Suspended sediment loads were calculated for the monitoring sites on the North Stann Creek (Middlesex and Melinda), the Sittee River and the South Stann Creek (Highway and Banana) sites. The loads were calculated by combining water discharge and suspended sediment concentration data. This calculation involved:

  1. Continuous monitoring of turbidity at each site every 15 minutes over the study period using optical sensors.
  2. Collection of river suspended sediment samples corresponding to observed turbidity values. These samples were then filtered in the laboratory to determine suspended sediment concentrations. Optical probes used for the continuous monitoring were calibrated using these measurements of actual suspended sediment concentration. The continuous turbidity records were then converted to a continuous suspended sediment record using the calibration relationship.
  3. The continuous suspended sediment concentration data were then combined with the continuous discharge record to calculate the suspended sediment load.
Optical sensors to monitor suspended sediment concentrations were installed at the same location as stage (discharge) monitoring equipment.  An assumption was made of total mixing of suspended sediment across the river, so sensors could be located at the edge of main flow (although backwater sites were avoided).

                                                                                         ^ analite sensor

Two types of turbidity probe were used during the study period; Partech single gap type probes (measuring light attenuation) and Analite probes (measuring backscattered light). Only the Partech probes were used during 1998 and a combination of Partech and the Analite probes during 1999. These probes were routinely calibrated using a formazine solution (NTU units) to check for stability of  the output.

The probes needed to be in the river at all times for measuring but also required to be removed for regular (weekly) maintenance and cleaning. To achieve this, the probes were fixed into a carrier that slid up and down a "H" shaped metal girder (>6m long) that was driven into the river bed and fixed at the top to a permanent structure (<<photo).

The probes were connected to loggers which both controlled the sensor operation and stored the data generated. The loggers were programmed to make measurements every 30 seconds and then average these readings over a 15 minute period and store the averaged data.  Concentrations of suspended sediment may vary significantly over short periods and consequently this sampling regime was chosen.

It is important to note that whilst these turbidity probes do represent a more accurate way of continuously monitoring suspended sediment concentrations, they have limitations. The probes require a relatively high level of maintenance to ensure their effective operation. Biological growth on the sensors affects their output.  Significant fouling was observed to occur in under 36 hours during the hottest weather. The Analite probes overcome this problem by having a self-cleaning device.

The voltage output from the Analite probes used during the 1999 season was found to have a strong linear relationship with formazine and also suspended sediment concentrations (see graphs below). It was not possible to collect suspended sediment samples over the whole range of turbidity measurements during the study period for all rivers but the calibration with formazine and the results from several of the rivers showed that a linear relationship could be assumed over the whole range of the probe. The Partech sensors were less easy to calibrate.

Calibration of Analite sensors to formazine solution (A) and calibration to filtered suspended sediment samples from the relevant river; South Stann Creek Highway (B), Sittee River (C), S. Stann Creek, banana (D), North Stann Creek, Melinda (E) and N. Stann Creek, Middlesex (F)