Share | 03/24/2020
Satellites and airborne multispectral cameras are not opposing technologies, in fact, the two sources of information are used to feed similar applications and their data can even be complementary. For instance, Aerobotics software Aeroview, combine both satellite and drone imagery to provide in-depth and varied data for their clients.
Both satellites and multispectral cameras are tools used in remote sensing applications. Objects on the earth’s surface absorb and reflect light from the sun. The light reflected is what satellites and sensors capture, which can be visible or invisible. Human eyes are designed to detect only visible light, however, multispectral cameras and satellites can measure both the visible and invisible bands of the electromagnetic spectrum.
Satellites observe the earth and provide data on how it changes over time. Multispectral cameras are used to monitor not the whole earth, but specific chunks of land, and also provide data on how those areas change over time. The data collected with both technologies is commonly used for the classification and mapping of vegetation, being cheaper and less time-consuming than manual field surveys.
Aeroview software processes both satellite and multispectral imagery to provide precise per-tree health analytics and highlight problem areas that could otherwise go undetected. With this vegetation map, farmers are able to keep track of underperforming trees and make actionable decisions on their farms.
With 3 to 10 bands, multispectral sensors capture data within specific wavelength ranges across the light spectrum, typically including green, red, and near-infrared. The combination of bands in a sensor will determine its uses and applications which are usually related to changes in land use, vegetation mapping, monitoring of natural resources, besides others.
The MicaSense series RedEdge-P, for example, utilizes five bands (red, green, blue, NIR, and red edge) to collect data used in applications like vegetation health mapping, disease detection, and irrigation management.
These capabilities are not exclusive to airborne sensors, some satellites also use multispectral sensors to capture data from the earth’s surface.
Satellites are used to monitor the earth and collect data on different natural phenomena. The specific applications vary widely depending on the instrument used and the altitude of the satellite orbit. Landsat and Sentinel are the most common and well-known satellite programs in the remote sensing community and their historic data is freely accessible.
Landsat: for monitoring climate change, urbanization, and wildfires
Launched in 1972, the Landsat program consists of six satellites of which two are still in orbit (Landsat 7 and 8). Landsat 8 operates over 9 bands (coastal, blue, green, red, NIR, SWIR-1, SWIR-2, panchromatic, cirrus, and two thermal bands) and has a ground resolution of 30 meters. The Landsat 8 captures imagery used to monitor changes on the earth’s surface and the relationship of those changes with other phenomena like climate change, urbanization, and wildfires. It is also used to identify and track the movement of pollutants such as oil slicks.
Sentinel: for land, ocean and atmospheric monitoring
Sentinel satellites carry a range of technologies such as radar and multispectral sensors. Aeroview software makes use of the Sentinel-2, which is currently in orbit and collects imagery at a higher spatial resolution (10m to 60m). It covers 13 wavelengths of the visible light spectrum, plus near-infrared and shortwave infrared, providing imagery used in vegetation mapping, soil & water monitoring, and detection & tracking of inland waterways and coastal areas.
The main differences between satellites and multispectral cameras lay in the cost, the accessibility of the information, the spectral and spatial resolution, and the control someone can have over atmospheric phenomena like clouds.
Historic satellite data is available for free, allowing users to compare changes in the surface of the earth through decades at no cost. However, in places where the weather conditions limit the number of cloud-free days, satellite data may be incomplete.
The spatial resolution of satellites can also be an issue on certain applications, like plant phenotyping. Both Landsat and Sentinel data have a ground resolution lower than 10m/pixel, while multispectral cameras have a resolution higher than 10cm/pixel.
Click here to compare the spatial resolution of our Dual Camera System and the Sentinel 2A.
Aerial photography has higher resolution but is more expensive per square meter.
The preference for one or the other depends on the applications for which the imagery is needed, but that does not mean that only one source can be used at a time since satellite and aerial imagery can be complementary. Satellite data has a longer historic record than drone-based sensor imagery and because the bands captured by both technologies are different, some issues can be more evident with satellite imagery than aerial imagery and vice-versa.
Aerobotics, our partner in South Africa, provides data analytics solutions using both satellite and multispectral aerial imagery to help farmers identify pests and diseases early.
Sign up to receive updates directly to your inbox.
