Environment Canada. Environnement Canada. The University of Toronto’s Balloon-Borne Fourier Transform Spectrometer Debra Wunch, Jeffrey R. Taylor, James R. Drummond Clive Midwinter, Kimberly Strong Department of Physics, University of Toronto Hans Fast Environment Canada.
The University of Toronto’s Balloon-Borne
Fourier Transform Spectrometer
Debra Wunch, Jeffrey R. Taylor, James R. Drummond
Clive Midwinter, Kimberly Strong
Department of Physics, University of Toronto
The University of Toronto’s Fourier transform spectrometer (U of T FTS), derived from an ABB-Analytical DA5 Michelson-type interferometer, was rebuilt and flown on the Middle Atmosphere Nitrogen TRend Assessment (MANTRA) high-altitude balloon platform in September, 2004. The U of T FTS has a resolution of 0.02 cm-1, a spectral range covering 1200-5000 cm-1, and InSb and MCT detectors that measure simultaneously.
The spectrometer was originally built in the 1980s and purchased by the Meteorological Service of Canada. To prepare the instrument for flight, new software was necessary and was replaced by LabVIEW control software, creating a robust and easily-controlled instrument, adaptable to either remote control or lab-based work. The software contained accessible housekeeping information, downlink capability and an embedded scheduler. In addition to replacing the software, nearly all the original electronics were replaced using off-the-shelf components. The dynamic alignment system is the only original electronic system remaining. A small suntracker was incorporated which facilitated alignment of the spectrometer with the balloon payload.
Despite telemetry and gondola pointing system failures during the MANTRA 2004 campaign, two spectra were recorded on each detector during sunset from a float height of 35 km. The data indicate that the instrument performed well throughout the flight, and, had the payload pointing been under control, would have retrieved a full set of occultation data. The data that were acquired will be shown. The spectrometer has since participated in a three-month-long ground-based FTS inter-comparison campaign. The results from the intercomparison show that the U of T FTS retrieves, on average, ozone columns that are within 3% of a co-located Brewer Spectrophotometer.
The Instrument and its Scientific Roles
The University of Toronto’s balloon-borne Fourier Transform Spectrometer (U of T FTS) is a Michelson-type interferometer with a 0.02 cm-1 resolution and a spectral range covering 1200-5000 cm-1 across two detectors. The two detectors, an InSb and an MCT detector, measure simultaneously. The instrument is configured to measure atmospheric trace-gas species in occultation mode: that is, measuring solar absorption through sunrise and sunset. It is designed primarily for high-altitude balloon-borne measurements.
The U of T FTS has participated in two Middle Atmosphere Nitrogen TRend Assessment (MANTRA) high-altitude balloon campaigns: in 2002 and 2004. The role of the FTS aboard MANTRA is to advance MANTRA’s science goal of understanding mid-latitude ozone depletion, and, specifically, to determine the role of nitrogen compounds in the depletion of ozone. The U of T FTS contributes to this project because it can simultaneously measure many trace gas species related to nitrogen and ozone, with high vertical and spectral resolution.
The FTS Refurbishment
The U of T FTS was initially purchased by the Meteorological Service of Canada in the early 1980s from what is now known as ABB-Analytical, a company that produced a research-grade FTS with an excellent optical design. The main asset of this design was the dynamic alignment system that used the stationary mirror to compensate for the inevitable angular changes of the scanning mirror (see Figure 1).
The instrument control software provided by ABB-Analytical allows the user to manually control the main functions of the FTS in a simple and repeatable way. This, however, was not appropriate for our balloon application, which requires that interferograms be recorded during solar occultation (sunrise and/or sunset) with limited telemetry uplink and downlink. For this application, it is necessary to control the instrument remotely and have a robust automatic mode. The original software is embedded in the original control computer for the instrument, so, in order to update the software, the control computer and nearly all of the original electronics had to be replaced, while ensuring that the dynamic alignment system was kept intact. This was done using off-the-shelf components, which resulted in a significantly smaller instrument that was reduced in mass from around 90kg to 55kg, with reduced power consumption from around 140W to 65W.
The original software was replaced by LabVIEW control software, creating a robust and easily-controlled instrument, adaptable to either remote control or lab-based work, containing accessible housekeeping information (voltages, temperatures and instrument status information), downlink capability and an embedded scheduler.
While gondola pointing systems can be good to a few degrees in zenith and azimuth, FTS instruments require more precise solar pointing abilities. A small, light-weight, 2-axis suntracker with ±10° zenith and azimuth field of view was used to partly decouple the FTS from the main gondola pointing system. This allowed for easy instrument integration on the payload and a reasonably simple transition to ground-based measurements (two extra mirrors were required to direct the sun into the tracker on the ground). Figure 1 shows a schematic of the instrument and delta-tracker.
Results from the MANTRA 2004 Flight
Due to a telemetry and gondola pointing system failure during the MANTRA 2004 flight, the payload rotated during sunset. Since the delta-tracker has only a ±10° zenith and azimuth field of view, only two spectra were recorded on each detector. There is every indication that, had the payload been properly oriented, the instrument would have recorded a full occultation of spectra.
The MCT spectrum, seen to the right in Figure 2, was recorded while looking through the atmosphere from a float altitude of approximately 35 km at a 89° solar zenith angle. Signatures of O3, N2O, CH4 and CO2 are clearly visible in the spectra, and a selection of microwindows for these gases are displayed in the lower panels.
Long-Term FTS Intercomparison Campaign
The U of T FTS participated in a long-term ground-based intercomparison campaign with the Toronto Atmospheric Observatory FTS (TAO) that lasted from May 2005 through early September 2005. The TAO FTS, is a linear, ABB-Analytical DA8 FTS, and is housed at a complementary NDSC site in a rooftop laboratory at the University of Toronto It is solely a ground-based FTS, and records solar absorption measurements during every clear-sky opportunity throughout the year. TAO has been in commission since 2001. Also on-site during the campaign was a Brewer Spectrophotometer, for ozone total column comparisons.
The main scientific goal of the intercomparison is to answer the question: What is the effect of instrument resolution on the retrieved column amounts of trace gas species? To be able to answer this question, we eliminated possible differences due to retrieval methods and a priori trace gas profiles, by using the retrieval algorithm SFIT2 and the same a priori information. We eliminated possible differences in the atmospheric conditions themselves by measuring simultaneously with both instruments, and using a pick-off mirror to supply the U of T FTS with sunlight from the TAO sun tracker. This isolates the problem to one of instrumentation.
Results from the Long-Term Intercomparison Campaign
Results for ozone, N2O, CH4, HCl and CO have been retrieved for this campaign. Only the ozone results are shown here. Figures 3a, b, and c below show the agreement between the U of T FTS, the TAO FTS and the Brewer Spectrophotometer. Figure 3a shows the total columns and Figure 3b and c show the percent difference between the TAO FTS and the Brewer and the U of T FTS and the Brewer, respectively. The U of T FTS is, on average, 3% lower than the Brewer, with no discernable structure in the residuals of Figure 3c. The TAO FTS reads lower than the Brewer by 3.5% on average.
Figure 1. A schematic of the University of Toronto’s Balloon-Borne Fourier Transform Spectrometer.
Figure 2. Top panel: an MCT spectrum recorded from the MANTRA 2004 balloon during sunset; bottom panels: a closer look at some molecular signatures.
Figure 3. Results from the Long-Term Intercomparison. In a, the total columns; in b, the percent difference between the TAO FTS and the Brewer; in c, the percent difference between the U of T FTS and the Brewer.
Summary and Future Work
The University of Toronto’s Fourier transform spectrometer was successfully modernized and recorded good data under difficult conditions. New software and a “delta-tracker” were key components in enabling its success on board the MANTRA 2004 payload.
The ground-based long-term intercomparison with the TAO FTS has revealed some interesting results. Future work will concern investigating the reasons for the apparent bias between both FTS instruments and the Brewer.
The authors wish to thank C. Tom McElroy, Pierre F. Fogal, John Olson and the MANTRA Science team for their invaluable advice and assistance.
MANTRA is supported by the Canadian Space Agency (CSA), Environment Canada, CFCAS and NSERC.
The figure in the background is from an ozonesonde launch during the MANTRA 2004 campaign.
To contact Debra Wunch, email firstname.lastname@example.org