Multi-satellite sensor study on precipitation-induced emission pulses of NO<sub><i>x</i></sub> from soils in semi-arid ecosystems
Abstract. We present a top-down approach to infer and quantify rain-induced emission pulses of NOx (≡ NO + NO2), stemming from biotic emissions of NO from soils, globally with a spatial resolution of 0.25° from satellite-borne measurements of NO2. This is achieved by synchronizing time series at single grid pixels according to the first day of rain after a dry spell of prescribed duration. The full track of the temporal evolution several weeks before and after a rain pulse is retained with daily resolution. These are needed for a sophisticated background correction, which accounts for seasonal variations in the time series and allows for improved quantification of rain-induced soil emissions. We find strong peaks of enhanced NO2 Vertical Column Densities (VCDs) on the first day of rainfall after prolonged droughts in many semi-arid regions of the world, in particular in the Sahel. Detailed investigations show that the rain-induced NO2 pulse detected by the OMI, GOME-2 and SCIAMACHY satellite instruments could not be explained by other sources, such as biomass burning or lightning, or by retrieval artefacts (e.g. due to clouds). For the Sahel region, absolute enhancements of the NO2 VCDs on the first day of rain based on OMI measurements 2007–2010 are on average 4 × 1014 molec cm−2 and exceed 1 × 1015 molec cm−2 for individual grid cells. Assuming a NOx lifetime of 4 h, this corresponds to soil NOx emissions in the range of 6 ng N m−2 s−1 up to 65 ng N m−2 s−1, in good agreement with literature values. Apart from the clear first-day peak, NO2 VCDs show moderately enhanced NO2 VCDs of 2 × 1014 molec cm−2 compared to background over the following two weeks suggesting potential further emissions during that period of about 3.3 ng N m−2 s−1.