Constituent Retrieval - MANTRA Case Study

The MANTRA stratospheric balloon missions were launched in August 1998 and August 2000. Each flight included a variety of instruments to examine the stratosphere. I have been working with the AES nitric acid instruments that are used to measure nitric acid concentration profiles between about 5 to 35 km in the atmosphere. The AES devices have an extensive flight history and have been flown on many missions over the last three decades. This latest instrument flight has been proposed for two reasons:

The extensive flight history of the device means that data collected during the MANTRA mission can be compared to previous data made by very similar devices. With some reprocessing of old flight data, using modern profile retrieval techniques and greatly increase processing power, it should be possible to estimate nitric acid trends over the last thirty years will some reliability.

The MANTRA mission also includes a number of more modern instruments using other sensing techniques such as FTIR. One of the primary objectives of the mission will be to conduct a cross-comparison of the data produced by the different measurement systems. Making simultaneous measurements in this way should allow a better interpretation between different instrument data sets. It should also lead to an increased understanding of the performance and uncertainties associated with each type of device.

Additionally, the spectroscopic range of the nitric acid instrument makes it suitable for the investigation of CFC-11 and CFC-12 absorption bands. These two bands occur on either side of the nitric acid absorption band, and although they are much weaker than the nitric acid response, it is hoped that it may be able to characterize their historical trends over the same period.

Results

The figure below shows the radiance data from one instrument flown in 1998. The data shown is in the nitric acid spectral region (around 11.6 microns), corrected using in-flight blackbody calibrations.

The radiance measurements were used to recursively estimate a mixing-ratio profile, minimising the error between theoretically calculated radiance and measured radiance. The results are shown below for the same instrument. The blue line shows low gain instrument data and the red line high gain instrument data (saturating towards the beginning of the flight). The green line shows the theoretical radiance fit.

The corresponding computed mixing ratio profile is shown below. This figure shows the results for both instruments. The calculation was performed on a 2 km vertical grid. Individual theoretical radiance measurements where interpolated to sub-2 km resolution. The level-by-level radiance calculations where also interpolated to a 1 km resolution grid.

Summary and Conclusions

The preliminary results from the two instruments show excellent agreement in the 10-35 km altitude range. The mixing ratio shows clear structure that is unlikely to be attributable to instrument effects as each instrument made measurements independently and the measurement sets are not synchronised in time. The large projected mixing ratio for the 35 km layer is a calculation artefact: the large value represents an equivalent nitric acid concentration for the total gas column above 35 km if all the observed radiance was attributable to gas in only the 35 km layer. Below 10 km atmospheric opacity is likely to have a significant effect on the nitric acid estimate and results are not reliable.

The mixing ratio results are broadly in line with previous calculations for the region and season. There may be a small revision of the estimated profile in future analysis when a more accurate calibration of the instrument field-of-view has been conducted. Future work will attempt to use the same recovery algorithm to estimate ozone-mixing ratios. Ozone concentrations are relatively well determined and were also measured directly by in-situ sonde sampling during the campaign. It is therefore hope to validate the recovery algorithms and nitric-acid mixing ratio results by validating ozone-profile recoveries compared with existing data.