Characterization and simulation of a ground-based millimeter wave observation system for Arctic atmospheric research

Abstract A preparatory performance and error characterization was carried out for a ground-based millimeter wave instrument designed for high Arctic atmospheric research. The instrument is a radiometer to measure rotational emission spectra of O 3 , ClO, HNO 3 , and N 2 O, between 265 and 280 GHz, using a sensitive superconductor–insulator–superconductor detector. Forward and inverse modeling tests have been performed to assess the instrument/inversion system and to determine the sources of the most significant errors in the retrieval of each trace gas. The altitude ranges over which retrievals of concentrations can be made were found to be ~13–62 km for O 3 , ~12.5–39 km for N 2 O, ~12–36 km for HNO 3 , and ~18–46 km for ClO. For each target species the measurement and smoothing errors calculated with an optimal estimation method (OEM) were compared to the errors calculated from inversions of 500 simulated spectra. The absolute error from these inversions agreed well the OEM results, but there were systematic differences that are attributed to nonlinearities in the forward model. The results of these nonlinearities can cause biases of the order of 5–10% of the a priori profile if they are not accounted for when averaging concentration profiles or when analyzing trends in concentration. The techniques used here can be applied to any ground-based remote sounder.