Temporal and spatial variability in the Guadalquivir estuary: a challenge for real-time telemetry
- First Online:
- Cite this article as:
- Navarro, G., Gutiérrez, F.J., Díez-Minguito, M. et al. Ocean Dynamics (2011) 61: 753. doi:10.1007/s10236-011-0379-6
- 223 Views
Meteorological, hydrological, and hydrodynamic data for 3 years (2008–2010) have been used to document and explain the temporal and spatial variability of the physical–biogeochemical interactions in the Guadalquivir River Estuary. A real-time, remote monitoring network has been deployed along the course of the river between its mouth and Seville to study a broad range of temporal scales (semidiurnal, diurnal, fortnightly, and seasonal). This network consists of eight hydrological monitoring stations capable of measuring temperature, conductivity, dissolved oxygen, turbidity, and chlorophyll fluorescence at four depths. In addition, six stations have been deployed to study hydrodynamics, obtaining 20-cell water column current profiles, and there is a meteorological station at the river mouth providing data for understanding atmospheric interactions. Completing this data-gathering network, there are several moorings (tide gauges, current/wave sensors, and a thermistor chain) deployed in the estuary and river mouth. Various sources of physical forcing, such as wind, tide-associated currents, and river discharge, are responsible for the particular temporal and spatial patterns of turbidity and salinity found in the estuary. These variables force the distribution of biogeochemical variables, such as dissolved oxygen and chlorophyll fluorescence. In particular, episodes of elevated turbidity (when suspended particle matter concentration >3,000 mg/l) have been detected by the network, together with episodes of declining values of salinity and dissolved oxygen. All these patterns are related to river discharge and tidal dynamics (spring/neap and high/low tide).
KeywordsGuadalquivir estuaryReal-time remote monitoringSensor technologyTelemetryWater quality
In spite of its importance, until very recently, only a few studies have been carried out in the estuary; these have focused mainly on fisheries, crustacean decapods (Baldó and Drake 2002; Fernández-Delgado et al. 2007; González-Ortegón et al. 2010), and chemical contamination following the Aznalcollar mining waste spill (Grimalt et al. 1999; Blasco et al. 1999). Three years ago, to obtain a better understanding of the physical and biological processes governing the estuary ecosystem, an interdisciplinary research program was established, with the following main goal: to “develop an integral method to diagnose and forecast the consequences of human actions on the Guadalquivir estuary.” To achieve this goal, a comprehensive monitoring program has been established in the estuary and river mouth. This program has comprised three main parts: the real-time remote monitoring (RTRM) network, the deployment of several moorings in the estuary and river mouth, and monthly cruises for water sampling. The stations of the network and moorings are strategically positioned at locations that are important for hydrological dynamics. The locations span the whole estuary but with increased spatial resolution and shorter distance between stations close to the mouth, where environmental gradients are more acute and dynamic (Fig. 1).
This monitoring program was set up in the summer of 2007 in the estuary and near the continental shelf and is providing detailed data for research, policy-makers, government agencies, and education/outreach applications on coastal and transition waters management. In this paper, we report on the technical layout of the various stations and on the experience gained of their performance during the last 3 years of operation. The paper is focused mainly on the RTRM network that has been operated in the study area by the Instituto de Ciencias Marinas de Andalucía (ICMAN). This network provides data with a high degree of temporal resolution within a Eulerian framework and has the advantage that constant surveillance can be carried out to rapidly detect any changes and trends in critical indicators. Remotely acquired data from continuous in situ monitoring provides important early warning information to decision-makers so that they can respond appropriately. RTRM technologies have emerged as an economically viable means of monitoring key hydrological parameters (Glasgow et al. 2004). The calibration/validation of the instruments is described in the “Results and discussion” section together with the approaches adopted to overcome the limitations that were encountered during the operational use of these stations. Also, in the “Results and discussion” section, we present a preliminary analysis of the data sets obtained with the network with the aim of understanding the temporal and spatial variability of the physical and biological processes that dominate in the estuary. Finally, in the “Conclusions” section, some conclusions about the RTRM are presented.
2 Materials and methods
The Guadalquivir estuary (SW Spain: 36 45′–37 15′ N, 6 00′–6 22′ W) is a well-mixed system with a longitudinal salinity gradient (Vannéy Vanney 1970; Álvarez et al. 2001). The tidal influence extends up to the Alcalá del Río dam and the maximum tidal range for the Guadalquivir mouth is 3.86 m (Rodriguez-Ramirez and Yáñez-Camacho 2008). Between the river mouth and the Port of Seville, several navigation buoys (Fig. 1) are deployed by the Port Authority of Seville, and these have been transformed into environmental laboratories.
2.2 Monitoring program
The three parts that comprised the monitoring program were set up on different dates: monthly oceanographic cruises started in June 2007 and the installation of the moorings and RTRM stations started in March 2008 and at the beginning of 2008, respectively. The real-time telemetry system was aimed at providing online continuous meteorological, hydrographic, and water quality information. The water samples were used both to supplement the telemetered water quality information and to provide a cross-check on some of the other important parameters. To complete the monitoring program, daily data on discharges from the Alcala del Rio dam were obtained from the regional water management agency (Agencia Andaluza del Agua, Junta de Andalucía, http://www.juntadeandalucia.es/agenciadelagua/saih/). This agency also carried out several cruises in the estuary to measure hydrological parameters (conductivity, dissolved oxygen, and turbidity), and its water quality stations are situated along the estuary (Fig. 1). Moreover, there are some meteorological stations located in Sanlucar de Barrameda and Chipiona at the estuary mouth that we have used to validate data obtained by our meteorological station. In addition, RGB MODIS images have been downloaded from the NASA server (http://rapidfire.sci.gsfc.nasa.gov/). These images are processed and projected using MATLAB(c) software. Satellite imagery can resolve patterns of turbidity on a large spatial scale but is confined to surface data.
Water samples were collected each month at the field stations from June 2007. Parameters measured in the routine water sampling included: nutrient concentration (nitrite, nitrate, phosphate, and silicate) following JGOFS standards (UNESCO 1994), chlorophyll concentration using fluorometric methods following JGOFS protocols (UNESCO 1994), dissolved organic carbon and total nitrogen (TOC-VCPH, Shimadzu, Japan), dissolved oxygen (Winkler method), total alkalinity (Pérez and Fraga 1987), suspended particle matter (SPM) using the gravimetric method, phytoplankton, and zooplankton. Vertical profiles of temperature, salinity, dissolved oxygen, turbidity, and chlorophyll fluorescence were measured with a CTD-Seabird 19 probe with external sensor. Water samples were taken at the same depths as the pumping levels of the RTRM station to validate quality data from the RTRM.
Sensor equipment installed in moorings
Aqualogger 520 PT, between March and June 2008. sampling rate, 10 min
NKE SP2T logger between June 2008 and March 2009. Sampling rate, 10 min
RBR-Thermometrics, between November 2007 and August 2008. Sampling rate, 1 min
NKE-S2T. From June 2009. Sampling rate, 10 min
AWAC-AST Nortek 1,000 kHz. Currents: sampling rate, 10 min; cell size, 1 m. Waves: sampling rate, 1 h
The RTRM network comprises three types of station: for water dynamics, six stations; for water quality, eight stations; and one meteorological station. These stations are installed on the navigation buoys positioned between the river mouth and Seville harbor for water dynamics and water quality and on the Salmedina buoy for meteorological monitoring (Fig. 1). The stations were designed for year-round operation, with planned maintenance quarterly. All stations were strategically positioned. The technologies employed offer several advantages over historical monitoring techniques: they streamline the data collection process, they potentially minimize human errors and time delays, they reduce the overall cost of data collection, and they significantly increase the quantity and potentially increase the quality of data obtained on temporal and spatial scales (Glasgow et al. 2004). Details of the technology applied in the construction of the prototype stations for water dynamics and water quality can be found in Gutierrez et al. (2009).
2.2.1 Meteorological station
The meteorological station was installed on the Salmedina buoy (Fig. 1e) located over the continental shelf off Chipiona (Fig. 1, map) in May 2008. The system comprises an array of sensors for meteorological measurements (air temperature, relative humidity, incident solar radiation, barometric pressure, and wind speed and direction). Meteorological data is collected as follows. Barometric pressure is obtained with a Young 61202 L barometer. Air temperature and relative humidity data are acquired via a Geonica STH-5031 instrument (Geonica, Spain). Wind speed and direction are obtained using a marine wind sensor (R.M. Young Wind Monitor). Incident solar radiation is collected by a pyranometer (model Licor Li200 LiCor Biosciences, Lincoln, NE, USA). Different measuring intervals can be selected but this station was designed to sample every second, and the average, maximum, and minimum values for a 10-min interval are sent to the laboratory by telemetry. A Geonica Hydrodata 2008CP datalogger serves as the central processing unit of the system and the energy supply is from a gel battery charged by a bank of three solar panels connected in parallel. Bi-directional communication between the meteorological station and the laboratory enables the instruments and sensors installed to be controlled and serviced remotely.
2.2.2 Water dynamics stations
Six stations have been deployed along the estuary (Fig. 1, map) since February 2008. Figure 1d, g also shows photographs of the transformed navigation buoy at a current telemetry station. These stations are equipped with a Nortek AS Aquadopp acoustic Doppler current profiler (ADP) operating at a frequency of 1,000 kHz. The deployment configuration has been set to a cell size of 1 m, with an integration period of 2 min and four profiles per hour. The ADP has been programmed independently from the telemetry module as the two modules work asynchronously. The ADP also measures temperature, head pressure, and pitch and roll. These parameters are used in real time to assess the quality of data. The ADP is controlled by a Geonica Hydrodata 2008CP datalogger, and the data are sent to the laboratory every 15 min by GPRS system. The energy supply to the ADP and datalogger is from a gel battery charged by a bank of three 45-W solar panels connected in parallel. The compass was calibrated following the manufacturer’s specifications. Information on the amount of suspended particles can be deduced from the intensity of the backscatter signal measured by the ADP (Hill et al. 2003).
2.2.3 Water quality stations
Eight stations have been deployed along the estuary (Fig. 1, map) since February 2008. Figure 1c, f also shows photographs of the transformed navigation buoys operating as water quality laboratories. This type of station comprises four modules: the power module (a bank of three 120-W solar panels connected in parallel, charging a gel battery), the hydraulic module (a SHURflo suction pump, flow meter, batch filter, and silicon piping), the measurement module, and the control module (a Geonica Hydrodata 2008CP datalogger). The functioning of this station is described in more detail in Gutierrez et al. (2009). Routine maintenance of the water quality stations was carried out quarterly and involved replacing the pumping module and cleaning the CTD and external sensors. When necessary, the anti-fouling cylinders (SBE A24173) inside the CTD were replaced.
The measurement module is comprised of a Seabird Electronics SBE16plus conductivity and temperature recorder with external sensors for dissolved oxygen (SBE43), a chlorophyll fluorometer (Turner Designs, model Cyclops-7), and a turbidimeter (Turner Designs, model Cyclops-7). The salinity (S) and density (ρ) are calculated from temperature, conductivity, and hydrostatic pressure (p) with the equations of state of seawater (UNESCO 1985). Measuring intervals can be selected, but the standard measuring rate is two cycles per hour.
During the first 9 months, the water quality stations took measurements at four depths (1-, 2-, 3-, and 4-m depth from the sea surface). After November 2008, measurements were only taken at a depth of 1 m because the estuary was well mixed and the maintenance cost could thus be reduced. Continuous monitoring generates more than 7,000 water quality records per day, accessible in near real-time.
Logged data were uploaded to a data acquisition computer located in the laboratory via GPRS communications and posted in graphical form on the website (http://www.guadalquivir.csic.es); this webpage is currently password-protected and therefore not accessible to the general public. Data acquisition programs were modified for the specific instrument packages.
3 Results and discussion
3.1 Calibration/validation program
The Guadalquivir estuary is characterized by high current velocities, high loadings of suspended matter (González-Ortegón et al. 2010), and high biological productivity (Ruiz and Navarro 2008). All of these characteristics affect the stability of the instruments, which have been deployed for extensive periods of time. Therefore, a calibration/validation program is required to monitor and maintain the quality of the data collected by the RMRT network during its operation.
3.1.1 Calibration program
ADP instruments (an AWAC-AST located at the river mouth and the several Aquapro’s deployed at water dynamics stations) and meteorological sensors were calibrated by their manufacturers. For the water dynamics stations, every compass was calibrated at the buoys to compensate for the spurious or residual magnetic effect. The tide gauges (Aqualogger 520 PT and NKE SP2T) and the thermistor sensor of the thermistor chain were also calibrated by their manufacturers.
Sensors for temperature (SBE16plus), conductivity (SBE16plus), and dissolved oxygen (SBE43) installed at the water quality stations were calibrated by the manufacturer, Seabird Electronics. Turbidity and chlorophyll fluorescence sensors (Turner Design's Cyclops) were calibrated in our laboratory as detailed in the following subsections.
Fourth-degree polynomial calibration coefficients for turbidity sensor. Turbidity (FNU) = p1 * V3 + p2 * V2 + p3 * V + p4, where V is the turbidity sensor response in volts (Gain 1x). R2 is regression coefficient, Voffset is minimum sensor response in millivolts, Vmax is 90% of the sensor saturation signal, and Tmax represents the turbidity in FNU corresponding to Vmax
With respect to chlorophyll fluorescence and due to the high turbidity levels found in the Guadalquivir estuary, turbidity has two primary effects on chlorophyll fluorescence readings: first, it may increase blank readings due to increased light scatter and, second, it may reduce the fluorescence reading due to light absorption. To calibrate the chlorophyll fluorometer in turbid water, we used the method proposed by the manufacturer (Turner Designs, Application Note S-0035), whereby we have simultaneously measured turbidity and fluorescence in estuary waters. We have demonstrated that fluorometers run correctly when the turbidity is lower than 1,500 FNU.
3.1.2 Validation program
Data for currents from the water dynamics stations were validated with a different ADP (ADP-1,000 kHz SonTek) integrating GPS and bottom tracking. During a tidal cycle, we measured water velocity profiles near the buoy using the same equipment configuration (cell size, integration time, etc). The results show that the water velocity profiles recorded by the RTRM stations are measured correctly (data not shown).
3.2 General field operation of the RTRM
Figure 6 shows that data availability can be considered very good since more than 97% of the total possible data was acquired. In addition, the data quality is sufficient in terms of accuracy, as is evident from Figs. 4 and 5. System reliability is guaranteed by the fact that measurement and analysis were still operational under severe conditions, including storms, high winds, spring tides, and huge river discharges. Results obtained during some of these extreme events will be analyzed, and they will be used to validate hydrodynamic and ecological models.
3.3 Multi-year time series data
This paper presents high-resolution, long-term hydrological and hydrodynamics measurements obtained at the Guadalquivir estuary. The data reveal the wide fluctuations of water quality parameters in highly dynamic tidal regimes. In addition, special events, such as a high river discharge, are captured (April–May 2008, March–April 2009, and December 2009; Fig. 5). These data are valuable for estimating the impact of such events on the biogeochemical dynamics in the Guadalquivir estuary. For instance, they can be used to validate model calculations. These topics will be discussed in more detail in a separate paper.
After more than 2 years of operation, the RTRM system described provides an effective means of monitoring estuarine Guadalquivir water, with high-temporal resolution. This platform is very important because estuarine ecosystems are characterized by high inherent spatio-temporal variability (Day et al. 1989). Reference has been made both to the mechanical modification of navigation buoys to convert them into valuable environmental laboratories and to the sensors and loggers shown to be adequate for reliably measuring the key meteorological, hydrodynamic, and hydrological parameters. These parameters are used by specialists to make decisions on maintenance and the operation of each station for monitoring water dynamics and water quality. The stations can be mounted without any alteration of the buoy, and this can be performed in situ in a few hours (two or three stations can be mounted per day). The modularity of the stations allows them to be adapted easily and deployed in other estuaries, ports, dikes, piers, etc., with minimal maintenance.
RTRM enables a variety of types of study and application to be carried out, as demonstrated in the previous section. The sampling rates were 10, 15, and 30 min for meteorological, water dynamics, and water quality stations, respectively. These rates allow processes such as overtides (harmonics like M6 and M4) and inter-annual variability to be resolved. Numerical models have been applied to this large volume of high-frequency information to make forecasts of how ecosystems in the estuary are likely to develop. In addition, real-time remote monitoring could be used to generate early warnings for navigation, algal bloom, turbidity increases, etc. The data can also be sent automatically to other cooperating agencies.
This type of system would aid in the earlier forward planning of measures to mitigate undesirable events or changes. In the near future and given the modularity of the network, we hope to install a new type of biosensor (Altamirano et al. 2004) to detect and give immediate warnings of phytoplankton blooms, a crude oil sensor to warn about oil spills, and a nutrient analyser to study the eutrophication of the estuary. Despite the present economic context, the long-term continuity of the system is under consideration by regional administration given its diagnosing power, autonomous simplicity, and low-cost robustness.
This project was supported by the Autoridad Portuaria de Sevilla (APS) and by the Consejería de Innovación, Ciencia y Empresa (Junta de Andalucía). MODIS images have been processed in the context of P09-RNM-4583 Project. Thanks are due to all the APS technicians and divers of AGUAYO S.L. who participated. The authors also thank, in particular, David Roque, Raúl García, Joaquín Pampin, and Antonio Moreno (ICMAN-CSIC) for their assistance.