Deformation Monitoring of the Pungu-Gumongo Steel Bridge Using Geodetic Techniques

Engineering structures such as bridges, dams, tunnels, and high-rise buildings undergo constant deformation due to natural and human-induced factors, including changes in groundwater levels, temperature fluctuations, seismic events, traffic loads, and construction-related ground disturbance. Excessive or uncontrolled deformation can threaten structural safety and usability, making systematic monitoring and analysis crucial for modern infrastructure management (Aktan & Catbas, 2003). Deformation monitoring, also called deformation surveying, involves systematically measuring and tracking changes in an object's position, shape, or size due to applied loads and environmental factors. Geodetic techniques, such as precise leveling, total station measurements, and Global Positioning System (GPS) observations, provide accurate three-dimensional data on structural movements and are commonly used to monitor bridges, dams, and other critical structures (Chrzanowski & Szostak, 2009). In Ghana, few studies have focused on systematic deformation monitoring of bridges and extensive civil infrastructure, despite increasing concerns about structural failures and building collapses in the region. The Pungu-Gumongo Bridge, which spans the river between the two communities, acts as a vital link. Its structural safety is paramount given its high daily traffic. This study aims to monitor short-term bridge deformation using geodetic techniques. Specifically, it seeks to determine whether observable movements occur in horizontal and vertical directions, estimate their magnitudes and directions, and assess whether temperature differences between morning and afternoon affect the structure's behavior.

The specific objectives are to:

1. Detect horizontal and vertical movements and estimate their rates.

2. Assess the relationship between temperature and the measured deformations.

LITERATURE REVIEW

Deformation monitoring techniques can be broadly categorised into geodetic and non-geodetic approaches. Geodetic methods (e.g., levelling, total station surveying, GPS, photogrammetry) provide georeferenced coordinates. In contrast, non-geodetic methods (e.g., strain gauges, extensometers, fibre-optic sensors) typically measure local relative changes such as strain, tilt, or temperature (Erol et al., 2004). Monitoring strategies are often classified as permanent, semi-permanent, or epoch-based. Permanent systems continuously record sensor data, semi-permanent systems acquire readings at regular intervals, and epoch monitoring uses repeated geodetic surveys at discrete epochs separated by days or weeks. The choice depends on the expected deformation rate, structural importance, and available resources (Lim et al., 2011). Conventional geodetic methods, including traversing, triangulation/ trilateration, intersection, and precise levelling, have long been used in deformation surveys because they provide redundant measurements, internal checks, and high accuracy, though they may be labour-intensive. Photogrammetry and remote sensing enable simultaneous monitoring of multiple points, whereas space-based methods such as GPS and InSAR are increasingly used for large-scale deformation problems (Chen, 1983). For bridges, deformation monitoring is critical for understanding structural behaviour under traffic and environmental loads, verifying design assumptions, and planning maintenance. Temperature-induced expansion and contraction can cause significant daily and seasonal movements in steel and concrete elements, which must be distinguished from long-term structural deterioration. In this context, the current study employs an epoch-based geodetic monitoring scheme on the Pungu-Gumongo Bridge, integrating GPS, levelling, and total station measurements to record both horizontal and vertical movements and to investigate the influence of temperature changes.

3. Study Area and Monitoring Design

The Pungu-Gumongo Bridge is located between Pungu and Gumongo in the Upper East Region. It is a road bridge constructed to enhance safety and facilitate movement among communities. The bridge features reinforced concrete supports and a steel superstructure that carries a vehicle roadway. An epoch monitoring approach was used. Measurements were taken every two weeks for approximately three months, yielding three main epochs. For epoch 1, only morning observations were recorded; epochs 2 and 3 included both morning and afternoon sessions to capture possible temperature-related changes. The monitoring network was designed as an absolute network, with control points located outside the deformation zone serving as stable reference stations. These control points were positioned to be mutually visible, located sufficiently far from the bridge to prevent shared deformation, and close enough to ensure good geometry and reduce atmospheric effects. Two permanent control pillars were constructed near the footbridge and labeled in accordance with the department’s permanent benchmark scheme. Target points were placed at likely deformation-sensitive areas of the bridge, including pillars and joints where structural loads transfer to the foundation, joints between steel plates on the deck, and selected points on the front face of the bridge for reflector less total station measurements. In total, four key target points (TX1–TX4) on the main structure and numerous leveling points along the driveway were monitored.

MATERIALS AND METHODS

Trimble GPS receivers were used for establishing control coordinates and monitoring selected targets, a Trimble total station for three-dimensional coordinate determination of target points using reflector less mode and resection, a Dini level and staff for high-precision vertical deformation monitoring along the bridge, and prisms, tripods, measuring tapes, and paint for marking and maintaining survey points and instrument setups. Control stations were first established and coordinated using GPS. SG1 and SG15, previously known stable points, served as the base and rover for GPS observations, with sessions lasting at least 120 minutes per station. A test of beacons (point-of-departure method) was carried out using total station observations to verify the stability of these control points. Distances and angles between control points were measured and compared with values derived from known coordinates, confirming the network's stability. Spirit leveling was employed to monitor vertical movements of points along the bridge, with balanced backsight and foresight distances to minimize systematic errors. Leveling was repeated at each epoch, with separate morning and afternoon observations where possible, and height differences were reduced and adjusted using standard procedures. For horizontal and vertical monitoring of the four critical points (TX1–TX4) on the bridge's front face, the resection method was used. The total station was set up on the two control stations, and angles and distances to the targets were measured. For each target point and coordinate component (height, Easting, Northing), differences between epochs were calculated and compared with one-third of the standard error of the differences. If the absolute value of a difference exceeded this threshold, the point was considered to show significant movement. Outlier analysis at a 95% confidence level found no gross errors in the observed data.

RESULTS

Levelling along the bridge driveway showed measurable height changes between the three observation periods. Several points exhibited differences between epochs 1, 2, and 3 that surpassed one-third of the corresponding standard errors, indicating statistically significant vertical movement rather than random noise. Generally, height differences were a few millimeters, with some locations showing slightly more uplift or settlement than in the initial epoch. Total station height measurements for the four critical targets on the bridge's front face (TX1–TX4) confirm this pattern. Tables 1 and 2 list the epoch-wise vertical coordinates for the morning and afternoon sessions, while Figures 1 and 2 display these values for visual comparison. Across all four points, small but systematic changes are observed between epochs, with cumulative vertical differences of a few millimeters relative to epoch 1. For epochs 2 and 3, the afternoon heights are generally slightly higher than the corresponding morning heights at the same epoch, suggesting a modest vertical expansion of the steel superstructure during the warmer afternoon periods. Most of these height differences exceed one-third of their standard errors, indicating significant vertical deformations rather than measurement scatter.

Table 1: Epoch-wise vertical coordinates of monitoring points TX1–TX4 from total station observations (morning sessions).

Table 2: Epoch-wise vertical coordinates of monitoring points TX1–TX4 from total station observations (afternoon sessions).

Horizontal coordinates derived from total station resection also indicate measurable movement of the bridge's front face. Tables 3 and 4 summarize the Easting and Northing coordinates for TX1–TX4 across the three epochs, along with displacement statistics. When Eastings are referenced to epoch 1, the east–west components (ΔE) for individual points reach approximately 2–6 mm by epoch 3, with all four points showing non-zero shifts. The corresponding Northing differences (ΔN) are generally smaller but still at the millimeter level. In all cases, the observed coordinate differences exceed one-third of the standard error of the differences, so each target is classified as “Moved” according to the adopted significance criterion.

Table 3 Epoch-wise horizontal coordinates (Eastings) of monitoring points TX1–TX4 from total station resection (morning sessions).

Table 4 Epoch-wise horizontal coordinates (Northings) of monitoring points TX1–TX4 from total station resection (morning sessions).

Figure 1: Total station vertical coordinates comparison at different epochs, morning session

Figure 2: Total station vertical coordinates comparison at different epochs, afternoon session

Figures 3 and 4 show ΔE and ΔN as functions of epoch for TX1–TX4. The ΔE plots reveal a consistent east–west displacement trend for all points, with some variation in magnitude among targets. The ΔN components, meanwhile, remain relatively small and show no strong systematic trend. This indicates that the main horizontal movement of the monitored part of the bridge occurs along the east–west axis, while north–south shifts are less significant.

Figure 3 East–West displacement components (ΔE) for TX1–TX4.

Figure 4 North–South displacement components (ΔN) for TX1–TX4.