Extreme air quality episodes represent a major threat to human health worldwide but are highly dynamic and exceedingly challenging to monitor. The 2018 Kīlauea Lower East Rift Zone eruption (May to August 2018) blanketed much of Hawai‘i Island in “vog” (volcanic smog), a mixture of primary volcanic sulfur dioxide (SO₂) gas and secondary particulate matter (PM). This episode was captured by several monitoring platforms, including a low-cost sensor (LCS) network consisting of 30 nodes designed and deployed specifically to monitor PM and SO₂ during the event. Downwind of the eruption, network stations measured peak hourly PM_2.5 and SO₂ concentrations that exceeded 75 μg m⁻³ and 1,200 parts per billion (ppb), respectively. The LCS network’s high spatial density enabled highly granular estimates of human exposure to both pollutants during the eruption, which was not possible using preexisting air quality measurements. Because of overlaps in population distribution and plume dynamics, a much larger proportion of the island’s population was exposed to elevated levels of fine PM than to SO₂. Additionally, the spatially distributed network was able to resolve the volcanic plume’s chemical evolution downwind of the eruption. Measurements find a mean SO₂ conversion time of ∼36 h, demonstrating the ability of distributed LCS networks to observe reaction kinetics and quantify chemical transformations of air pollutants in a real-world setting. This work also highlights the utility of LCS networks for emergency response during extreme episodes to complement existing air quality monitoring approaches.

极端的空气质量事件对全世界人类健康构成了重大威胁，但具有高度动态性，监测起来极具挑战性。2018 年基拉韦厄下东裂谷带喷发（2018 年 5 月至 8 月）将夏威夷岛的大部分地区覆盖在“vog”（火山烟雾）中，这种烟雾是原生火山二氧化硫 (SO ₂ ) 气体和二次颗粒物 (PM)的混合物。这一事件被多个监控平台捕获，包括一个低成本传感器 (LCS) 网络，该网络由 30 个节点组成，专门设计和部署用于在事件期间监控 PM 和 SO ₂。在火山喷发的下风向，网络站测量到峰值每小时 PM _2.5和 SO ₂浓度超过 75 μg m ^-3和 1,200 十亿分之 (ppb)，分别。LCS 网络的高空间密度能够对火山喷发期间人类暴露于两种污染物的情况进行高度精细的估计，这是使用预先存在的空气质量测量结果无法实现的。由于人口分布和羽流动力学的重叠，与 SO ₂相比，该岛人口中暴露于高水平细颗粒物的比例要大得多。此外，空间分布的网络能够解决火山喷发顺风处的火山羽流的化学演化问题。测量发现平均 SO ₂大约 36 小时的转换时间，证明了分布式 LCS 网络在现实环境中观察反应动力学和量化空气污染物化学转化的能力。这项工作还强调了 LCS 网络在极端事件期间应急响应的效用，以补充现有的空气质量监测方法。