Satellite Imagery: An Aerial Alternative - 25/08/2020
Remote sensing projects often begin with the question: 'Should I use aerial imagery or satellite imagery?' During long-term projects, this question may arise again, particularly when unforeseen circumstances change the ability to collect data, the reliability of the data or the scope of the work. Budgets, technology and government restrictions constantly change, making it critical to choose the right data inputs.
There are a number of key parameters that must be taken into account when beginning any remote sensing project. These include the size and accessibility of the area of interest, the timeliness of the data, whether the data can be sourced from existing imagery or if a new collection is required, and perhaps the most important aspect: the project budget.
Efficiency and Scale
Satellites can complete remote sensing projects more efficiently than aircraft, resulting in enormous impacts on cost-benefit analysis. The key difference is the speed and ease with which satellites can collect thousands of square kilometres in minutes without the complicated logistics of aircraft flight planning. Aircraft face greater restrictions: they must obtain airspace permits, plan for suitable take-off and landing points, and adhere to ever-changing border and travel restrictions. Aircraft are also vulnerable to weather conditions such as heavy winds. Satellites simply do not have any of these issues. They can collect data from isolated, conflicted or cross-border locations with ease. This applies particularly to large-scale mapping projects that may require multiple flights for manned or unmanned aircraft. These savings are significantly important in budget/time-sensitive government mapping projects.
Tasking and Processing
The planning of satellite tasking is fully customizable. This allows users to prioritize their areas for collection, define the resolution and spectral bands as well as specify collection angles. There is added flexibility for complex projects to adjust these requirements shortly before the acquisition takes place. Real-time weather updates ensure that the data acquisition will be as cloud-free as possible, further narrowing the competitive gap between aerial data and satellite imagery. After collection, satellite imagery is directly downloaded through a ground station where it can be delivered to the user within hours of collection. Users can choose from several processing options and delivery methods.
Satellite imagery providers can collect data in various multispectral band combinations as well as stereo imagery in a single pass, eliminating the need for multiple flights by multiple specialized aircraft. Stereo imagery offers reliable data for the creation of Digital Elevation Models (DEM) and virtual 3D reconstructions. The suitability of 30cm satellite imagery for aerial imaging applications is confirmed by the National Imagery Interpretability Rating Scale (NIIRS), which is used by the imaging community to define and measure the quality of images and performance of imaging systems. Imagery captured in 30cm from Maxar’s WorldView-3 has a rating of NIIRS 5.7. This means that it is possible to identify objects such as above-ground utility lines in a residential neighbourhood, impervious surfaces, crop species and their boundaries, vehicle types, manhole covers and much more.
Satellites can reach areas of interest that are difficult to reach or inaccessible by other means and offer predictable and frequent refresh schedules. With high frequency refresh rates, users can confidently request the exact same area of interest to be collected at specific intervals. This is a crucial feature for automated analysis. As with aerial data, satellite imagery can also be integrated into programmes using artificial intelligence to automatically extract and classify features and thereby streamline workflows. The amount of imagery collected over time by satellites compared to aerial offers increased training data for machine learning programmes. Additionally, users can take advantage of historical data to model predictive analytics that are incredibly useful for trend analysis, anomaly detection at a mass scale and profitability estimates.Last updated: 28/08/2020