Severe atmospheric pollution in the Middle East is attributable to anthropogenic sources

The study addresses a prevailing assumption that natural desert dust dominates air pollution in the Middle East. The region lies in the global dust belt and experiences frequent dust storms, strong radiative impacts, and dry conditions that maintain elevated mineral dust levels. At the same time, the Middle East has a substantial anthropogenic footprint, emitting more than 15% of global SO2 and about 7.5% of greenhouse gases. Emission inventories are uncertain and observational data sparse, hampering accurate air quality modeling. To clarify the relative roles of natural versus anthropogenic pollution, the authors conducted and analyzed the AQABA ship-borne campaign (summer 2017) around the Arabian Peninsula and used a regional chemistry-transport model to interpret the measurements, focusing on ozone chemistry and particulate matter in the marine boundary layer, distinguishing fine (submicron, PM1/PM2.5) from coarse aerosols. The research tests whether anthropogenic emissions, rather than desert dust, dominate the health-relevant fine PM and significantly contribute to aerosol optical depth and climate forcing in the region.

Prior studies have highlighted natural drivers of Middle Eastern air quality and climate, including dust radiative effects and circulation. The region’s ozone problem has been linked to stratosphere-troposphere exchange (up to ~25% of tropospheric O3 column in summer), long-range transport from Europe, North America and Asia, and strong local photochemistry from NOx and VOC emissions. Emission inventories are known to underestimate regional anthropogenic sources, especially SO2 and NH3 point sources, and to miss contributions from specific activities such as gas flaring. Satellite-based work has exposed missing SO2 sources and large uncertainties in NH3 point sources. Global and regional modeling has struggled with black carbon underestimation. Conventional dust emission parameterizations often employ bimodal size distributions that may misattribute accumulation-mode aerosol to dust, obscuring anthropogenic contributions in observations and retrievals.

Observations: The AQABA (Air Quality and climate in the Arabian Basin) ship campaign (late June–early September 2017) sampled around the Arabian Peninsula, with stops at Jeddah and Kuwait, covering conditions from remote marine air to heavy pollution and dust storms. Instruments characterized aerosol size distributions (5 nm to 32 µm), aerosol composition and trace gases. Fine aerosol composition (PM1) was measured by an Aerosol Mass Spectrometer (non-refractory sulfate, ammonium, nitrate, organics) and an aethalometer (black carbon). Ion chromatography provided dust and sea salt contributions and aided closure with fine and coarse fractions. Aerosol water was estimated with thermodynamic modeling. Modeling: The WRF-Chem model (10 km resolution; domain 6.67°–45.56°N, 22.92°–69.66°E) simulated meteorology-chemistry interactions using RACM for gas-phase chemistry, MADE for aerosols, ISORROPIA-II for thermodynamics, and a VBS scheme for SOA formation/aging. Boundary and initial conditions for key species (O3, SO2, CO; sulfate, organic and black carbon, dust, sea salt) were prescribed from MERRA-2. Methane was set to 1.8 ppmv for OH sink representation. Anthropogenic emissions were prepared with HERMESv3, combining inventories: CAMS-GLOB-SHIPv2.1 (shipping incl. VOC), HTAPv2 (VOC excluding shipping), OMI-HTAPv2 (SO2), EDGARv5 (broad pollutants excluding shipping, VOC and certain industrial SO2), waste burning, GFASv12 (biomass burning), natural alkene emissions from the Red Sea, NH3 sources (satellite-derived constraints), and BC point sources from flaring. Diurnal/weekly cycles of emissions were applied (regional weekend Friday–Saturday). Biogenic emissions were assumed negligible. Model skill was evaluated against observations for trace gases, aerosol size/composition, and meteorology. Dust representation: Using AQABA optical properties and PM observations, the dust emission flux and size distribution were optimized. A unimodal coarse dust emission size distribution was applied, consistent with scaling theory for brittle fragmentation, omitting an accumulation-mode dust source to match observed size/composition partitioning. Coarse dust comparisons accounted for inlet transmission losses (e.g., estimated losses: 8% at 5 µm, 60% at 10 µm, 99% at 15 µm) by weighting the simulated size distribution with a spectral loss function. Sensitivity and inventory adjustments: Analyses indicated large underestimates of NH3 emissions; agricultural NH3 was scaled by ~15× to reproduce observed ammonium background and variability. BC emissions were underestimated; inclusion of non-speciated PM2.5 (PM2.5 − BC − OC) as BC improved background fits. Over the Arabian Gulf, additional missing sources were inferred from backward trajectories pointing to Mesopotamia petrochemical activities; flaring-related BC emissions were scaled by ~30× to reduce biases. SO2 observations showed extreme spikes near shipping lanes and industrial areas; model daily shipping resolution captured gradients but not all spikes. Potential fast SO2 oxidation pathways (e.g., Criegee intermediates, aqueous-phase oxidation) were considered to explain sulfate underestimation in the Arabian Gulf. Health impact assessment used modeled annual PM2.5 and O3 with MR-BRT exposure-response functions and GBD baseline mortality for 2017 to estimate excess mortality with 95% confidence intervals.

- Fine particulate matter (submicron, PM1) is predominantly anthropogenic (>90% by mass), distinct from coarse natural dust; sea salt and dust contribute only minor, infrequent fractions to fine PM. - Composition of fine PM: ammonium sulfate is substantial (median ~61%); organics are prominent in northern areas (median ~40%); black carbon shows sharp spikes near urban/industrial centers. Over the Arabian Gulf, fine aerosol concentrations reach 30–50 µg m−3. - Ozone: Mean summer O3 exceeds WHO seasonal guideline (~30 ppbv) and daily maximum 8-hourly O3 exceeds 50 ppbv across large areas; net O3 production up to ~32 ppbv/day with hotspots over the Arabian Gulf, northern Red Sea, and Gulf of Suez. Chemistry is mostly NOx-limited or transitional, implying NOx reductions are most effective region-wide, with VOC controls beneficial in transition regimes (Mediterranean, northern Red Sea, Gulf of Oman). - Emission inventory gaps: NH3 emissions are severely underestimated (by >15× needed for agricultural sources to match NHx/PM1 ammonium). BC emissions are underestimated by factors of 2–3 regionally and up to 10–100 in the Arabian Gulf; including non-speciated PM2.5 as BC removes much of the bias. Scaling flaring BC by ~30× further reduces Arabian Gulf biases. Extreme SO2 (>100 ppbv) observed near industrial/port regions; sulfate underestimation in the Arabian Gulf suggests missing rapid oxidation pathways and/or primary sulfate. - Dust: A unimodal coarse-mode dust emission better matches observations; fine-mode dust in boundary layer is ~1–3 µg m−3, while coarse dust varies ~10–100+ µg m−3. Accounting for inlet spectral transmission losses enables unbiased comparisons for coarse dust. - Aerosol optical properties: Anthropogenic (accumulation-mode) aerosols explain a median ~53% of total AOD at 500 nm, reaching ~80–95% in polluted regions (e.g., northern Red Sea, Gulf of Suez). AE patterns reflect fine anthropogenic dominance near pollution centers and coarse dust dominance in dusty regions. - Health burden: WHO PM2.5 guidelines (5 µg m−3 annual mean; 15 µg m−3 24-hour mean) are broadly exceeded except over the Arabian Sea. Excess mortality attributable to air pollution in the Middle East is 745 (514–1097) per 100,000 per year; region-wide, PM2.5 accounts for ~89% of the mortality burden over O3. Country-level PM2.5-attributable excess mortality ranges from ~5.9% (Cyprus) to ~15.9% (Kuwait). While ~52% of regional PM2.5 mass is dust (including sparsely populated deserts), the anthropogenic fraction dominates above-guideline exposures in populated areas. O3 unhealthy levels are entirely anthropogenic. - Deposition and ecosystems: Coastal regions (Red Sea, Arabian Gulf, Eastern Mediterranean) experience intense sulfur deposition, with implications for acidification and regional greening initiatives.

The findings overturn the notion that natural dust dominates health-relevant air pollution in the Middle East. Anthropogenic emissions control the fine (accumulation-mode) aerosol population, which dominates particle numbers and contributes over half of column AOD, implying a climatic radiative forcing comparable to that of abundant desert dust. Elevated O3 across the region, largely from NOx-limited or transitional regimes, highlights NOx reductions as a primary mitigation strategy, with VOC controls beneficial in specific subregions. Inventory gaps—particularly in NH3, BC, SO2, and petrochemical/flaring-related emissions in Mesopotamia—lead to model underestimates and suggest substantial unreported or misallocated sources. Over the Arabian Gulf, rapid SO2-to-sulfate conversion and strong RO2/O3 variability indicate intense local petrochemical VOC oxidation. Partitioning of aerosol optical properties shows anthropogenic aerosols exert major climate-relevant direct radiative effects, challenging assumptions that AOD is dust-dominated. The demonstrated unimodal coarse dust emission scheme aligns better with observed size/composition and avoids misattribution of accumulation-mode AOD to dust, with implications for models and satellite retrieval algorithms. Health impacts are severe and comparable to leading risk factors (e.g., high LDL cholesterol, tobacco), underscoring the urgency of targeting anthropogenic sources—industrial, petrochemical, shipping, traffic, and agriculture—to reduce PM2.5 and O3 exposures.

Anthropogenic emissions are intrinsic and prominent in the Middle East, dominating submicron particulate matter and contributing more than half of aerosol optical depth, thus exerting a climate forcing comparable to desert dust. Fine and coarse particles are largely segregated in size and optical properties, with anthropogenic aerosols governing the accumulation mode and dust confined to the coarse mode. Air quality guidelines for PM2.5 and O3 are widely exceeded, and associated excess mortality is high region-wide. Improving emission inventories (especially NH3, BC, SO2, and petrochemical/flaring sources), adopting a dust emission representation consistent with observations (coarse-mode only), and implementing NOx (and regionally VOC) controls are key to mitigating air pollution and its climate and health impacts. Future work should extend observations beyond a single season/year to assess seasonal and interannual variability and refine model skill and emission estimates.

- Observations are from a single season (summer) and year (2017), limiting assessment of seasonal and interannual variability. - Emission inventories show substantial uncertainties: NH3 (agricultural and point sources), BC (especially petrochemical/flaring), SO2 (shipping and industrial spikes), and VOC emissions are likely underestimated or misallocated. - Model underestimation in the Arabian Gulf points to missing sources and/or processes (e.g., rapid SO2 oxidation pathways, primary sulfate). Shipping emissions represented at daily resolution do not capture subhourly spikes. - Assumptions include negligible biogenic emissions and specific treatments (e.g., including non-speciated PM2.5 as BC, scaling flaring emissions), introducing uncertainty in source apportionment. - Dust comparisons require corrections for inlet transmission losses; residual uncertainties in coarse particle sampling and parameterizations remain.