for assessing Chl-a variation within this period. Overall, the TBON data successfully fill the gap between monthly field sampling and the satellite remote sensing for monitoring Chl-a.

Providing continuous monitoring in response to acute disturbance events

Autonomous sensors from TBON can provide continuous meteorological and water quality data to provide early warning and also support emergency response during acute disturbance events (e.g., extreme weather, contaminant spills) when traditional field sampling is impractical or limited in scope. Here, we present time-series meteorological and water quality observations from TBON Station TB05 (Bayboro Harbor) during August 4–6, 2024, when Tampa Bay was impacted by Hurricane Debby (which developed from a tropical storm prior to August 4). We describe these temporal trends and perform a simple, qualitative interpretation using established marine ecosystem principles. While data users may consider alternative explanations for these changes, our goal is to demonstrate how TBON data can support such valuable analyses and scientific discussions.

Debby was located approximately 100 miles west of Tampa Bay around 20:00 EDT on August 4, 2024, at which point the lowest air pressure (~ 1004 mb, Fig. 6a) and the highest wind speed (~ 12.7 m/s, Fig. 6b) were observed at TB05. Due to its significantly higher specific heat capacity, water temperature changed more slowly and was more "buffered" compared to the overlying atmosphere. From 12:00 on August 3 to 20:00 on August 4, while the air temperature dropped substantially with high variability (Fig. 6c), the water temperature exhibited a more gradual and steady decrease (Fig. 6d). The drop in air temperature was associated with reduced solar radiation from cloud cover and the cooling effect of rainfall on the atmosphere (i.e., the fluctuation of air temperature was associated with the precipitation peaks shown in Fig. 6e). Air temperature started to increase immediately after the storm's closest passage (around 20:00 on August 4), reaching ~ 27 °C around midnight on August 4. The water temperature, in contrast, continued to decrease, reaching its minimum (~ 28 °C) around 5:00 on August 5, before increasing. A significant peak in precipitation rate was observed between 09:00 to 20:00 on August 4 (Fig. 6e) and corresponded with a subsequent and substantial decrease in salinity (reaching a low of 28, Fig. 6f). Following the storm's passage on August 5, salinity began to increase with peak values (~ 32) observed during the latter half of the day driven by tidal ingress.

Fig. 6.

Fig. 6

Open in a new tab

Meteorological and water quality data obtained from Station TB05 at Bayboro Harbor from August 4–6, 2024, during the impact of Hurricane Debby (Category 1). The panels show (a) air pressure (mb), (b) wind speed (m/s), (c) air temperature (°C), (d) water temperature (°C), (e) precipitation rate (mm/h), (f) salinity, (g) dissolved oxygen (mg/L), (h) chlorophyll-a concentration (µg/L), (i) pH, and (j) turbidity (NTU)

Dissolved oxygen (DO) variations are driven by complex biogeochemical processes, and its concentration is affected by many factors, including, for example, temperature and salinity (solubility increases when both temperature and salinity decrease), high-speed wind and wave-induced air bubble injection (increases DO); mixing of water with varying DO concentrations; and biological activity (respiration consumes oxygen, while photosynthesis releases it). Chl-a often serves as proxy measure for biological production, with higher Chl-a values indicating elevated algal biomass and associated photosynthesis. From 12:00 on August 3 to 00:00 on August 4, DO (Fig. 6g) and Chl-a (Fig. 6h) concentrations exhibited similar temporal variability, implicating phytoplankton metabolism as the primary determinant of DO. Early on August 4, DO was generally lower but exhibited several pronounced fluctuations. This was most likely a result of wind-induced vertical mixing of oxygen-rich surface water and low-oxygen bottom water. Later in the day on August 4, atmospheric injection became the dominant process, which led to the large increase and notable fluctuations in DO. The two variables (DO and Chl-a) were again highly synchronized for most of August 5 after the storm passed. On August 6, the increased DO may be due to the increasing wind speed at the end of this period. The deviation from the Chl-a trend indicated that air injection was a more important controlling factor than biological processes on that day. During the storm's impact, Chl-a was generally lower than that observed before and after the event. This was most likely due to enhanced mixing, which redistributed Chl-a throughout the entire water column, thereby lowering the surface Chl-a concentration. Additionally, the decreased light availability resulting from reduced solar radiation and increased particulate matter (due to freshwater runoff or sediment resuspension) also led to reduced primary production.

Various physical, chemical, and biological processes control pH of the water in the Bay, including, for example, equilibrium with atmospheric carbon dioxide (CO2), CO2 solubility (and its dependence on temperature and salinity), source water (precipitation and river water typically have a much lower pH than surface seawater), and biological activity (photosynthesis consumes CO2, increasing pH, while respiration releases CO2, decreasing pH). From 12:00 on August 3 to 00:00 on August 4, like DO, the pH change (Fig. 6i) also mirrored the temporal variation of Chl-a, indicating that biological processes were the primary control. During the first half of August 4, pH mirrored the trends observed in DO, suggesting that vertical mixing with low-pH bottom water was likely the primary control. pH gradually decreased from 12:00 on August 4 to late August 5, which could be attributed to the input of low-pH freshwater from precipitation and runoff. After that, the pH gradually recovered.

Turbidity (Fig. 6j) started to increase with wind in the early hours of August 4. Since the strong wind led to sediment resuspension and the heavy rain led to significant freshwater runoff, a turbidity peak was observed in the afternoon and night of August 4, when the storm's impact was strongest. This impact gradually subsided after August 5.

Leveraging high-frequency TBON data to advance model development

The Tampa Bay coastal ocean model (TBCOM) is a high‑spatial‑resolution, daily automated numerical circulation model designed to provide nowcasts and forecasts of oceanographic conditions (e.g., sea level, currents, salinity, and temperature) for Tampa Bay, Sarasota Bay, the Intracoastal Waterway, and all of the inlets connecting these water bodies with each other and with the adjacent coastal ocean (e.g., Chen et al., 2019, 2023b). The TBCOM has successfully been used for various applications, including evaluation of potential environmental effects of municipal engineering projects (e.g., Chen et al., 2023a; Zhu et al., 2015b), storm surge forecasting during hurricanes (Chen et al., 2018), and tracking HABs (e.g., Liu et al., 2023) and other pollution events (e.g., Chen et al., 2026; Liu et al., 2024). Future work will integrate new biogeochemical modules into TBCOM, expanding its simulation capabilities to include processes like biological productivity, inorganic carbon cycling, and organic carbon export. The high-frequency, continuous TBON data, such as Chl-a, pH, and FDOM, will be leveraged as a vital resource for both the model's development and validation, similar to those for the Chesapeake Bay Environmental Forecast System (Friedrichs et al., 2025).

Limitations and challenges

The TBON system faces several practical limitations, many of which stem from the inherent constraints of current in-situ sensor technologies. Biofouling remains a main operational challenge for collecting long time series in the subtropical estuary, particularly in summer. Occasional data gaps also occur due to instrument connector issues or power outage associated with blocked solar panels that affect real-time data transmission. To minimize instrument downtime, sensor maintenance is performed via frequent boat trips. Maintenance intervals vary from weeks to months depending on seasonal biofouling and environmental conditions. If signal anomalies are detected and confirmed to be an operational issue, targeted field maintenance is conducted promptly; otherwise, routine servicing occurs monthly. Furthermore, TBON sensors do not directly measure nutrients (such as nitrogen species), which are key water quality indicators for Tampa Bay and other estuaries (e.g., Kahl et al., 2007). This can be addressed in the future so that the real-time data can be used by the Tampa Bay Nitrogen Management Consortium (TBNMC) for better decision-making (Sherwood et al., 2016; Yang et al., 2026). Finally, the system’s current spatial coverage across the bay is limited. Expanding the network remains a primary goal for future development and will require additional investment.

Conclusions

The case studies in this work demonstrate that, even with its current footprint, TBON successfully bridges critical information gaps between traditional monthly field sampling and satellite remote sensing. This is particularly vital during acute disturbance events, such as tropical cyclones, when manual sampling is logistically impractical. By providing high-frequency data feeds and an accessible data portal, TBON serves as a functional tool for resource managers and stakeholders. Unlike other large, nationally significant estuaries in the United States that have long benefited from sustained monitoring, Tampa Bay has historically lacked a continuous data acquisition system to support real-time decision-making (Sherwood et al., 2016). Ultimately, this system will facilitate the collaborative research and informed policy formulation necessary to protect and restore the water quality of the Tampa Bay ecosystem.

Acknowledgements

TBON maintenance and data QA/QC were contributed by Mitchell Lemon, Orion Scharton, Jeffery Donovan, Amanda Muni-Morgan, Luis Sorinas, James Ivey, Grace Humphrey, Olivia Smith, Lilian Ring, and Alexander Trail.

Author contributions

Conceptualization: Yonggang Liu, Robert H. Weisberg, Jason A. Law, Bo Yang; Data curation: Jason A. Law, Samantha E. D’Angelo; Formal analysis and investigation: Bo Yang; Writing—original draft preparation: Bo Yang; Writing—review and editing: Thomas K. Frazer, Marcus W. Beck, Edward T. Sherwood, Yonggang Liu, Robert H. Weisberg, Jason A. Law, Kaili Qiao; Funding acquisition: Yonggang Liu; Supervision: Yonggang Liu.

Funding

The TBON was initiated and primarily supported by the State of Florida Department of Environmental Protection (Agreement # AT007). Partial support was also provided by the Tampa Bay Estuary Program (PO7445) and the Environmental Protection Commission of Hillsborough County Pollution Recovery Fund (Agreement # 25–01).

Data availability

TBON data can be accessed at https://tbon.marine.usf.edu/graph.php. MODIS satellite data can be downloaded from https://oceandata.sci.gsfc.nasa.gov/l3/. Data from EPC Station 82 can be found at: https://www.epchc.org/divisions/water/water-monitoring-maps-and-data.

Declarations

Ethical approval

All authors have read, understood, and have complied as applicable with the statement on “Ethical responsibilities of Authors” as found in the Instructions for Authors.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Ahmed, A. N., Othman, F. B., Afan, H. A., Ibrahim, R. K., Fai, C. M., Hossain, M. S., Ehteram, M., & Elshafie, A. (2019). Machine learning methods for better water quality prediction. Journal of Hydrology,578, 124084. 10.1016/j.jhydrol.2019.124084 [DOI] [Google Scholar]
  2. Beck, M. W. (2016). SWMPr: An R package for retrieving, organizing, and analyzing environmental data for estuaries. R Journal. 10.32614/RJ-2016-015 [DOI] [Google Scholar]
  3. Beck, M. W., Altieri, A., Angelini, C., Burke, M. C., Chen, J., Chin, D. W., Gardiner, J., Hu, C., Hubbard, K. A., Liu, Y., Lopez, C., Medina, M., Morrison, E., Philips, E. J., Raulerson, G. E., Scolaro, S., Sherwood, E. T., Tomasko, D., Weisberg, R. H., & Whalen, J. (2022). Initial estuarine response to inorganic nutrient inputs from a legacy mining facility adjacent to Tampa Bay, Florida. Marine Pollution Bulletin,178, 113598. 10.1016/j.marpolbul.2022.113598 [DOI] [PubMed] [Google Scholar]
  4. Beck, M. W., Burke, M. C., Raulerson, G. E., Scolaro, S., Sherwood, E. T., & Whalen, J. (2023). Coordinated monitoring of the Piney Point wastewater discharge into Tampa Bay. Florida Scientist,86(2), 288–300. [Google Scholar]
  5. Beck, M. W., Sherwood, E. T., Henkel, J. R., Dorans, K., Ireland, K., & Varela, P. (2019). Assessment of the cumulative effects of restoration activities on water quality in Tampa Bay, Florida. Estuaries and Coasts, 42(7), 1774–1791. s12237-019-00619-w [PMC free article] [PubMed]
  6. Cerco, C. F., & Noel, M. R. (2013). Twenty-one-year simulation of Chesapeake Bay water quality using the CE-QUAL-ICM eutrophication model. JAWRA Journal of the American Water Resources Association,49(5), 1119–1133. 10.1111/jawr.12107 [DOI] [Google Scholar]
  7. Chen, J., Weisberg, R. H., Liu, Y., & Zheng, L. (2018). The Tampa Bay coastal ocean model performance for Hurricane Irma. Marine Technology Society Journal,52(3), 33–42. 10.4031/MTSJ.52.3.6 [DOI] [Google Scholar]
  8. Chen, J., Weisberg, R. H., Liu, Y., Zheng, L., & Zhu, J. (2019). On the momentum balance of Tampa Bay. Journal of Geophysical Research: Oceans,124(7), 4492–4510. 10.1029/2018JC014890 [DOI] [Google Scholar]
  9. Chen, J., Weisberg, R. H., Liu, Y., & Zheng, L. (2023a). Hillsborough bay inflow modification study: An application of the Tampa Bay Coastal Ocean Model. Estuarine, Coastal and Shelf Science,281, 108213. 10.1016/j.ecss.2023.108213 [DOI] [Google Scholar]
  10. Chen, J., Weisberg, R. H., Liu, Y., Zheng, L., Law, J., Gilbert, S., & Murawski, S. A. (2023b). A Tampa Bay coastal ocean model (TBCOM) nowcast/forecast system. Deep Sea Research Part II: Topical Studies in Oceanography,211, 105322. 10.1016/j.dsr2.2023.105322 [DOI] [Google Scholar]
  11. Chen, J., Weisberg, R. H., Zheng, L., Hu, C., Zhang, Y., Nickerson, A. K., Donovan, J. C., & Liu, Y. (2026). Frontogenesis by material convergence in an estuary and its adjacent coastal ocean. Limnology and Oceanography,71, e70276. 10.1002/lno.70276 [DOI] [Google Scholar]
  12. Chen, Z., Hu, C., & Muller-Karger, F. (2007). Monitoring turbidity in Tampa Bay using MODIS/Aqua 250-m imagery. Remote sensing of Environment, 109(2), 207–220, 10.1016/j.rse.2006.12.019 [DOI]
  13. Etcheber, H., Schmidt, S., Sottolichio, A., Maneux, E., Chabaux, G., Escalier, J. M., Wennekes, H., Derriennic, H., Schmeltz, M., Quéméner, L., Repecaud, M., Woerther, P., & Castaing, P. (2011). Monitoring water quality in estuarine environments: Lessons from the MAGEST monitoring program in the Gironde fluvial-estuarine system. Hydrology and Earth System Sciences,15(3), 831–840. 10.5194/hess-15-831-2011 [DOI] [Google Scholar]
  14. Ferla, M., Cordella, M., Michielli, L., & Rusconi, A. (2007). Long-term variations on sea level and tidal regime in the lagoon of Venice. Estuarine, Coastal and Shelf Science,75(1–2), 214–222. 10.1016/j.ecss.2007.03.037 [DOI] [Google Scholar]
  15. Friedrichs, M. A. M., Bever, A., Musick, S., & Hudson, K. (2025). The Chesapeake Bay environmental forecast system. Benefits of Ocean Observing Catalog (BOOC),3(1), 4. 10.15351/3068-2320.1116 [DOI] [Google Scholar]
  16. Gilbert, M., Needoba, J., Koch, C., Barnard, A., & Baptista, A. (2013). Nutrient loading and transformations in the Columbia River estuary determined by high-resolution in situ sensors. Estuaries and Coasts,36(4), 708–727. 10.1007/s12237-013-9597-0 [DOI] [Google Scholar]
  17. Goodwin, C. R. (1987). Tidal-flow, circulation, and flushing changes caused by dredge and fill in Tampa Bay, Florida (No. 2282). Department of the Interior, US Geological Survey. 10.3133/wsp2282 [DOI]
  18. Greening, H., & Janicki, A. (2006). Toward reversal of eutrophic conditions in a subtropical estuary: Water quality and seagrass response to nitrogen loading reductions in Tampa Bay, Florida, USA. Environmental Management,38(2), 163–178. 10.1007/s00267-005-0079-4 [DOI] [PubMed] [Google Scholar]
  19. Greening, H., Janicki, A., Sherwood, E. T., Pribble, R., & Johansson, J. O. R. (2014). Ecosystem responses to long-term nutrient management in an urban estuary: Tampa Bay, Florida, USA. Estuarine, Coastal and Shelf Science,151, A1–A16. 10.1016/j.ecss.2014.10.003 [DOI] [Google Scholar]
  20. Hanisak, M. D., & Davis, K. S. (2018). Interannual variability in the Indian River Lagoon, Florida, measured by a network of environmental sensors. Ocean Sensing and Monitoring X,10631, 83–98. 10.1117/12.2304609 [DOI] [Google Scholar]
  21. Hu, C., Feng, L., Lee, Z., Franz, B. A., Bailey, S. W., Werdell, P. J., & Proctor, C. W. (2019). Improving satellite global chlorophyll a data products through algorithm refinement and data recovery. Journal of Geophysical Research: Oceans,124(3), 1524–1543. 10.1029/2019JC014941 [DOI] [Google Scholar]
  22. Hu, C., Lee, Z., & Franz, B. (2012). Chlorophyll a algorithms for oligotrophic oceans: A novel approach based on three‐band reflectance difference. Journal of Geophysical Research: Oceans, 117(C1). 10.1029/2011JC007395 [DOI]
  23. Kahl, J. S., Nelson, S. J., Fernandez, I., Haines, T., Norton, S., Wiersma, G. B., Jacobson, G., Jr., Amirbahman, A., Johnson, K., Schauffler, M., Rustad, L., Tonnessen, K., Lent, R., Bank, M., Elvir, J., Eckhoff, J., Caron, H., Ruck, P., Parker, J., … Grygo, A. (2007). Watershed nitrogen and mercury geochemical fluxes integrate landscape factors in long-term research watersheds at Acadia National Park, Maine, USA. Environmental Monitoring and Assessment,126, 9–25. 10.1007/s10661-006-9328-0 [DOI] [PubMed] [Google Scholar]
  24. Karlen, D. J., Campbell, K., Ash, T., Goetting, B., Pratt, C., Mathiak, E., & Schuman, M. (2023). Long-term water quality trends in Tampa Bay (1974–2020). Florida Scientist,86(2), 400–408. [Google Scholar]
  25. Le, C., Hu, C., English, D., Cannizzaro, J., Chen, Z., Feng, L., Boler, R., & Kovach, C. (2013). Towards a long-term chlorophyll-a data record in a turbid estuary using MODIS observations. Progress in Oceanography,109, 90–10. 10.1016/j.pocean.2012.10.002 [DOI] [Google Scholar]
  26. Liu, Y., Weisberg, R. H., Zheng, L., Hubbard, K. A., Muhlbach, E. G., Garrett, M. J., Hu, C., Cannizzaro, J. P., Xie, Y., Chen, J., John, S., & Liu, L. Y. (2023). Short-term forecast of Karenia brevis trajectory on the West Florida Shelf. Deep Sea Research Part II: Topical Studies in Oceanography,212, 105335. 10.1016/j.dsr2.2023.105335 [DOI] [Google Scholar]
  27. Liu, Y., Weisberg, R. H., Zheng, L., Sun, Y., Chen, J., Law, J. A., Hu, C., Cannizzaro, J. P., & Frazer, T. K. (2024). A tracer model nowcast/forecast study of the Tampa Bay, Piney Point effluent plume: Rapid response to an environmental hazard. Marine Pollution Bulletin,198, 115840. 10.1016/j.marpolbul.2023.115840 [DOI] [PubMed] [Google Scholar]
  28. Lopez, C. B., Shankar, S., Kaminski, S. G., & Hubbard, K. A. (2023). Pyrodinium bahamense bloom dynamics in Old Tampa Bay, FL, with a focus on Feather Sound. Florida Scientist,86, 62–71. [Google Scholar]
  29. Milbrandt, E. C., Bartleson, R. D., Martignette, A. J., Siwicke, J., & Thompson, M. (2016). Evaluating light attenuation and low salinity in the lower Caloosahatchee Estuary with the River, Estuary, and Coastal Observing Network (RECON). Florida Scientist,79, 109–124. [Google Scholar]
  30. Morrison, E. S., Phlips, E., Badylak, S., Chappel, A. R., Altieri, A. H., Osborne, T. Z., Tomasko, D., Beck, M. W., & Sherwood, E. (2023). The response of Tampa Bay to a legacy mining nutrient release in the year following the event. Frontiers in Ecology and Evolution,11, 1144778. 10.3389/fevo.2023.1144778 [DOI] [Google Scholar]
  31. Murphy, R. R., Curriero, F. C., & Ball, W. P. (2010). Comparison of spatial interpolation methods for water quality evaluation in the Chesapeake Bay. Journal of Environmental Engineering,136(2), 160–171. 10.1061/(ASCE)EE.1943-7870.0000121 [DOI] [Google Scholar]
  32. Navarro, G., Gutiérrez, F. J., Díez-Minguito, M., Losada, M. A., & Ruiz, J. (2011). Temporal and spatial variability in the Guadalquivir estuary: A challenge for real-time telemetry. Ocean Dynamics,61(6), 753–765. 10.1007/s10236-011-0379-6 [DOI] [Google Scholar]
  33. Rosenau, N. A., Galavotti, H., Yates, K. K., Bohlen, C. C., Hunt, C. W., Liebman, M., Brow, C. A., Pacella, S. R., Lagier, J. L., Nielsen, K. J., Hu, X., McCutcheon, M. R., Vasslides, J. A., Poach, M., Ford, T., Johnston, K., & Steele, A. (2021). Integrating high-resolution coastal acidification monitoring data across seven United States estuaries. Frontiers in Marine Science,8, 679913. 10.3389/fmars.2021.679913 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sherwood, E. T., Greening, H. S., Janicki, A. J., & Karlen, D. J. (2016). Tampa Bay estuary: Monitoring long-term recovery through regional partnerships. Regional Studies in Marine Science,4, 1–11. 10.1016/j.rsma.2015.05.005 [DOI] [Google Scholar]
  35. Sherwood, E. T., Greening, H. S., Johansson, J. R., Kaufman, K., & Raulerson, G. E. (2017). Tampa Bay (Florida, USA): Documenting seagrass recovery since the 1980’s and reviewing the benefits. Southeastern Geographer,57(3), 294–319. 10.1353/sgo.2017.0026 [DOI] [Google Scholar]
  36. Tango, P. J., Zhang, Q., Tian, R., Murphy, R. R., Sullivan, B., Mallonee, M., Ghosh, D., Goldfischer, A., & Gootman, K. S. (2025). Adaptive monitoring for change: Record low hypoxia in Chesapeake Bay in 2023. Ocean-Land-Atmosphere Research,4, 0076. 10.34133/olar.0076 [DOI] [Google Scholar]
  37. Weisberg, R. H., & Zheng, L. (2006). Circulation of Tampa Bay driven by buoyancy, tides, and winds, as simulated using a finite volume coastal ocean model. Journal of Geophysical Research,111, 2005JC003067. 10.1029/2005JC003067 [DOI] [Google Scholar]
  38. Wilson, W. D., & Siegel, E. (2011). Current and wave measurements in support of the Chesapeake Bay Interpretive Buoy System. In 2011 IEEE/OES 10th Current, Waves and Turbulence Measurements (CWTM) (pp. 94–99). IEEE. 10.1109/CWTM.2011.5759533 [DOI]
  39. Xian, G., Crane, M., & Su, J. (2007). An analysis of urban development and its environmental impact on the Tampa Bay watershed. Journal of Environmental Management,85(4), 965–976. 10.1016/j.jenvman.2006.11.012 [DOI] [PubMed] [Google Scholar]
  40. Yang, B., Liu, Y., Tang, W., Law, J. A., Weisberg, R. H., Beck, M. W., Sherwood, E. T., & Frazer, T. K. (2026). A machine learning model for real-time total nitrogen monitoring in Tampa Bay. Marine Pollution Bulletin,233, 120184. 10.1016/j.marpolbul.2026.120184 [DOI] [PubMed] [Google Scholar]
  41. Yates, K. K., Greening, H., & Morrison, G. (2011). Integrating science and resource management in Tampa Bay, Florida (No. 1348, pp. 1–280). US Geological Survey, 10.3133/cir1348 [DOI]
  42. Yentsch, C. S., Lapointe, B. E., Poulton, N., & Phinney, D. A. (2008). Anatomy of a red tide bloom off the southwest coast of Florida. Harmful Algae,7(6), 817–826. 10.1016/j.hal.2008.04.008 [DOI] [Google Scholar]
  43. Zhang, Q., Shenk, G. W., Bhatt, G., & Bertani, I. (2025). METRIC: An interactive framework for integrated visualization and analysis of monitored and expected load reductions for nitrogen, phosphorus, and sediment in the Chesapeake Bay watershed. Environmental Modelling & Software,188, 106379. 10.1016/j.envsoft.2025.106379 [DOI] [Google Scholar]
  44. Zhu, J., Weisberg, R. H., Zheng, L., & Han, S. (2015a). On the flushing of Tampa Bay. Estuaries and Coasts,38, 118–131. 10.1007/s12237-014-9793-6 [DOI] [Google Scholar]
  45. Zhu, J., Weisberg, R. H., Zheng, L., & Han, S. (2015b). Influences of channel deepening and widening on the tidal and non-tidal circulation of Tampa Bay. Estuaries and Coasts,38, 132–150. 10.1007/s12237-014-9815-4 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

TBON data can be accessed at https://tbon.marine.usf.edu/graph.php. MODIS satellite data can be downloaded from https://oceandata.sci.gsfc.nasa.gov/l3/. Data from EPC Station 82 can be found at: https://www.epchc.org/divisions/water/water-monitoring-maps-and-data.

Articles from Environmental Monitoring and Assessment are provided here courtesy of Springer

RESOURCES