According to the tenth-century Arabic geographer Al-Masudi, travellers through the mountains between Samarkand and China had to pass through a valley where the smoke was so dense that the Sun’s rays could not penetrate. Al-Masudi recorded1 how paid porters would “use sticks to drive the passengers on their journey; for any stoppage or rest would be fatal to the traveller, in consequence of the irritation which the ammoniacal vapours of this valley produce on the brain, and on account of the heat”. His graphic account describes the earliest known industrial source of ammonia emissions, and has fresh significance in light of a study reported on page 99 by Van Damme and colleagues2. The authors have mapped atmospheric ammonia levels with unprecedented precision around the globe, and have quantified emissions from this ancient source for the first time — along with those from a host of previously uncharacterized industrial and agricultural hotspots.
In the valley mentioned by Al-Masudi, locals were exploiting the spontaneous natural combustion of surface coal seams. They used stone huts to collect ammonium chloride and other ammonium salts3,4 carried by the fumes, with the remaining emissions contributing to air pollution. Although this oldest of ammonium industries is no longer in business, Van Damme et al. identify two sites, at Abakan in Russia and Jharia in India (Fig. 1), that are emitting ammonia from burning coal mines. Their demonstration that global satellite observations can now detect such ammonia sources represents a historic moment for science.
Al-Masudi’s example makes it clear why we should care about ammonia. Emitted ammonia reacts rapidly with other air pollutants, and thereby helps to form fine particulate matter that shortens the human lifespan through respiratory and coronary diseases5. Moreover, gaseous ammonia and ammonium compounds formed from it in the atmosphere are deposited into ecosystems, damaging sensitive habitats — especially those naturally adapted to need clean air. Ammonia emissions from agricultural sources also reduce the efficiency with which nitrogen is used though the food-production chain, which has knock-on consequences that increase greenhouse-gas emissions and contribute to water pollution6.
Van Damme et al. carried out a high-spatial-resolution analysis based on nine years of data derived from the Infrared Atmospheric Sounding Interferometer (IASI) — an instrument that takes twice-daily measurements of atmospheric ammonia levels — on the Metop-A meteorological satellite. This allowed the researchers to estimate ammonia emissions from 248 hotspots (defined by the authors as areas with diameters of less than 50 kilometres) and a further 178 regional sources (which have no clearly defined hotspot).
This is not the first report of ammonia distribution mapped from satellite observations. Earlier publications used IASI data7 or measurements from other infrared-observing platforms8,9 to produce global maps and characterize source regions. What sets Van Damme and colleagues’ analysis apart is the comprehensiveness and diversity of quantified ammonia sources. The study shows how satellite technology is coming of age as it starts to fulfil multiple scientific and policy-assessment objectives.
For example, the authors used a method called oversampling to produce a much more precise global map than was previously available. The IASI instrument scans the entire globe daily, and records observations at each observation point at 09:30 and 21:30 hours (local time), but always at slightly different positions. By averaging the nine-year record, the researchers were able to derive fine-scale maps with a resolution of 0.01° × 0.01°, from which they identified hundreds of ammonia sources. They also used a simple ‘inversion’ method to estimate the emissions produced from all of the sources. The authors complemented these approaches by using aerial photographs and other independently obtained data to help them characterize the nature of the hotspots.
Most of the hotspots were found to be associated with intensive livestock farming and industrial activities. Van Damme et al. also discovered a previously unknown natural source — Lake Natron in Tanzania, where drying mudflats are found to release substantial amounts of ammonia to the atmosphere.
One of the authors’ key findings is that many of the industrial sources are missing from ‘bottom-up’ inventories of ammonia emissions (which combine data on the intensity of ammonia-emitting activities with estimated quantities called emission factors). This shows that satellite technology can now be used as an auditing tool for the national reporting of ammonia emissions. Some countries might not report emissions from specific sources because, for example, national regulations do not require polluting activities to be registered. The use of standard emission factors in bottom-up inventories might also lead to errors in national reporting, emphasizing the need for independent verification methods.
As with any emerging technology, some limitations of satellite monitoring remain to be overcome. The largest of these are the requirement for the atmosphere to be cloud-free when measurements are made, and the need for a sizeable temperature difference between land or sea surface and the atmosphere — which limits the zones at which measurements can usefully be made to warm-temperate and tropical climates. There is also huge potential to improve on Van Damme and colleagues’ inversion technique, which assumes that the atmospheric lifetime of ammonia is constant everywhere. This simplification will underestimate ammonia emissions at sources in windy locations, such as on coasts or in mountain areas. Together, these limitations might explain why the authors do not detect high ammonia levels at any seabird colonies, which are known to be substantial ammonia hotspots, especially in sub-polar regions10. Curiously enough, the authors identify several fire-based sources (including the second-largest global ammonia hotspot, found in West Africa), but exclude many of these from their detailed analysis.
Perhaps the most important feature of the new analysis, however, is that it has demonstrated ammonia trends at specific locations. For example, the authors detected changes of 15–20% in ammonia levels at hotspots over the period of the study (see Fig. 4 of the paper2). The achievment of this accuracy for hotspots implies that even better precision could probably be achieved for observations at national and regional scales.
Achieving this capability now is especially timely. Ammonia emissions in many countries are currently increasing (Fig. 2), even in the European Union, which has committed to achieving an overall reduction of 6% by 2020 and 19% by 2030, compared with 2005 levels (see Annex II at go.nature.com/2e2gphe). Combined with atmospheric models (which are necessary for considering the effects of ammonia’s interactions with acidic gases and particulate matter), satellite technology offers a valuable independent tool with which to check whether countries are really achieving their goals.
Nature 564, 49-50 (2018)