RICS Draft Global Guidance Note: Earth observation and aerial surveys, 6th edition

RICS Draft Guidance Note: Earth observation and aerial surveys, 6th edition

7 Earth observation

Earth observation from satellite platforms offers the advantage of covering large areas, up to 1,000,000km2 every day.

Sensors on-board satellite platforms tend to have either a push broom scanner, also known as an along track scanner, or a whisk broom scanner, also known as an across track scanner. Both types of scanner can produce both mono and stereo imagery. Whisk broom scanners have a greater number of moving parts, which tend to make the design of these instruments heavier, more expensive, and more prone to wearing out than their push broom counterparts. However, the whisk broom design does have the potential to offer a better spatial resolution.

Scanner-based imagery must be flown in a continuous swathe with a minimum of 20% overlap (25% in elevated or urban areas).

The key decision is whether the imagery specification can be met by the supply of tasked imagery, to be captured at some point in the future, or from previously captured imagery from the existing archives of satellite imagery providers. With tasked imagery, the same point on the earth's surface can potentially be revisited daily, offering monitoring solutions as well as applications in mapping, change detection and responding to environmental disasters.

7.1 Types of imagery

Visible, radar, and multispectral sensors are the most common sensors employed from satellite platforms. It is common for individual satellites to carry multiple sensors, including panchromatic, RGB and NIR sensors and additional multispectral options.

Radar imaging relies on an active sensor, emitting electromagnetic radio waves. Unlike optical methods that measure the wave amplitude, radar sensors measure the phase of the backscattered radio waves and can therefore operate in the dark and in all weather conditions.

Modern satellite sensors offer spatial resolutions of between 0.35m and 1.5m GSD for panchromatic sensors and from 1m to 6m GSD in the multispectral bands.

7.2 Flying and coverage

7.2.1 Satellite orbits

Earth observation satellites typically operate in sun synchronous geostationary or polar orbits. Geostationary orbits are at altitudes of approximately 36,000km, where the satellites take approximately 24 hours to orbit the earth. This enables continent-wide areas to be monitored continuously for environmental conditions or weather patterns.

Polar orbits are the most common orbits for remote sensing applications. In a polar orbit, the satellites pass over the earth's polar regions within 20 to 30 of the poles. Typically passing over the polar regions several times a day, they are at an altitude of between 200 and 1,000km from the earth's surface. The cameras onboard the satellites can be tasked with capturing either nadir imagery or off-nadir imagery.

When using off-nadir imagery, the imagery sensor is rotated so that it views the AOI from the side, rather than waiting until the satellite is directly overhead. This reduces the length of time that will elapse between observations, known as the satellite revisit time. However, because of the greater distance between the sensor and the surface of the earth, off-nadir imagery results in poorer resolution than nadir imagery.

7.2.2 Acceptable quality limits

The following list is intended to act as a set of AQLs to provide guidance on the subjective topic of image quality. The client and contractor should work closely together to ensure a mutually acceptable result.

  • The specified coverage and imagery accuracy requirements should be met.
  • The imagery should be sharp.
  • Colour, contrast and light balance should be uniform across the whole AOI. This is particularly true for stereo photography.
  • The time lapse between stereo pairs should not be too great to affect the quality of the stereo models.
  • The imagery should only be accepted if it is substantially free of cloud, dust, atmospheric haze dense shadow or smoke. Isolated areas of cloud, dense shadow or smoke should not be cause for rejection of the imagery provided the intended use is not impaired. Typical tolerances for cloud and cloud shadows may be less than 5% or 10% in a single image.
  • The photography should conform to any specific radiometric values specified by the client, including:
    • mean histogram luminosity values
    • mean of the individual colour bands and
    • standard deviation for each colour band.

The photography may be captured at any time when the weather conditions are suitable to achieve the specified standards of image quality, except where special time constraints are defined.

Earth observation satellites operate on specific orbits and cross the equator at the same time every day. It is therefore not possible to capture data over a target at a specific time of day. The intended use of the imagery may impose limitations upon times of capture. For earth observation imagery, it is common to specify winter or seasonal imagery capture, acquiring photographs when trees are not in leaf, or during a part of the growing season.

7.3 Earth observation accuracy and resolution table

Table 8 shows the achievable accuracy and resolution values for visible satellite imagery.

Platform

Height AGL

Achievable accuracy (m) (RMSE figures, at 1 sigma)

Achievable resolution - GSD (m)

m

Ft

Plan X,Y

Height Z

Satellite imagery

450- 770 km

279- 478 mi

3-4m (CE90)

3-4m (LE90)

0.35- 0.8

Table 8: Achievable accuracy and resolution values for satellite imagery

Achievable accuracies are quoted as circular and linear errors, at the 90% confidence level.

Table 9 shows the achievable resolution values for satellite imagery outside of the visible spectrum.

Platform

Height AGL

Achievable resolution - GSD (m)

m

Ft

Multispectral

Hyperspectral

Satellite hyperspectral and multispectral imagery

450 -770 km

279-478 mi

1-6

1.25-3.7

Table 9: Achievable resolution values for hyperspectral and multispectral satellite imagery

7.4 Satellite imagery products

7.4.1 Stereo imagery

Imagery from satellite platforms can be formed into stereo models. These are created from pairs of images captured from the same orbit at different times and the associated camera model and geo-referencing information. Using this imagery requires specialist software with advanced image processing and photogrammetric tools.

Stereo imagery finds applications in the creation of DTMs, DSMs, mapping and 3D visualisation.

7.4.2 Orthophotos

Imagery providers frequently offer processed imagery in the form of orthophotos, where the rectification to create the true-to-scale 2D imagery has already been completed.

They can be created in any common imagery format, in any specified coordinate system. The measurement of a distance on the image such as a road width will be replicated in the terrain, making them an indispensable tool for a wide range of applications.7.4

Multispectral and hyperspectral imagery from satellite platforms offer the same set of products as from an aerial platform, (see sections 6.4.2 and 6.4.3) with the obvious advantage of potentially a much larger area of coverage captured in a shorter time span.

As well as the traditional three- and four-band orthorectified products, modern satellite multispectral sensors tend to offer a wider range of spectral bands than their aerial counterparts, which can be automatically classified in the same way using a spectral signature.

It is particularly important to verify the accuracy of the spectral analysis with ground truth observations.

7.4.4 Radar

Radar imaging has the advantage of the ability to operate at night, through thin cloud and thick vegetation layers.

Radar imaging has found applications in digital elevation model (DEM) generation, structural monitoring of dams, buildings and bridges as well as in land management, monitoring subsidence and detecting changes such as deforestation.

Interferometric synthetic aperture radar is a technique that uses pairs of radar images to generate maps of surface deformations. These are generated by the analysis of the differences of the waves returning to the satellite or aircraft-based instrument.