Satellite imagery offers a powerful tool for understanding our planet, but the abundance of vendors and specifications can be overwhelming. This guide offers a high-level overview of visible spectrum satellite data, and walks you through the key factors to consider as you select the imagery that best fits your use case.  Whether you're monitoring pipeline integrity, optimizing solar farm placement, or tracking changes in biodiversity, understanding the right satellite imagery specifications is crucial to select the best data for your use case.

Establish your satellite imagery use case 

Before diving into the technical details, clearly define your project objectives. 

• What are you trying to achieve with satellite imagery? 
• What specific features or phenomena are you interested in observing? 
  • For example, analyzing individual buildings in an urban environment requires different considerations than monitoring large-scale deforestation.
• Is your use case static (i.e. a single point in time) or dynamic (i.e. monitor change over a time period)?

Once you’ve established these objectives, you can start to select the best satellite imagery based on the following criteria.  

Spatial resolution - what level of detail do you need?

Spatial resolution is the cornerstone of imagery selection. It dictates the size of the smallest object you can identify. For use cases requiring higher scalability and less detail, open-source options like Sentinel-2, Landsat-9 (the most recent satellite in the Landsat program), and GOES-R offer medium and low-resolution imagery. 

GOES-R offers low spatial resolution data at 0.5 - 1.0 km/pixel for visible spectrum bands, which is not suitable for object detection or habitat segmentation, but can be used for atmospheric conditions monitoring. Sentinel-2 (10m/pixel) and Landsat-9 (30m/pixel) provide medium-resolution visible spectrum bands and have been used for land cover mapping. 

Higher resolution images, like those from WorldView-4 (30cm/pixel), SkySat (50cm/pixel), Maxar (30cm/pixel and 15cm/pixel), and National Agriculture Imagery Program (NAIP) (60cm/pixel), allow for detailed analysis and are essential for applications like urban planning or precision agriculture. They reveal intricate details, enabling you to distinguish individual vehicles or trees. However, high resolution often comes with a high associated cost. It’s best to establish the lowest resolution needed for your use case (see more in the Balancing budget with data needs section). 

Figure 1: A comparison of different spatial resolutions of a Landsat-8 image of Reykjavik, Iceland, taken on July 7, 2019. Credit: NASA Earth Observatory.

Temporal resolution - how often an image is captured

Temporal resolution refers to the frequency at which a satellite captures imagery of the same location on Earth. This is a crucial factor in remote sensing, as it determines how often you can obtain updated images of a particular area. The higher the revisit frequency, the higher the temporal resolution; and vice versa. 

Understanding the required temporal resolution of your use case helps you select the right satellite products for specific project needs. For example, some projects demand freshly acquired imagery, particularly when monitoring rapidly changing events like natural disasters or crop growth. Despite its low spatial resolution, GOES-R offers the highest temporal resolution among the free satellite imagery sources at 15 minutes, making it ideal for a wide range of weather, oceanographic, climate, and environmental applications (Schmit et al., 2005). Daily revisits, offered by PlanetScope and WorldView-4, also prove invaluable for use cases requiring tracking daily changes (e.g. urban development, disaster response, agriculture, etc.) Less frequent revisits might suffice for projects focused on long-term trends, like glacier retreat or coastal erosion. NAIP provides imagery with a temporal resolution of no more than three years, making it suitable for regional analysis and long-term monitoring. Ultimately, the diverse range of temporal resolutions available enables you to tailor your satellite data selection to your specific use case. 

Figure 2: A comparison of seasonal changes across February, April, July, and October of Lake George, New York, USA taken using NASA’s Terra satellite. Credit: NASA Jet Propulsion Laboratory

Balancing budget with data needs

Satellite imagery pricing varies significantly, ranging from free (e.g. Sentinel-2) to thousands of dollars for high-resolution, on-demand acquisitions. The price adds up for use cases like environmental monitoring, which may require multiple images of the same area over a designated period of time (i.e. a high temporal resolution). Table 1 shows cost estimates of 1000 sq km of data from a variety of popular satellite imagery providers. It’s worth noting that 1000 sq km could be a one time purchase of imagery for a 1000 sq km area, or could represent a 100 sq km area with 10 revisits to that site. You’ll notice that as spatial resolution increases, the price increases exponentially. This magnifies the importance of selecting the lowest spatial resolution that still meets your project requirements.dis

*Price is based on published pricing from Apollo Mapping and LandInfo for archive imagery (typically >90 days old). For exact pricing, please reach out directly to a vendor. Please note that real-time images can be almost double this cost versus the data archive. More information about specific satellites and vendors can be found at Apollo Mapping and Landinfo.

Table 1: Satellite imagery pricing from common providers

Historical data coverage - understand recent changes

Another consideration when selecting your optimal imagery is whether you require a past look. For use cases such as glacial monitoring, it’s best to have existing historical data to feed into the analyses. The Landsat missions provide the longest historical context as the first satellite launched in 1972. In contrast, NAIP data has been available since 2003 and is collected at 2-3 year intervals. Open source Sentinel-2 data dates back to 2015, providing a ten year look-back for monitoring use cases, whereas newer Satellies like the Pléiades Neo from Airbus only have historical data since its launch in 2021 (Apollo Mapping, Pléiades Neo).

Use cases for satellite imagery

Circling back to spatial and temporal resolutions, we summarize multiple potential monitoring projects categorized by different spatial and temporal resolution combinations in Table 2. Each combination of spatial and temporal resolution is associated with specific monitoring applications, highlighting the versatility of satellite data in diverse fields.

Table 2: Applications of satellite imagery for potential monitoring projects, categorized by different spatial and temporal resolution combinations. 

In Conclusion

By carefully evaluating these factors and aligning them with your specific project goals, you can unlock the power of satellite imagery and gain valuable insights into our ever-changing planet. We hope you’ve found this guide useful as you begin your satellite imagery journey. 

…Next time

You probably noticed the exponential costs of high-resolution satellite imagery in this post - tune in next time for a technical solution to improve the resolution of free, open source satellite data using AI/ML. 

Watch our webinar

Learn how to uncover the full potential of satellite imagery for your business. Kira Dickey, Manager of Innovation Science, showcases practical applications of ML based super resolution technology in industries like energy, urban planning, and beyond.

>> Watch on our YouTube channel

References

Satellite Imagery Providers: 

●    Sentinel-2
●    Landsat
●    Landsat
●    GOES-R
●    Planet
●    NAIP
●    PlanetScope
●    Airbus
●    Maxar WorldView-4
●    Maxar

Other linked websites: 

●   USGS: How often is orthoimagery in The National Map updated and what are the acquisition dates

Images:

●    NASA Earth Observatory: comparison of different spatial resolutions of a Landsat-8 image of Reykjavik
●    NASA Jet Propulsion Laboratory: comparison of seasonal changes across Lake George, New York, USA taken using NASA's Terra satellite

Data Vendors and Pricing Sources:
●   Apollo Mapping
●   Landinfo
●   Up42 SPOT tasking

Academic journals:

●    Granero-Belinchon, C.; Adeline, K.; Lemonsu, A.; Briottet, X. 2020: Phenological Dynamics Characterization of Alignment Trees with Sentinel-2 Imagery: A Vegetation Indices Time Series Reconstruction Methodology Adapted to Urban Areas. Remote Sens, 12, 639. 
●    Pena-Regueiro, J.; Sebastiá-Frasquet, M.-T.; Estornell, J.; Aguilar-Maldonado, J.A. 2020: Sentinel-2 Application to the Surface Characterization of Small Water Bodies in Wetlands. Water, 12, 1487. 
●    Schmit, T. J., M. M. Gunshor, W. P. Menzel, J. J. Gurka, J. Li, and A. S. Bachmeier, 2005: Introducing the next-generation advanced baseline imager on GOES-R. Bull. Amer. Meteor. Soc., 86, 1079–1096.