NCAR Annual Report > RAL Annual Report Contents > Discoveries > 2. Looking at the Earth from Space

A New Way of Looking at the Earth from Space

By David B. Johnson, RAL


Figure 1.  An example of the possible transformation of a geostationary satellite image of the earth using a new kind of optical adapter.  The optical device, a hardware component that would have to be incorporated into the system design before launch, progressively stretches the edges of the earth image prior to data collection to offset the normal loss of image resolution due to earth curvature effects.  The transformed image will provide essentially uniform resolution imagery over almost the whole earth disk as viewed from space, enhancing the quality of spaced-based observations for use in numerical models or for applications involving integration into operational decision support systems.

The international system of geostationary meteorological satellites, augmented by a growing constellation of operational and research satellites in low earth orbit (LEO), are essential tools in efforts to understand the atmosphere and earth system. Figure 1 (left-most earth image) shows the traditional view of the earth from geostationary orbit, in this case from the GOES-EAST operational satellite position, while figure 2 shows the earth-satellite distances and viewing geometry of geostationary and LEO satellites, drawn in true relative scale. 

Views from space, particularly from geostationary orbit, however, are inherently limited by an increasing loss of image resolution as one approaches the edge of the visible earth disk. This foreshortening of earth features is due to the curvature of the earth and the increasingly oblique viewing angle as seen from the satellite.  The red circle drawn on the geostationary earth view in Figure 1, for example, highlights the position at which an imaging system’s sensors will have their resolution reduced by a factor of three from their resolution at nadir (in this case meaning the center of the full disk earth image).  Observations outside of the red line are generally not useful for quantitative applications, and the gradual loss of resolution as one approaches the red line significantly reduces the utility of the satellite observations over areas far from nadir, such as the bulk of the continental United States (CONUS) and Europe, areas that are often of particular interest. 

As an unexpected result of a NASA-funded project to enhance the use of advanced satellite products in support of aviation safety and efficiency, RAL scientists realized that new technology imaging systems based on two-dimensional charge coupled devices (CCDs, such as those used in digital cameras) and focal plane arrays may be able to be “corrected” to offset the loss of resolution due to earth curvature.  The correction would be based on an optical adapter, either a lens or mirror depending on the design of the imaging instrument, which would stretch the imagery as one moves away from nadir just enough to offset the normal loss in resolution as one moves towards the edge of the earth disk.  Figure 1 shows a detailed example of such a correction, starting with a normal view of the earth from geostationary orbit, including full parallax and angle of view considerations, and then applying a two-dimensional numerical lens model to produce the transformed image on the right.  The transformation is limited to the area within the red circle and necessarily produces a larger image.  The features in the center of the transformed image (e.g. the countries in the northwest corner of South America), however, are still the same size as they were in the original earth view.  The image stretching is seen in the enlarged relative depiction of North America and Patagonia, now shown at the same scale and relative proportions as Colombia and Equador.   


Figure 2.  Scale drawing of the Earth-satellite geometry for geostationary and low earth orbit (LEO) satellites. 

The transformed earth image, with its improved resolution over areas away from nadir, would be well-suited for earth systems science and for initializing numerical models.  The transformation is particularly valuable since it is applied before the space-based observations are made and directly enhances the quality of the observations rather than remapping the imagery after the data is collected.  The transformation allows users to have the same quality of data over northern Canada and Labrador as one would expect over Panama, and to view virtually all of CONUS with uniform resolution and uniformly spaced observations from a single satellite. 

This approach is best suited for use with the emerging new generation of two-dimensional imaging systems, rather than the normal scanning radiometers that up to now have been the mainstay of satellite-based observing systems.  The imaging capabilities of the Hubble Space Telescope, however, have long made use of exactly this sort of two-dimensional imaging system and the hyperspectral imaging and sounding systems currently being proposed for the next generation of meteorological satellites also employ 2-D imaging devices.  This approach also permits faster data collection and longer dwell times for improved radiometric accuracy. While this concept is particularly well-matched for observations from geostationary satellites, it can also be adapted to applications for satellite sensing from low earth orbit. 

One of the unique aspects of this discovery is that it was based on consideration of how to apply satellite observations, which then led to an idea for improving how the observations are made.  In essence, the approach envisioned is to consider the observing system (e.g. the satellite imaging device) and the earth as a single system, and then customize the space-based observation for optimal utility for earth-based applications. 

This development directly supports the NCAR Strategic Plan’s emphasis on integration and innovation and offers a promise of improved earth system modeling through advances in observing technologies.  The discovery is the direct result of RAL’s efforts in support of the NASA-funded Advanced Satellite Aviation-weather Products; additional support was provided by the National Science Foundation.