5 steps to track ground surface changes at solar power plants using drone surveying

In solar power plants, gradual changes that are easy to overlook during daily operations can occur, such as post-construction ground subsidence, scour caused by drainage, ruts on access roads, slope failures, and changes to the ground surface after mowing or repair work. Visual patrols alone make it difficult to record the location and extent of such changes, and comparisons with the past tend to rely on the memory of the person in charge. Drone surveying, which repeatedly measures the same area and allows comparison as topographic data, is useful for this. In this article, we explain the practical workflow for tracking ground-surface changes at solar power plants using drone surveying, divided into five steps.

Table of Contents

• Determine the purpose of assessing the current situation and the scope of changes to track

• Set up flight plans and reference points so they can be compared under the same conditions.

• Perform imaging and 3D data generation capable of capturing the ground surface

• Organize the locations and amounts of changes compared to the previous data.

• Retain as records that can be used for inspection, repair, and reporting.

• Summary

Determine the purpose of understanding the current situation and the scope for tracking changes

When tracking ground surface changes at a solar power plant using drone surveying, the first thing to clarify is the purpose—why you are checking for changes. Simply photographing the site from above makes it difficult later, when reviewing, to determine "what changed," "to what extent it changed," and "whether any action is required." To make ground surface change management useful, it is important to narrow down the targets you want to check in advance and to accumulate records that are easy to compare.

Areas that are particularly likely to be checked at solar power plants include drainage routes, slopes, around retention ponds, access roads, beneath panel rows, boundaries of embankments and cuttings, and areas around buried cable routes. These locations are susceptible to the effects of rainwater flow, vehicle traffic, mowing operations, repair work, and soil compaction, and their condition can change over time. In particular, after periods of heavy rainfall or from immediately after site development through the early operational phase, subtle subsidence or topsoil erosion can progress, so there is value in performing periodic inspections under the same conditions.

However, if you try to check every area in detail every time, flight planning, data processing, and comparison work become burdensome and it becomes difficult to sustain. In practice, it is easier to operate if you separate surveys that broadly capture the entire power plant from surveys that inspect in detail the priority areas where changes are likely to occur. For example, in the overall check, roughly grasp the site's elevation differences, drainage directions, and the condition of access roads, while in the focused check, prioritize recording areas such as around drainage ditches, the toe and shoulder of slopes, and places prone to mud or ruts.

When setting objectives, also clarify whether surface changes will be used for "anomaly detection," "before-and-after comparisons of repairs," or "explanations to management companies and stakeholders." If anomaly detection is the purpose, it is important to repeatedly check a wide area at consistent times so that differences from the previous survey are easier to spot. If the purpose is before-and-after comparisons of repairs, you need to decide on comparison points—such as before construction, immediately after construction, and after a certain period has elapsed—and record them. If the data will be used as explanatory material, retaining a location map, overhead images, and close-up images of the changed areas that are easy to understand for non-specialists will make it easier to share the site conditions.

When assessing changes to the ground surface, it's also necessary to distinguish whether a change falls within the range of what occurs naturally or is a sign that requires action. On bare ground and grassy areas within a power plant, vegetation and the appearance of the surface change with the seasons. Immediately after rain, puddles and moisture can make the ground surface appear lower, and during periods when grass is growing, the ground itself can become difficult to see. Do not rush to judge based solely on drone survey results; it is important to interpret them in conjunction with the time of imaging, weather, mowing conditions, and findings from on-site inspections.

Also, if drawings of the power plant or past development documents are available, checking them at an early stage makes it easier to establish a baseline for comparison. Understanding the drainage plan, development plan, slope geometry, road plan, panel layout, and so on will make it easier to interpret the significance of changes when you compare them with surface data obtained from drone surveys. Rather than looking only at the current shape, deciding the survey extent based on "what the terrain should originally be," "the expected direction of water flow," and "where vehicles are planned to pass" leads to survey results that are useful in practice.

What you should avoid in this procedure is conducting flights and capturing imagery before first defining the objective. Because drones can cover large areas in a short time, it can be tempting to fly first and decide later. However, when tracking ground surface changes, if the area you need wasn’t captured or the reference points required for comparison aren’t present, you may need to re-fly or reprocess the data. During power plant operations there are constraints such as access restrictions, working hours, weather, and safety management, so deciding the objective and scope up front will ultimately improve efficiency.

Arrange flight plans and reference points so they can be compared under the same conditions

In drone surveying that tracks changes to the ground surface, it is important to ensure that each dataset can be compared using the same reference frame. If the survey is conducted only once to capture the current conditions, its main objective is to clearly record the terrain at that time. However, when tracking changes, it is necessary to minimize positional and elevation discrepancies when overlaying the initial, subsequent, and later datasets. If this is unclear, differences may appear even though the ground has not actually changed, or small changes may be overlooked.

In flight planning, keep the capture area, flight altitude, capture direction, image overlap, capture time, and flight route as stable as possible. It can be difficult to make conditions exactly the same every time due to site conditions, but when planning for comparisons, ensure at least that the image density and angles over the important areas do not change significantly.

On slopes and drainage ditches where there are inclines or steps, images taken directly overhead may be insufficient to interpret the shape. Therefore, in areas where you want to understand surface elevation differences and the condition of the sides, combining oblique images as needed makes the records of change easier to use.

Establishing reference points is also important. In drone surveying, how reference points are handled is important for aligning the three-dimensional data and orthophotos created from images with local coordinates. At solar power plants, sites are often large and rows of panels, fences, roads, and slopes are arranged in a complex manner, so if the alignment conditions change with each comparison, judgments about changes to the ground surface become unstable. When placing control points or check points on site, choose locations that are unlikely to move due to vehicle traffic, mowing, or runoff, and record them so they can be verified in future surveys.

With reference points, it is more important to continuously manage which point is used for which purpose than merely that they were installed. For example, distinguishing points used for coordinate referencing, points used for accuracy checks, and points used as markers during site inspections makes it easier to avoid confusion during data processing. Even when using the same point every time, verify that the point has not moved, that sight lines are maintained, and that it is not obscured by surrounding vegetation or materials. If a reference point can no longer be verified on site, an alternate point will be used, but you must record that the comparison conditions have changed.

Before flight, we also verify on-site safety conditions. At solar power plants, attention to panels, racking, electrical equipment, fences, power lines, surrounding roads, and adjacent land is essential. We also check the movement paths of workers and vehicles within the plant, scheduled inspection work, and whether mowing or repair work is planned, and coordinate so that drone flights do not overlap with on-site operations. When surveying ground surface changes, it may be tempting to capture images immediately after rain or a typhoon, but if the ground is muddy, entering the site or confirming control points can be hazardous. Prioritize safety and consider this as part of the overall on-site inspection plan, not just the flight.

The timing of image capture also affects the accuracy of comparisons. When grass is growing versus immediately after mowing, the same location can look different on the ground surface. Fallen leaves, snowfall, mud, puddles, and the way shadows fall also influence how the data appears. If you want to track changes in the ground surface itself, it is easier to compare images taken after mowing or when the ground is clearly visible. On the other hand, if you want to understand poor drainage or the occurrence of puddles, it makes sense to check conditions after rainfall. Thus, the timing of image capture should be decided according to the purpose.

It is also advisable to retain flight plans and control point information so they can be reproduced even if the person in charge changes. In power plant management, multiple stakeholders may be involved, such as inspection personnel, surveyors, construction personnel, management companies, and clients. If the previous flight altitude and coverage, the arrangement of control points, and precautions taken during capture are not shared, comparisons with subsequent surveys will be difficult. By keeping work records of the capture date, weather, flight coverage, control points used, on-site conditions, and any special notes, drone surveying becomes continuous management documentation rather than a one-off record.

Perform imaging and three-dimensional data processing that can capture the ground surface

Once the flight plan and ground control points are set, perform the actual capture and convert it into data that allows comparison of the ground surface. When tracking ground surface changes at a solar power plant, it is important to produce purpose-specific deliverables—not just aerial photos but orthophotos, 3D point clouds, terrain models that organize surface conditions, cross-sections, difference maps, and so on. In particular, if you want to check ground subsidence, scouring, embankment deformation, or road surface irregularities, three-dimensional data that includes elevation information is useful.

When shooting, ensure that the ground surface is as visible as possible. Shadows tend to appear beneath panels and around the mounting racks, and if grass is overgrown the shape of the ground becomes difficult to discern. Because panel rows in a power plant are arranged regularly, similar patterns repeat in the images and position matching can become difficult. Therefore, in areas where ground-surface features are insufficient, it is important to appropriately place control points and markers and to secure sufficient image overlap. Data tends to become unstable at the edges of the shooting area and in locations with elevation changes, so capturing a slightly wider area than required will make later processing easier.

When creating 3D data, point clouds and terrain models are generated from images. One thing to be careful of is that the resulting data does not necessarily represent only the pure ground surface. If grass, materials, vehicles, temporary structures, or parts of panels or mounting frames are included in the data, they can be mistakenly treated as ground surface elevation. To track ground surface changes, unnecessary objects should be excluded as much as possible and the data organized so that the ground condition you want to compare can be read. When vegetation has a large influence, it is important to use site photographs and inspection records together to determine whether the ground itself has changed or only the condition of the vegetation has.

On the ground surface at a solar power plant, not only abrupt changes but also gradual changes can have management significance. For example, if soil slowly flows around a drainage ditch and accumulates at the toe of a slope, it may not appear as a major anomaly in the early stages. However, by repeatedly collecting data at the same location, it becomes easier to grasp the direction and extent of the change. Conversely, if temporary vehicle tracks or puddles are treated as terrain changes, they can appear to be a larger problem than they actually are. Three-dimensional data is useful, but it is essential to interpret it in the context of the site.

When processing data, choose the output format with the purpose of the deliverable in mind. If you are sharing an overview with managers, materials that use an orthophoto viewed from above and indicate areas of change are easier to use. If the data will be used for considering construction or repairs, materials that allow verification of cross-sections and elevation differences at the relevant locations are helpful. For drainage-related studies, materials that let you check the surface slope and the directions in which water is likely to flow are important. In other words, even with the same drone survey data, the way you process it needs to change depending on the viewer and the intended use.

When generating three-dimensional data, the limitations of the survey results should also be made clear. Drone surveying can efficiently capture large areas of the ground surface, but it may not adequately capture areas such as ground hidden beneath grass or structures, dark areas directly beneath panels, or the bottoms of narrow drainage channels. If necessary, it should be decided to combine drone surveying with on-the-ground verification or alternative surveying methods. In particular, for decisions that directly affect repair quantities or design changes, it is safer not to finalize based solely on drone survey results but to carry out on-site verification and any required accuracy checks.

In practical work, it is also important to prepare the initial data as reference data. If the initial imaging or processing is unstable, it will affect all subsequent comparisons. When creating the reference data, check for missing captures, positional shifts, errors in reference points, inclusion of unwanted objects, and insufficient processing coverage. In particular, when using it for managing an entire power plant, clearly record the date, imaging conditions, target coverage, coordinate conditions, and processing conditions in the initial data to make future comparisons smoother.

Organize the locations and magnitudes of changes compared with the previous data

The most important step in tracking changes to the ground surface is comparing the data with the previous survey. The strength of drone surveying lies in repeatedly capturing data over the same area and being able to confirm visual and numerical differences between past and present. However, producing only the differences does not provide information usable in practice. Only when the change’s location, extent, depth and height, time of occurrence, possible causes, and response priority are organized can it be used to inform inspection and repair decisions.

In comparison work, first confirm the alignment between the previous dataset and the current dataset. If the entire dataset is shifted in the same direction, it may not be that the ground surface has changed but that there is an alignment error between the datasets. In particular, when the conditions of control points have changed or only part of the captured area was processed, the difference map can show a broad-area shift. Making judgments about areas of change in this state can lead to incorrect conclusions. Before looking at the differences, check places that should not have moved—such as sections of paved road or areas around fixed structures—to see whether the reference for comparison has shifted significantly.

Next, classify the locations of change. Surface changes at a solar power plant can include subsidence, uplift, erosion, deposition, ruts, slope failures, sediment movement around drainage channels, and shape changes caused by repair work. In difference maps, changes may be represented by colors or shades, but do not judge based solely on visual impressions; compare them with on-site conditions. For example, linear changes on a road may be ruts caused by vehicle traffic. Scouring on the downstream side of a drainage ditch may indicate erosion from flowing water. A bulge at the toe of a slope may be deposition of sediment that flowed from above. In this way, infer the cause from the shape and location of the differences and, if necessary, proceed to an on-site inspection.

When organizing changes, care is required in handling numerical values. Elevation information obtained from drone surveys includes variability due to shooting conditions, how the ground surface appears, control points, and processing conditions. Therefore, it is dangerous to treat every small difference as a ground change. In practice, consider whether the survey conditions are appropriate for the magnitude of change you want to confirm, and handle minute differences cautiously. What is important is not to rely excessively on single measurements, but to comprehensively assess whether change is continuing at the same location, whether the affected area is expanding, and whether drainage or structures are being impacted.

Organize the comparison results in a way that makes priorities clear. For example, changes that affect drainage function, changes close to panel mounts or cable routes, changes that impede passage on access roads, and changes related to slope stability should be checked promptly. On the other hand, differences caused by grass condition or temporary surface disturbances should not immediately be treated as repair targets; it may be appropriate to monitor their progress at the next inspection or during site patrols. Drone survey comparison results are not intended to treat everything as a problem, but are most effective when used as material to decide where to focus limited maintenance resources.

When there are multiple past datasets, it is useful to look at trends over time rather than just a simple comparison with the previous one. Separating changes from the first to the current measurement, changes before and after the rainy season, changes before and after repairs, and differences in appearance before and after mowing makes it easier to judge whether surface changes are temporary or are progressing continuously. For example, if scour on the downstream side of the same drainage route is getting slightly deeper each time, even small changes in isolation may require a review of drainage plans and sediment runoff countermeasures.

When preparing comparative materials, be mindful to use expressions that on-site personnel can easily verify. Even if you provide only specialized three-dimensional data, not all stakeholders will necessarily be able to interpret it the same way. Indicate locations on an overall view, show the shape of changes in enlarged views, and add cross-sections or explanatory notes where necessary to make decision-making easier. This is especially important when the power plant’s management company, construction contractor, maintenance personnel, and owner are different entities; creating materials that establish a shared understanding of the locations of change is crucial. By clarifying at which points, from when, and to what extent changes have occurred, on-site inspections and repair discussions can proceed more smoothly.

Keep records that can be used for inspection, repair, and reporting

When you confirm changes to the ground surface through drone surveying, record the results as documentation that can be used for inspection, repair, and reporting. Investigating changes to the ground surface is not finished by simply photographing and comparing differences. Rather, the important thing is to connect the findings to follow-up actions and store them in a form that can be used for comparisons in the future. If records are kept inadequately, the valuable survey results will become one-off materials and will be difficult to use for long-term management.

First, organize for each change point the location, details, inspection date, summary of the change, on-site photos, and the current response status. For the location, it is helpful to record it together with expressions that are easy to share on site, such as the power plant’s site map or plot name, panel row number, road name, and drainage system. Coordinates alone may not allow field personnel to immediately identify the location. On the other hand, verbal descriptions alone may lack precision. Therefore, using both coordinate information and names familiar to on-site staff can reduce misunderstandings during inspections or repairs.

When proceeding to repairs, it is also important not to be too quick to assume the cause of changes. From comparative results of drone surveys, it is possible to read the potential for settlement, scour, or deposition, but confirming the cause may require on-site inspection, construction history, drainage conditions, and verification of ground conditions. In reports, before asserting "scour has occurred," it is safer and more practical to use expressions such as "a lowering of the ground surface is observed around drainage routes" or "confirm the presence or absence of scour through on-site inspection." By avoiding definitive conclusions beyond what is necessary, misunderstandings among stakeholders are easier to prevent.

Organize records so they can be tracked in chronological order. For surface changes, it is important not only to note when they occurred but also to determine whether they have progressed afterward. For example, if a small collapse is found on a slope, you need to check whether it has expanded after subsequent rainfall or whether it has stabilized after repairs. Manage drone survey data by capture date and save comparative materials for the same area, which makes it easier to explain past conditions. This also helps when explaining the need for repairs or when verifying the effectiveness after construction.

When keeping records as management documentation, store the raw data and the deliverables separately. Raw images, control point information, processing parameters, point clouds, orthophotos, difference maps, and reporting materials each serve different purposes. Data that may be needed for future reanalysis or additional comparisons should be retained rather than deleted. Include information in filenames and folder structures—such as the power plant name, acquisition date, coverage, purpose, and processing stage—so files are easier to find later. This is especially important when managing multiple power plants: if naming rules are ambiguous, simply locating the required data can take a lot of time.

In reports, rather than presenting specialized survey results as-is, organize them into the information needed for management decisions. Stakeholders want to know what has changed, how significant it is, and what should be done next. Therefore, concisely summarizing the overall view, locations of changes, whether an on-site inspection is required, whether repair should be considered, and the schedule for the next inspection makes the report more useful. At solar power plants, not only the generation equipment itself but also the surrounding environment—site, drainage, roads, and slopes—affects long-term operation. Records of surface changes also contribute to revising maintenance plans and to preventive maintenance.

Furthermore, linking the results of drone surveys with on-site inspections improves management accuracy. By extracting areas suspected of change from drone surveys and establishing a workflow to focus checks on them during inspections, you can inspect large sites efficiently. Conversely, unusual conditions noticed during inspections can be prioritized for focused verification in the next drone survey. Combining wide-area aerial awareness with detailed ground-level checks reduces oversights and makes it easier to determine response priorities.

When keeping records, it is useful to note the conditions for the next survey as well. If you record when the next flight will take place, which areas will be emphasized, whether the same reference points will be used, whether photos will be taken after vegetation cutting, or whether checks will be made after rainfall, ongoing management becomes easier. Tracking changes in the ground surface is more valuable when repeated with the same approach than from a single survey. By reflecting survey results in the next plan, drone surveying becomes established within power plant management.

Summary

To track ground-surface changes at solar power plants with drone surveying, it is important to consider the whole workflow—not just imaging techniques but also objective setting, comparison conditions, data organization, on-site verification, and reporting operations. The plant’s ground surface gradually changes under the influence of rainwater, wind, vehicle traffic, mowing, repair work, and site formation/ground conditions. These changes can be hard to notice by visual inspection alone in the early stages, but if the same area is continuously recorded by drone surveys, it becomes easier to grasp their location and extent.

To achieve results in practical work, first clarify the area and purpose you want to check, and establish flight plans and reference points that make comparison easy. Then capture imagery in a condition that allows the ground surface to be read, and organize it as three-dimensional data and orthophotos. When comparing with previous data, do not judge solely by the visual appearance of differences; interpret them taking into account alignment, imaging conditions, the effects of vegetation and shadows, and on-site conditions. Record confirmed changes as documentation that can be used for inspection and repair, and incorporate them into the conditions for the next survey.

Drone surveying is an effective means to efficiently grasp the extent of a solar power plant site and to continuously manage changes to the ground surface. However, survey results are not万能 (not omnipotent), and their value increases when used in combination with on-site verification and management decision-making. If you want to detect ground subsidence, scour, poor drainage, road deterioration, or slope deformation early, operational design that assumes comparisons rather than one-off flights is important. If you want to make plant maintenance management more visible and link it to decisions about upcoming inspections and repairs, consider a drone surveying approach that is easy to operate continuously while confirming site conditions, required accuracy, safety management, and consistency with existing documentation.