Tracking Reactive Nitrogen Plumes and their Evolution from Satellite Observations

Abstract

Satellite remote sensing offers near-continuous global coverage and plays a critical role in addressing observational gaps for many atmospheric trace gases. Reactive nitrogen species are critical in atmospheric chemistry because they drive tropospheric ozone (O3) formation and contribute to the production of secondary aerosols. This dissertation presents novel satellite- based observations and methodologies for analyzing reactive nitrogen trace gases, specifically Peroxyacyl Nitrates (PANs) and ammonia (NH3). We investigate the variability in intercontinental air pollution transport of PANs and the chemical evolution of NH3 within smoke plumes. We introduce new approaches for using multiple satellite products together to isolate enhancements in plumes and extracting information on chemical evolution in the context of highly variable background concentrations. We further develop these methods by implementing a machine learning–based retrieval framework for trace gas observations of NH3. In Chapter 2, we leverage global satellite observations of PANs from the Cross-track Infrared Sounder (CrIS) on the Suomi National Polar-orbiting Partnership (S-NPP) satellite to evaluate the seasonal and interannual variations of intercontinental transport in the Northern Hemisphere between 2016 and 2022. We find that April and July are dominant months for transpacific transport of PANs and summer months (June, July and August) are dominant months for transatlantic transport. There is significant interannual variability over the study period during the months where the intercontinental transport of PANs is largest. We use CrIS PANs combined with NO2 from the Ozone Monitoring Instrument (OMI) to explore changes to the intercontinental transport of PANs associated with major decreases in precursor emissions in response to COVID-19. CrIS observations indicate statistically meaningful decreases in PANs over regions in both the Pacific and the Atlantic Ocean basins compared to Pre-COVID years 2016-2019; the changes in PANs are smaller than the changes in NO2. May 2020 CrIS observations indicate PANs (OMI NO2) declined over the NW Pacific by ~11% (~33%), NE Pacific by ~8% (~15%), USA-Atlantic outflow region by ~4% (~4%), and Atlantic by ~11% (~11%). The largest change in PANs occurred over the NW Pacific in February 2020, where PANs (OMI NO2) decreased ~16% (~42%) compared to Pre-COVID years. We also use a chemical transport model to simulate PAN changes in response to the pandemic emissions changes and find the model is consistent with the observed changes in PANs. Our observations suggest the values of PANs over the ocean basins have not fully rebounded to Pre-COVID values which is consistent with the trend in tropospheric column NO2. Biomass burning is one of the largest natural sources of gas-phase ammonia (NH3) to the atmosphere which has implications for secondary aerosol formation. In Chapter 3, we use measurements of NH3 and carbon monoxide (CO) from the Cross-track Infrared Sounder (CrIS) aboard the Suomi-National Polar orbiting Partnership (S-NPP) satellite to investigate the emissions and evolution of NH3 in smoke plumes. Our analysis focuses on wildfires over the western United States during summer 2018, a period that coincides with the Western Wildfire Experiment for Cloud Chemistry, Aerosol Absorption, and Nitrogen (WE-CAN) aircraft campaign. We present approaches for 1) separating NH3 enhancements from transient biomass burning events from persistent agricultural hotspots from satellite platforms, and 2) correcting height bias in the standard NH3 retrievals to improve applicability to lofted wildfire smoke. We calculate normalized excess mixing ratios (NEMR) of NH3 with respect to CO within smoke plumes to characterize evolution as the fires evolve in time and the plumes are transported downwind. The calculated NEMRs are also used to track the temporal evolution of NH3 enhancements near fire centroids over 25 days of burning and to examine their relationship with fire radiative power (FRP). We find higher distributions of NEMRs to be associated with lower FRPs, where median NEMRs > 0.10 are always associated with median FRPs < 50 MW. We incorporate data from the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) platform, which enables further interpretation of the height of a smoke plume. Chapter 4 presents the first application of a machine learning (ML)-based retrieval approach to extend optimal estimation (OE)-based retrievals of NH3 and CO from CrIS in the context of wildfire smoke analysis. Using this novel dataset, we examine NH3 enhancements relative to CO in smoke plumes during active fire years, demonstrating the method’s ability to capture the evolution of NH3 within individual plumes. Comparative analysis during the 2018 wildfire season shows qualitative agreement between ML- and OE-based retrievals in isolating smoke plumes and assessing plume characteristics, including NH3 NEMRs and their decay with plume age. Case studies, such as the Pole Creek Fire in Utah, illustrate this alignment across key metrics. These findings demonstrate the capability of ML-enhanced CrIS retrievals to assess chemical evolution in wildfire smoke, and this ML-based approach offers the potential for significantly more data to be processed and used.