Observing and Monitoring Earth Using Optical Filters

The agricultural revolution happened over 10,000 years ago, and since then humans have continued to explore new ways of managing and improving their local and global environments.

Measurement is the first step that leads to control and eventually to improvement. If you can’t measure something, you can’t understand it. If you can’t understand it, you can’t control it. If you can’t control it, you can’t improve it.

H. James Harrington

In the modern world it is has become progressively more important to be able to control, or at least comprehend not only the environment but also any impact/interaction upon it. This has been a central requirement in areas such as commerce, science, safety and security.

There is also a strong need to understand both natural and man‐made or influenced phenomena such as forest fires, sea‐ice, lightning, atmospheric constituents (clouds and aerosols), crops/surface cover and albedo, weather (storms), forests and deforestation, ocean temperatures/currents, and oil spills.

Being able to better manage or comprehend these events is crucial in guaranteeing that we preserve a healthy planet which can continue to sustain humanity as part of its diverse ecosystems. So, naturally, understanding these phenomena and how they alter over time requires the need to measure them.

Perspective

It can be extremely difficult to objectively observe any system from within that system, so the best approach is to view it from the outside.

Once the capacity to “slip the surly bonds of earth”¹ was achieved in the 1960s, it rapidly became a race for government sponsored space agencies to develop the capability to observe the earth from above. They did this via the deployment of satellites with capabilities for earth observation (EO).

The first space‐based multispectral imager (MSI) was Landsat‐1, launched by NASA in 1972, more recent additions being the ESA’s Sentinel‐2B - “Europe’s eyes on earth” - launched in 2017 and the upcoming GCOM‐C launch from JAXA.

Table 1 shows just some of the different optical EO instruments deployed currently.

Table 1. Examples of EO missions employing optical instruments²

Observation from orbit has presented specific challenges and their associated solutions. These are outlined below:

• Seeing through the atmosphere (clouds/aerosols) or observing only these atmospheric phenomena or constituents can be addressed via wavelength selective imaging.
• Observing small signals in a large background scene may require the use of large, highly uniform collection optics.
• Including as much measurement capability as possible into a small and lightweight package to reduce launch costs can be addressed using compact/multi‐spectral imaging.
• Ascertaining the type of phenomena (‘what’) and location (‘where’) under observation from a distance such as low earth orbit (160‐2000 km above the earth’s surface) may involve using a combination of high spatial (‘where’) and spectral (‘what’) resolution.
• Surviving launch conditions and operating outside of the earth’s protective atmospheric blanket requires the use of robust and reliable optical components.

Multiple designs and formats of optical filter are at the core of each of these solutions, with photonics-based EO instruments acting as “the eyes of the instruments”.

Wavelength Selective Imaging

Whether band‐pass (BPF), edge‐pass or notch designs are employed, optical filters will provide certain wavelength selectivity to the instruments in which they are used. In EO systems specifically, single selective wavelength bands are frequently required to enable observation of the distinctive spectral characteristics related to the phenomena of interest.

Different environmental and atmospheric constituents will reflect, absorb or transmit different wavelength bands, depending on their chemical composition. The table below illustrates the optical band capability of NASA’s Landsat 8, and the way in which these relate to the constituents being measured.

Table 2. Landsat 8 optical bands [Ref NASA]. 

Selecting optical filters with a band‐pass region that corresponds to the wavelength band being examined, users are able to observe only the signal from the constituent or phenomena they are investigating, allowing them “more signal with less background” in their data.

Large, Highly Uniform Collection Optics

Sometimes the constituent of interest has spectral bands which are narrow, in close spectral proximity to ‘background’ bands or the signal is a slight contributor in contrast with the background. In these cases, large, narrow band‐pass filters (NBPFs) may be one means of providing enhanced wavelength selectivity.

For example, mapping or measuring lightning from orbit requires high wavelength selectivity over a sizeable field of view. Lightning can be examined by studying a narrow atomic oxygen triplet line at 777.4 nm.

As it unclear as to when and where a lightning strike may occur, this necessitates a large detector area over which extremely narrow wavelength selectivity is maintained. This would be best served by an exceptionally uniform, narrow optical filter.

Iridian has the proven capability to produce such an NBPF that is centered to within 20 pm of the target wavelength, over an operating clear aperture >125 mm in diameter.³

Figure 1. Spatial variation in CWL across 125 mm diameter demonstrating highly uniform large NBPF.

The advantages of this ‘field of view’ and precision must be balanced with the additional weight and cost implications of producing such large, complex filters and their related optical components.

Compact/Multi‐Spectral Imaging

In order to reduce weight and cost per ‘science line’, various EO imaging systems aim to include as much science into each instrument as possible. They do this by using a single detector which has the capacity to interrogate multiple spectral bands of interest.

Multi‐spectral imaging (MSI) such as this has pushed forward the development of filter arrays where spectral performance varies spatially across the part. This allows the detector’s different pixel bands to be sensitive to different spectral bands, and therefore different phenomena or constituents of interest.

Figure 2. Multi‐spectral array of 10 BPFs (developed under a subcontract from ABB Canada for the Space Technology Development Program of the Canadian Space Agency)

Two methods are commonly used for the manufacture of multi‐zone/multi‐spectral filter (MZF or MSF) arrays: A butcher block construction, whereby different filters are manufactured, singulated, and assembled together into an array, is one such approach. Alternatively, patterning (via masking) different spectral bands on a monolithic substrate may be employed in the arrays’ manufacture.

Butcher block MZF arrays simplify the coating process by avoiding compounding coating run yields. Here, the filter manufacturer will coat a single band on any given substrate. This is the approach used in the majority of standard optical filter manufacturing.

Butcher block arrays are a good option when several (>4 or 5) bands of interest need to be observed, or when individual filter bands are complex, driving increased filter coating thickness and decreased run success yields.

Patterned MZF arrays offer the possibility of creating almost any shape or size of spectral band (a butcher block is limited to stripes or rectangles) with significant improvement (2x or more) in the size transition zone.

Pixels below the transition zone do not function in terms of analysis, so decreasing the loss of good pixels through the use of a patterned array can be beneficial. Also, coating on monolithic substrates avoids the issues often linked with alignment tolerances between differing spectral bands.

High Spatial and Spectral Resolution

Better discrimination of a ‘science line’ can be improved by decreasing the bandwidth of the filter related to the band of interest. Within MSI applications, this increase in spectral resolution comes at the cost of decreased spatial resolution, because this essentially shrinks the signal to noise (less total light) and the aperture size (pixels available) for any band.

On the contrary, where a broad band, panchromatic (PAN) filter spanning the entire visible spectrum is utilized, the increase in total light at these pixels provides enhanced spatial resolution.

Landsat 8’s PAN band has double the spatial resolution (15 m versus 30 m) when compared to other optical bands. This can be seen in Table 2. Using PAN alongside wavelength selective bands and ‘pan sharpening’, an array can be used for the simultaneous mapping of a specific signal’s origin and what the signal represents.

Robust and Reliable Optics

Lastly, there is little point in a system using a high resolution optical filter if this filter is not robust enough for use in orbit.

Through precise control of deposition processes (enhanced evaporating or sputtering) and materials, an experienced manufacturer is able to produce optical filters with good adhesion and density, able to resist standard terrestrial requirements for changes in temperature and humidity without altering or degrading over time.

Reliability specifications usually include:

• No difference in spectral performance or degradation in surface quality after 24 hours of damp heat exposure (95% RH; 49 °C).
• Rub and tape tests for adhesion
• Thermal cycling/shock tests from ‐60 to 70 °C

Once they reach orbit these filters or filter arrays may need to endure extreme temperature ranges (as low as to 70 K), as well as the electron and solar radiation normally filtered out by the earth’s atmosphere.

Filters and filter arrays used in these settings may need further specialized testing to confirm that they will survive in this harsh environment.

Challenges for the Future

Launch costs continue to decline, so the commercialization of space is growing. While NASA, ISRO and ESA continue to plan and launch satellites with EO capabilities, it is anticipated that constellations of EO satellites will be owned and operated by private organizations such as Planet, Satellogic, Urthecast and BlackSky Global, as opposed to governmental space agencies.

In the “2016 Nano/Microsatellite Market Forecast”, SpaceWorks estimated that from 2018‐2020 there will be ~1,000 nano/microsatellite (1‐50 kg) launches with 70% of these being commercial rather than government projects.

Over 70% of these projects are anticipated to be for EO applications (up from <40% over the past 5 years).

The capitalism of space will see more information available than ever before, with this data being used for more than just defense and intelligence purposes. Once extra‐terrestrially gathered data becomes more commonplace, new applications will likely emerge which leverage this information to better manage and control commerce and other terrestrial activities.

Optical filters and photonics are set poised to play a key role many of these new developments.

References and Further Reading

1. [Ref. John Gillespie Magee, Jr.]
2. [Ref NASA,ESA,ISRO,JAXA]
3. Large format BPF – a uniformity challenge (Iridian)

This information has been sourced, reviewed and adapted from materials provided by Iridian Spectral Technologies.

For more information on this source, please visit Iridian Spectral Technologies.