Infrared Satellite Detection Limits for Monitoring Atmospheric Ammonia

This study investigates single pixel nadir viewing detection limit of atmospheric ammonia for a range of area flux mapping satellite infrared sensor spectral resolutions (0.05 cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> to 2.0 cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup>) and measurement noise levels. The detection level of ammonia is computed directly from simulated satellite ammonia spectral signatures, which has the advantage of being independent of retrieval methodologies. Information on the frequency of a given detection limit, and the cumulative probability of detection, are provided as a function of instrument spectral resolution and noise. For example, a Cross-track Infrared Sounder-like instrument with a modest spectral resolution of 0.625 cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> and excellent signal-to-noise ratio of ∼1600 would be able to detect ammonia on average ∼70% of the time; these same instrument specifications will have a detection limit of 0.2 ppbv (surface) or 1.6x10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">15</sup> molec cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-2</sup> (total column) that can be achieved at a detection rate of 10% as it requires favourable infrared remote sensing conditions (large thermal contrast). Under more typical atmospheric states a detection limit of 0.5 ppbv (3.5x10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">15</sup> molec cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-2</sup>) is achieved at a 50% detection rate. This detection limit information is valuable for applications that incorporate remote sensing data in conditions when the atmospheric ammonia amounts are below the detection limit of the satellite sensor (e.g. non-growing season in crop fertilizer source regions). As the simulations use real-world atmospheric state observations as inputs the results can be used to provide general guidance on the detection limits of past, current, and potential new environmental flux mapping instruments used for ammonia monitoring covering large geographical regions.