A 4 U laser heterodyne radiometer for methane (CH₄) and carbon dioxide (CO₂) measurements from an occultation-viewing CubeSat

CubeSat instruments provide a unique and often underrated opportunity to make targeted atmospheric measurements at a significantly lower cost and faster turn-around than traditional flight instruments. Though the size limits imposed on a CubeSat payload (typically under 4 U⁷) were originally thought to severely limit the potential for a useful data product, the miniaturization and commercial availability of many laser components that have been brought on by growth of the telecommunications industry allows for construction of small and effective instrumentation.

Here we present the design of a passive, occultation-viewing, 4 U CubeSat laser heterodyne radiometer (LHR) instrument that measures CH₄, CO₂, and H₂O in the limb that has been developed and built at NASA Goddard Space Flight Center and will be housed on a 6 U CubeSat bus developed and built at Lawrence Livermore National Laboratory (LLNL). Deployed in a low earth orbit (LEO), this instrument will be capable of observing the region between the upper troposphere to the lower stratosphere. Compared to current greenhouse gas observation satellites such as the greenhouse gas observing satellite (GOSAT) [1] and the orbiting carbon observatory-2 [2, 3] that measure the total column absorption of greenhouse gases in sunlight reflected from the Earth's surface, this CubeSat has much smaller payload, much lower cost and much higher spectral resolution which is appropriate to measure the narrower vibration-rotational lines in the upper air with lower atmospheric pressure. In addition, the LHR sun-pointing limb measurements in the upper air will have much less atmospheric scattering effect which has known to be the largest error in those satellite observations using surface reflected sunlight [4–6].

Measurements of CH₄, CO₂, and H₂O from the upper troposphere to the middle stratosphere provide critical information on stratospheric circulation and how it responds to increasing greenhouse gas (GHG) concentrations. The circulation affects stratospheric temperature by controlling the distribution of radiatively important gases such as CH₄, CO₂, O₃, and H₂O, but also affects photochemistry through its impact on the distribution of gases that control O₃ loss processes. Measuring stratospheric circulation and its variability is essential for projecting how climate change will affect stratospheric ozone.

The projected CH₄ and CO₂ increases in the 21st century will impact photochemical, dynamical, and radiative processes in the stratosphere. Increased CH₄ results in increased stratospheric water vapor and OH production, directly impacting O₃ loss by OH. Changes in stratospheric ozone lead to local temperature changes and produce a dynamical response. Methane is long-lived below the upper stratosphere (~5 hPa), and thus its distribution is controlled by seasonal and interannual variations in the strength of stratospheric circulation. Most chemistry climate models predict a speed-up of the stratospheric circulation in this century due to increasing GHGs [7]. Because of the methane's role in the changing chemistry of the ozone layer and because of its long lifetime, methane measurements are especially valuable to measuring composition change and for identifying and quantifying GHG-driven changes in stratospheric circulation.

The 4 U LHR presented here is based on a similar ground-based instrument that was developed at NASA Goddard Space Flight Center for measurements of CO₂, CH₄ and H₂O in the atmospheric column [8–11]. Some changes in components were required to reduce the mass, volume and power of the LHR as well as to remove the use of problematic moving parts such as the optical chopper (replaced with an optical switch). What the 4 U instrument shares with the ground instrument is the use of low-cost, commercial, off-the-shelf telecommunications components. A challenge of using these components is that they are intended for use at the telecom wavelengths of ~1550 nm and not at 1640 nm where methane absorption features have strong enough line strengths to be measured in our atmosphere (CO₂ and H₂O absorption features in contrast also have good line strengths at lower wavelengths such as the 1570 nm region). For this reason, these components are often not operating at peak efficiency and it is likely that as the technology progresses, the performance of this design will also improve.

In the 4 U design shown in figure 1(a), collimation optics with a 0.2° field of view (Thorlabs part #F810APC-1550) point at the sun and collect light that has undergone absorption by methane in the limb. Collected sunlight is modulated with a fibre optic switch (CrystaLatch^™, Agiltron part #CLSW-115231323) at 500 Hz and then launched into a 50/50 single-mode fibre coupler (Thorlabs part #10202A-50-APC) where it is superimposed with IR laser light from a distributive feedback (DFB) laser (Eblana photonics part #EP1653-7-DM-B01-FA). Laser light and sunlight are mixed in an InGaAs fast (>10 GHz) photoreceiver (Electro-Optical Technology model #ET-3500FCEXT/APC) to produce a beat signal in the radio frequency (RF) region. The RF receiver (custom) amplifies and detects the beat signal [8]. The resulting signal is measured with an analog lock-in amplifier (custom PCB) that is referenced to the fibre switch frequency. The analog output from the lock-in amplifier is scaled and converted to a digital signal through the microcontroller's (Freescale, model #MK20DX256) integrated 16 bit A/D converter. To scan an absorption feature, the microcontroller's external 16 bit D/A (analog devices AD5686) provides the laser controller (wavelength electronics part #LDTC2/2E-00000B) with a constant temperature set voltage and a sawtooth voltage to produce a changing laser current to repetitively scan the laser across the wavelength region of interest. The sawtooth voltage input to the laser, shown in more detail in figure 1(b), is a series of voltage steps that are taller at the start (α) and end (γ) of each line, and shorter at the center (β) of each line. This results in a higher scan resolution over the main body of the absorption feature where the curvature of the line is largest and a lower resolution over the wings as shown in figure 1(c). Each voltage step (corresponding to a wavelength of the laser) for example, could persist for 500 ms with the lock-in amplifier integrating for 250 ms. While the processor samples the signal from the lock-in continuously, only integrated points that fall on each wavelength step are ultimately averaged; meaning that if a data point has integrated a portion of the previous wavelength into the signal, this data point is not included in the averaged data for that wavelength. Figure 2 shows modeled spectroscopy of the 1640 nm region with simulated line transmittances at limb tangent heights of 6 km through 30 km [12].

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

A 6 U CubeSat bus has been developed to support the 4 U payload that is based on the CubeSat next generation bus (CNGB) 3 U architecture and design. CNGB is being developed for the National Reconaissance Office (NRO) by a team led by LLNL, with significant contributions from the Naval Postgraduate School and the Space Dynamics Laboratory (SDL) at Utah State [13]. The bus is designed to be compatible with the Nanosatellite Launch Adapter System (NLAS) developed by NASA Ames Research Center and builds upon the CNGB mechanical, electrical and software standard. The 6 U bus with the 4 U LHR payload is shown in figure 3 and the spacecraft block diagram is shown in figure 4.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

3.1. Command and data handling (C&DH)

3.2. Electric power system (EPS)

3.3. Attitude determination and control system (ADCS)

3.4. Global positioning system (GPS)

3.5. Radio communication system

3.6. Environmental

A one-year observation period following a high inclination orbit is projected to be utilized in this mission. This orbit will produce the longest data collection window for the occultation instrument as the observation path crosses the interface between night and day (terminator). The right side of figure 5 shows a graphical representation of the data that could be collected with an orbit similar to the ICESat-2 satellite [16]. This orbital path would produce continuous measurements for up to 530 s with an average duration of 120 s. During a 30 d period the instrument will capture a cumulative total 17.6 h of measurement data.

Zoom In Zoom Out Reset image size

Orbits with high inclination typically allow for uninterrupted communication with the Iridium constellation with a downlink of 100–500 kB d⁻¹. This cost effective approach is pay-per-use (data is exchanged over the internet) and removes the need for custom receiving infrastructure on the ground. The transceiver also allows for short burst transmissions of a few hundred bytes which are more likely to be transmitted in the case of orbital difficulties.

Mole fractions are determined by fitting simulated limb measurements of a high spectral resolution radiative transfer model to LHR's absorption feature scans. Simulated transmittance, τ_ν, at a wavenumber ν, is calculated with the Beer–Lambert law:

$$\begin{eqnarray}&&{{\tau}_{\nu}}=\exp \left(-\upsilon \ast S\ast {{g}_{\nu}}\right)\end{eqnarray}$$

where υ is the mass path describing the number of molecules per cross-sectional area in the path (molecules cm⁻²), Ѕ is the HITRAN line intensity [17], and g_ν is the simulated line shape, broadened by Doppler and pressure effects [18]. Simulated transmittance is then scaled to account for altitude and position at the time of measurement and corrected for instrumental broadening. Furthermore, scene detection is used to exclude data taken through clouds. Other external factors common to all passive remote sensing instruments, such as solar spectrum variability, are excluded in analyzing the expected performance of the 4 U LHR. Mole fractions are extracted by fitting simulated transmittance spectra to LHR's absorption feature scans. This fit is achieved through matching critical features of the simulated and measured transmittance spectra and adjusting expected mole fraction parameters. To achieve a high signal-to-noise ratio, multiple scans, typically 20, are individually analyzed with the algorithm and then averaged.

The retrieval algorithm described is based on similar software that is currently used to analyze ground-based data and will provide a data-fitting similar to that shown in figure 6. The main difference is that transmission data will be taken from a horizontal atmospheric slice consisting primarily of the stratosphere, as shown in figure 5; whereas ground-based measurements are the result of light absorption occurring vertically through all atmospheric layers. Furthermore, the high-altitude, low pressure setting of the 4 U will result in less line broadening and narrower absorption lines. It should be noted that to optimize the dynamic range of our observations in these different portions of the atmosphere, we have selected slightly different wavelength regions for each of these instruments; 1639.4 nm–1639.9 nm for the CubeSat and 1640.3 nm–1640.45 nm for the ground instrument. While it would be convenient to measure the exact absorption features in both instruments, CH₄ and CO₂ lines measured by the ground instrument start to saturate at limb tangent heights below 20 km, and CH₄, CO₂ and H₂O lines that will be observed by the CubeSat have a weak transmittance when viewed through the column by the ground instrument. Space based measurements of CH₄ and CO₂ require precisions of ~10 ppb (~0.5%) and ~1 ppm (~0.25%) on regional scales, respectively, to produce data products that resolve spatial gradients and reduce uncertainties in CH₄ and CO₂ budget estimates [19–25]. For climate studies, water vapor in the upper troposphere and lower stratosphere is of equal importance to its role in the lower atmosphere [26]. With this LHR design, limb water vapor measurements could achieve a better than 1% precision, and their addition would improve the retrieval of CO₂ and CH₄ concentrations with regard to dry air in the same pathlength. Both high spectral resolution and high signal-to-noise are essential to these high precision retrievals [4, 27]. The wavelength precision of the LHR measurements are expected to be on the order of ~3  ×  10⁻⁶ nm (~10⁻⁵ cm⁻¹) based on the wavelength characterization of a DFB laser, which is well established [28]. Compared to traditional grating spectrometers and Fourier-transform spectrometers, the LHR approach is intrinsically higher in spectral resolution since it is based on interference of sunlight with a DFB laser, which has a linewidth of less than 10⁻⁴ cm⁻¹. Such high measurement sensitivity is critical for high-altitude calculations where less line broadening occurs due to low pressures. Traditional spectrometers are limited in spectral resolution by the length of the interferometer arms, whereas the resolution of heterodyne measurements is set by the integration bandwidth. The present spectral resolution of the Mini-LHR ground instrument is the FWHM of RF band-pass filter which is about 400 MHz or ~0.013 cm⁻¹ (~3.6 pm at 1640 nm). The higher spectral resolution of the LHR makes it more sensitive to changes in the absorption, and one sees less of the broadening and line blurring that affects traditional spectrometers.

Zoom In Zoom Out Reset image size

We have presented a compact (6 U) CubeSat design that implements a passive, sun-pointing LHR to observe CH₄, CO₂ and H₂O in the limb at 1640 nm (6098 cm⁻¹). The CubeSat design is based on a ground LHR instrument that measures CH₄ and CO₂ in the atmospheric column that has a current spectral resolution of ~3.6 pm (~0.013 cm⁻¹) and a precision of 10 ppb for CH₄ and 1 ppm for CO₂ for 20 averaged scans. Space observations require precisions of ~0.5% for CH₄ and 0.25% for CO₂ and are achievable with this design based on the expected wavelength precision of ~3  ×  10⁻⁶ nm (~10⁻⁵ cm⁻¹) and better than 500:1 SNR. Water vapor precisions are expected to be ~1% and will contribute to CO₂ and CH₄ retrievals with respect to dry air in the same pathlengh. LHR measurements will fully resolve a single line of each greenhouse gas, including line shape and line width, for low-bias and high precision retrievals. This CubeSat will ideally be deployed into a LEO, high-inclination orbit where the ascending node is oriented 90° from the sun to produce the longest observations. Measurements of CH₄, CO₂, and H₂O from the upper troposphere to the middle stratosphere will provide critical information on stratospheric circulation and how it responds to increasing greenhouse gas (GHG) concentrations.

Mini-LHR development was supported by the NASA Goddard Space Flight Center Internal Research and Development program and Science Innovation fund. This work was performed in part under the auspices of the US Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.