Skip to Content

Data Gap Bridging Study

The SciSAT-1 satellite was launched in August 2003 and is now in its 7th year of operation. The Atmospheric Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS) onboard SciSAT-1 utilises solar occultation to measure temperature and pressure as well as vertical profiles of over thirty different chemical species. Global coverage for each species is obtained approximately over one year and with a vertical resolution of typically 3-4 km. Eventually, the ACE mission will terminate and the Canadian Space Agency (CSA) is looking toward future options to replace ACE with an instrument that uses similar viewing geometry, such as one that will be aboard the Chemical and Aerosol Sounding Satellite (CASS). However, as the CASS launch date has been delayed, it means the mission will not be commissioned in time to overlap with the ACE/SciSAT-1 mission. As a result, a solution is required to bridge the two time series to facilitate a calibration between the two missions. In this project, we have proposed two different scenarios of how the gap should be bridged, namely, using a limb/occultation instrument, whilst using a nadir instrument as a second alternative. In the first method, we use ACE data to bridge a gap between two limb sounding instruments: Odin SMR and Aura MLS. For each instrument, profiles are binned by month, geographically between 30°N - 60°N and 30°S - 60°S, whilst also between 20 and 40 km. Partial column monthly averages are calculated using the binned profiles. To deal with the differences in spatial coverage between instruments, we apply spatial limits to SMR and MLS defined by the maximum and minimum latitudes ACE covers in each latitude band, for a given month. We note that it is not necessary to account for the differences in vertical resolution between instruments as they are of similar magnitudes. We remove biases by examining the linear best fit lines of each time series and applying offsets to SMR and MLS so that they fit relative to the ACE data. The individual time series can then be merged to create an all instrument average. We analyse different bridge lengths utilising the 30°S - 60°S latitude band, by stopping the SMR data in April 2007 and starting MLS at alternate 6 month periods around this date, where the ACE data acts as a bridge after SMR’s termination. In each case, a new all instrument average is created. We find the method works reasonably well, however different bridge lengths can impact the overall final merged data set and is influenced by; agreement between individual monthly values, the total contribution of each individual instrument to the all instrument average, the offset applied, and the first and end points of each time series. The second method uses MOPITT carbon monoxide data to bridge a gap of one year between a stop and start of an ACE carbon monoxide time series. Here, we emphasize the importance of differences in vertical resolutions between occultation and nadir viewing geometry. By using collocation criteria in terms of time and space between ACE and MOPITT, coincident pairs are found. Using the averaging kernels of the MOPITT data, we convolve each coincident ACE profile so that it shares the same vertical resolution as each respective MOPITT profile. Partial column CO time series are created for three latitude bands: 60°N - 90°N, 30°N - 60°N, 30°S - 30°N, and between 100 and 300 hPa. Agreement between the smoothed ACE and MOPITT time series is typically better than 10% in each latitude band. As the ACE orbit is annually repetitive, it means the same locations are sampled at roughly the same times over a given time period. By being able to predict these locations, and if the ACE mission was to temporarily terminate (as we show for a 12 month period), the MOPITT data could bridge the gap using the hypothetical locations and times of the ACE instrument. Limitations are that there is very little vertical overlap between nadir and occultation viewing instruments, hence this method will not work for studies of the stratosphere and above. Moreover, due to ACE’s viewing geometry, few data are found in the tropics, meaning that it would be hard to rationalize a final trend analysis in this region even if a bridge could be applied.