Stochastic Time-Inverted Lagrangian Transport Research Articles

Potential of using CO2 observations over India in a regional carbon budget estimation by improving the modelling system

Abstract. Devising effective national-level climate action plans requires a more detailed understanding of the regional distribution of sources and sinks of greenhouse gases. Due to insufficient observations and modelling capabilities, India's current carbon source–sink estimates are uncertain. This study uses a high-resolution Lagrangian transport model to examine the potential of available CO2 observations over India for inverse estimation of regional carbon fluxes. We use four different sites in India that vary in the measurement technique, frequency and spatial representation. These observations exhibit substantial seasonal (7.5 to 9.2 ppm) and intra-seasonal (2 to 12 ppm) variability. Our modelling framework, a high-resolution Weather Research and Forecasting Model combined with the Stochastic Time-Inverted Lagrangian Transport model (WRF–STILT), performs better in simulating seasonal (R2=0.50 to 0.96) and diurnal (R2=0.96) variability (for the Mohali station) of observed CO2 than the current-generation global models (CarboScope, CarbonTracker and ECMWF EGG4). The seasonal CO2 concentration variability in Mohali, associated with crop residue burning, is largely underestimated by the models. WRF–STILT captures the seasonal biospheric variability over Nainital better than the global models but underestimates the strength of the CO2 uptake by crops. The choice of emission inventory in the modelling framework alone leads to significant biases in simulations (5 to 10 ppm), endorsing the need for accounting for emission fluxes, especially for non-background sites. Our study highlights the possibility of using the CO2 observations from these Indian stations for deducing carbon flux information at regional (Nainital) and suburban to urban (Mohali, Shadnagar and Nagpur) scales with the help of a high-resolution model. On accounting for observed variability in CO2, the global carbon data assimilation system can benefit from the measurements from the Indian subcontinent.

Improved estimates of smoke exposure during Australia fire seasons: importance of quantifying plume injection heights

Abstract. Wildfires can have a significant impact on air quality in Australia during severe burning seasons, but incomplete knowledge of the injection heights of smoke plumes poses a challenge for quantifying smoke exposure. In this study, we use two approaches to quantify the fractions of fire emissions injected above the planetary boundary layer (PBL), and we further investigate the impact of plume injection fractions on daily mean surface concentrations of fine particulate matter (PM2.5) from wildfire smoke in key cities over northern and southeastern Australia from 2009 to 2020. For the first method, we rely on climatological, monthly mean vertical profiles of smoke emissions from the Integrated Monitoring and Modelling System for wildland fires (IS4FIRES) together with assimilated PBL heights from NASA Modern-Era Retrospective Analysis for Research and Application (MERRA) version 2. For the second method, we develop a novel approach based on the Multi-angle Imaging SpectroRadiometer (MISR) observations and a random forest, machine learning model that allows us to directly predict the daily plume injection fractions above the PBL in each grid cell. We apply the resulting plume injection fractions quantified by the two methods to smoke PM2.5 concentrations simulated by the Stochastic Time-Inverted Lagrangian Transport (STILT) model in target cities. We find that characterization of the plume injection heights greatly affects estimates of surface daily smoke PM2.5, especially during severe wildfire seasons, when intense heat from fires can loft smoke high in the troposphere. However, using climatological injection profiles cannot capture well the spatiotemporal variability in plume injection fractions, resulting in a 63 % underestimation of daily fire emission fluxes injected above the PBL in comparison with those fluxes derived from MISR injection fractions. Our random forest model successfully reproduces the daily injected fire emission fluxes against MISR observations (R2=0.88, normalized mean bias = 10 %) and predicts that 27 % and 45 % of total fire emissions rise above the PBL in northern and southeastern Australia, respectively, from 2009 to 2020. Using the plume behavior predicted by the random forest method also leads to better model agreement with observed surface PM2.5 in several key cities near the wildfire source regions, with smoke PM2.5 accounting for 5 %–52 % of total PM2.5 during fire seasons from 2009 to 2020.

Anthropogenic CO2 emission estimates in the Tokyo metropolitan area from ground-based CO2 column observations

Abstract. Urban areas are responsible for more than 40 % of global energy-related carbon dioxide (CO2) emissions. The Tokyo metropolitan area (TMA), Japan, one of the most populated regions in the world, includes various emission sources, such as thermal power plants, automobile traffic, and residential facilities. In order to infer a top–down emission estimate, we conducted an intensive field campaign in the TMA from February to April 2016 to measure column-averaged dry-air mole fractions of CO2 (XCO2) with three ground-based Fourier transform spectrometers (one IFS 125HR and two EM27/SUN spectrometers). At two urban sites (Saitama and Sodegaura), measured XCO2 values were generally larger than those at a rural site (Tsukuba) by up to 9.5 ppm, and average diurnal variations increased toward evening. To simulate the XCO2 enhancement (ΔXCO2) resulting from emissions at each observation site, we used the Stochastic Time-Inverted Lagrangian Transport (STILT) model driven by meteorological fields at a horizontal resolution of ∼1 km from the Weather Research and Forecasting (WRF) model, which was coupled with anthropogenic (large point source and area source) CO2 emissions and biogenic fluxes. Although some of the diurnal variation of ΔXCO2 was not reproduced and plumes from nearby large point sources were not captured, primarily because of a transport modeling error, the WRF–STILT simulations using prior fluxes were generally in good agreement with the observations (mean bias, 0.30 ppm; standard deviation, 1.31 ppm). By combining observations with high-resolution modeling, we developed an urban-scale inversion system in which spatially resolved CO2 emission fluxes at >3 km resolution and a scaling factor of large point source emissions were estimated on a monthly basis by using Bayesian inference. The XCO2 simulation results from the posterior CO2 fluxes were improved (mean bias, −0.03 ppm; standard deviation, 1.21 ppm). The prior and posterior total CO2 emissions in the TMA are 1.026 ± 0.116 and 1.037 ± 0.054 Mt-CO2 d−1 at the 95 % confidence level, respectively. The posterior total CO2 emissions agreed with emission inventories within the posterior uncertainty, demonstrating that the EM27/SUN spectrometer data can constrain urban-scale monthly CO2 emissions.

Understanding greenhouse gas (GHG) column concentrations in Munich using the Weather Research and Forecasting (WRF) model

Abstract. To address ambitious goals of carbon neutrality set at national and city scales, a number of atmospheric networks have been deployed to monitor greenhouse gas (GHG) concentrations in and around cities. To convert these measurements into estimates of emissions from cities, atmospheric models are used to simulate the transport of various trace gases and help interpret these measurements. We set up a modelling framework using the Weather Research and Forecasting (WRF) model applied at a high spatial resolution (up to 400 m) to simulate the atmospheric transport of GHGs and attempt a preliminary interpretation of the observations provided by the Munich Urban Carbon Column Network (MUCCnet). Building on previous analyses using similar measurements performed within a campaign for the city of Berlin and its surroundings (Zhao et al., 2019), our modelling framework has been improved regarding the initialization of tagged tracers, model settings, and input data. To assess the model performance, we validate the modelled output against two local weather stations and two radiosonde observations, as well as observed column GHG concentrations. The measurements were provided by the measurement campaign that was carried out from 1 to 30 August 2018. The modelled wind matches well with the measurements from the weather stations, with wind speeds slightly overestimated. In general, the model is able to reproduce the measured slant column concentrations of CH4 and their variability, while for CO2, a difference in the slant column CO2 of around 3.7 ppm is found in the model. This can be attributed to the initial and lateral boundary conditions used for the background tracer. Additional mismatches in the diurnal cycle could be explained by an underestimation of nocturnal respiration in the modelled CO2 biogenic fluxes. The differential column method (DCM) has been applied to cancel out the influence from the background concentrations. We optimize its application by selecting suitable days on which the assumption of the DCM holds true: a relatively uniform air mass travels over the city, passing from an upwind site to a downwind site. In particular, the Stochastic Time-Inverted Lagrangian Transport (STILT) model is used here and driven by our WRF-modelled meteorological fields to obtain footprints (i.e. the potential areas of influence for signals observed at measurement stations), further used for interpreting measurement results. Combining these footprints with local knowledge of emission sources, we find evidence of CH4 sources near Munich that are missing or underestimated in the emission inventory used. This demonstrates the potential of this data–model framework to constrain local sources and improve emission inventories.

Latitudinal distribution and sources analysis of greenhouse gases and air pollutants observed during the 2021 Yellow Sea Air Quality campaign aboard a research vessel

The Yellow Sea serves as a pathway for air pollution transport from the Asian continent into Korea and is annually monitored by the Korean Meteorological Administration (KMA) research vessel during the spring season. This study investigated the spatio-temporal variation of long-range transported greenhouse gases (CO2 and CH4) and air pollutants (CO, NOx, SO2, O3) in the Yellow Sea during the Yellow Sea Air Quality (YES-AQ) campaign. Our analysis reveals that GHGs and air pollutants exhibit a clear latitudinal distribution. Lower concentrations are observed in the range of 31.2–34.5°N, primarily influenced by air masses traversing the ocean and the western part of Korean Peninsula. In contrast, higher concentrations occur more frequently at 34.5–37.8°N, primarily affected by continental air masses. The regression slopes of CO and CO2 vary depending on regional characteristics, particularly in air originating from Korea and China. Regression slope analysis between CO, CO2, and CH4, except for high plume of CH4 (>2.05 ppm), indicates that fossil fuel sources play a significant role. A high plume of CH4 (>2.05 ppm) was observed on April 22nd, coinciding with the maximum O3 concentrations during the period of highest temperatures throughout the voyage. The potential source regions were identified in the west-southern coast of Korea, using footprint analysis from Stochastic Time-Inverted Lagrangian Transport (STILT) and column-averaged dry-air mole fraction of CH4 (XCH4) retrieval from the TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5 Precursor (S5P) satellite. The high CH4 enhancements are most likely due to biogenic emissions triggered by favorable near-surface temperatures for microbial activities in soil. Furthermore, the high CH4 enhancements were not correlated with CO and CO2, indicating that the source was not related to fossil fuel combustion. On March 29–30, a high plume of NOx was detected and is likely attributed to the Bohai Sea, which is located just above the Yellow Sea and surrounded by highly industrialized and semi-enclosed areas.

Validation of the Atmospheric Dispersion Model NAME against Long-Range Tracer Release Experiments

Abstract The Met Office’s atmospheric dispersion model Numerical Atmospheric-Dispersion Modeling Environment (NAME) is validated against controlled tracer release experiments, considering the impact of the driving meteorological data and choices in the parameterization of unresolved motions. The Cross-Appalachian Tracer Experiment (CAPTEX) and Across North America Tracer Experiment (ANATEX) were long-range dispersion experiments in which inert tracers were released and the air concentrations measured across North America in the 1980s. NAME simulations of the experiments have been driven by both reanalysis meteorological data from European Centre for Medium-Range Weather Forecasts (ECMWF) and data from the Advanced Research version of the Weather Research and Forecasting (WRF) Model. NAME predictions of air concentrations are assessed against the experimental measurements, using a ranking method composed of four statistical parameters. Differences in the performance of NAME according to this ranking method are compared when driven by different meteorological sources. The effect of changing parameter values in NAME for the unresolved mesoscale motions parameterization is also considered, in particular, whether the parameter values giving the best performance rank are consistent with values typically used. The performance ranks are compared with analyses in the literature for other particle dispersion models, namely, Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT), Stochastic Time-Inverted Lagrangian Transport (STILT), and Flexible Particle (FLEXPART). It is found that NAME performance is comparable to the other dispersion models considered, with the different models responding similarly to differences in driving meteorological data.

Sources and sinks of carbonyl sulfide inferred from tower and mobile atmospheric observations in the Netherlands

Abstract. Carbonyl sulfide (COS) is a promising tracer for the estimation of terrestrial ecosystem gross primary production (GPP). However, understanding its non-GPP-related sources and sinks, e.g., anthropogenic sources and soil sources and sinks, is also critical to the success of the approach. Here we infer the regional sources and sinks of COS using continuous in situ mole fraction profile measurements of COS along the 60 m tall Lutjewad tower (1 m a.s.l.; 53∘24′ N, 6∘21′ E) in the Netherlands. To identify potential sources that caused the observed enhancements of COS mole fractions at Lutjewad, both discrete flask samples and in situ measurements in the province of Groningen were made from a mobile van using a quantum cascade laser spectrometer (QCLS). We also simulated the COS mole fractions at Lutjewad using the Stochastic Time-Inverted Lagrangian Transport (STILT) model combined with emission inventories and plant uptake fluxes. We determined the nighttime COS fluxes to be -3.0±2.6 pmol m−2 s−1 using the radon-tracer correlation approach and Lutjewad observations. Furthermore, we identified and quantified several COS sources, including biodigesters, sugar production facilities and silicon carbide production facilities in the province of Groningen. Moreover, the simulation results show that the observed COS enhancements can be partially explained by known industrial sources of COS and CS2, in particular from the Ruhr Valley (51.5∘ N, 7.2∘ E) and Antwerp (51.2∘ N, 4.4∘ E) areas. The contribution of likely missing anthropogenic sources of COS and CS2 in the inventory may be significant. The impact of the identified sources in the province of Groningen is estimated to be negligible in terms of the observed COS enhancements. However, in specific conditions, these sources may influence the measurements in Lutjewad. These results are valuable for improving our understanding of the sources and sinks of COS, contributing to the use of COS as a tracer for GPP.

Anthropogenic carbon dioxide origin tracing study in Anmyeon-do, South Korea: Based on STILT-footprint and emissions data

Tracking the origin of greenhouse gases (GHGs) is critical for understanding regional GHG concentration variation and developing effective policies to reduce GHG emissions. This study provides quantitative information about the surface contribution to enhancement of carbon dioxide (CO2) concentration at Anmyeon-do (AMY), South Korea, using Stochastic Time-Inverted Lagrangian Transport (STILT) model and anthropogenic CO2 emission data. CO2 enhancement simulated by the STILT and emission data was positively correlated with measured CO2 anomalies at AMY with a correlation coefficient above 0.5. High and low CO2 days were selected using ground measurements of the CO2 mixing ratio at AMY during the winter season of 2018–2019. The surface contributions for the high and low CO2 days at AMY were compared quantitatively. When the high concentration was observed in AMY, the CO2 enhancements were dominated by domestic regions, especially from the metropolitan area in South Korea, due to the high footprint and large CO2 emissions. From the perspective of foreign regions, the surface contribution of eastern China regions (Shandong, Jiangsu-Shanghai) increased during high CO2 days compared to low CO2 days at AMY. During the high CO2 days, the ratio between CO2 and carbon monoxide, a co-emitted species, is large when the surface contribution of eastern China regions is relatively strong due to different regional combustion efficiency (i.e., high combustion efficiency in South Korea compared to that in China). The surface contribution based on STILT and emission data is useful for understanding the cause of high GHG concentration at the receptor (AMY in this study).

Why do inverse models disagree? A case study with two European CO2 inversions

Abstract. We present an analysis of atmospheric transport impact on estimating CO2 fluxes using two atmospheric inversion systems (CarboScope-Regional (CSR) and Lund University Modular Inversion Algorithm (LUMIA)) over Europe in 2018. The main focus of this study is to quantify the dominant drivers of spread amid CO2 estimates derived from atmospheric tracer inversions. The Lagrangian transport models STILT (Stochastic Time-Inverted Lagrangian Transport) and FLEXPART (FLEXible PARTicle) were used to assess the impact of mesoscale transport. The impact of lateral boundary conditions for CO2 was assessed by using two different estimates from the global inversion systems CarboScope (TM3) and TM5-4DVAR. CO2 estimates calculated with an ensemble of eight inversions differing in the regional and global transport models, as well as the inversion systems, show a relatively large spread for the annual fluxes, ranging between −0.72 and 0.20 PgC yr−1, which is larger than the a priori uncertainty of 0.47 PgC yr−1. The discrepancies in annual budget are primarily caused by differences in the mesoscale transport model (0.51 PgC yr−1), in comparison with 0.23 and 0.10 PgC yr−1 that resulted from the far-field contributions and the inversion systems, respectively. Additionally, varying the mesoscale transport caused large discrepancies in spatial and temporal patterns, while changing the lateral boundary conditions led to more homogeneous spatial and temporal impact. We further investigated the origin of the discrepancies between transport models. The meteorological forcing parameters (forecasts versus reanalysis obtained from ECMWF data products) used to drive the transport models are responsible for a small part of the differences in CO2 estimates, but the largest impact seems to come from the transport model schemes. Although a good convergence in the differences between the inversion systems was achieved by applying a strict protocol of using identical prior fluxes and atmospheric datasets, there was a non-negligible impact arising from applying a different inversion system. Specifically, the choice of prior error structure accounted for a large part of system-to-system differences.

Near-real-time CO2 fluxes from CarbonTracker Europe for high-resolution atmospheric modeling

Abstract. We present the CarbonTracker Europe High-Resolution (CTE-HR) system that estimates carbon dioxide (CO2) exchange over Europe at high resolution (0.1 × 0.2∘) and in near real time (about 2 months' latency). It includes a dynamic anthropogenic emission model, which uses easily available statistics on economic activity, energy use, and weather to generate anthropogenic emissions with dynamic time profiles at high spatial and temporal resolution (0.1×0.2∘, hourly). Hourly net ecosystem productivity (NEP) calculated by the Simple Biosphere model Version 4 (SiB4) is driven by meteorology from the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis 5th Generation (ERA5) dataset. This NEP is downscaled to 0.1×0.2∘ using the high-resolution Coordination of Information on the Environment (CORINE) land-cover map and combined with the Global Fire Assimilation System (GFAS) fire emissions to create terrestrial carbon fluxes. Ocean CO2 fluxes are included in our product, based on Jena CarboScope ocean CO2 fluxes, which are downscaled using wind speed and temperature. Jointly, these flux estimates enable modeling of atmospheric CO2 mole fractions over Europe. We assess the skill of the CTE-HR CO2 fluxes (a) to reproduce observed anomalies in biospheric fluxes and atmospheric CO2 mole fractions during the 2018 European drought, (b) to capture the reduction of anthropogenic emissions due to COVID-19 lockdowns, (c) to match mole fraction observations at Integrated Carbon Observation System (ICOS) sites across Europe after atmospheric transport with the Transport Model, version 5 (TM5) and the Stochastic Time-Inverted Lagrangian Transport (STILT), driven by ECMWF-IFS, and (d) to capture the magnitude and variability of measured CO2 fluxes in the city center of Amsterdam (the Netherlands). We show that CTE-HR fluxes reproduce large-scale flux anomalies reported in previous studies for both biospheric fluxes (drought of 2018) and anthropogenic emissions (COVID-19 pandemic in 2020). After applying transport of emitted CO2, the CTE-HR fluxes have lower median root mean square errors (RMSEs) relative to mole fraction observations than fluxes from a non-informed flux estimate, in which biosphere fluxes are scaled to match the global growth rate of CO2 (poor person's inversion). RMSEs are close to those of the reanalysis with the CTE data assimilation system. This is encouraging given that CTE-HR fluxes did not profit from the weekly assimilation of CO2 observations as in CTE. We furthermore compare CO2 concentration observations at the Dutch Lutjewad coastal tower with high-resolution STILT transport to show that the high-resolution fluxes manifest variability due to different emission sectors in summer and winter. Interestingly, in periods where synoptic-scale transport variability dominates CO2 concentration variations, the CTE-HR fluxes perform similarly to low-resolution fluxes (5–10× coarsened). The remaining 10 % of the simulated CO2 mole fraction differs by >2 ppm between the low-resolution and high-resolution flux representation and is clearly associated with coherent structures (“plumes”) originating from emission hotspots such as power plants. We therefore note that the added resolution of our product will matter most for very specific locations and times when used for atmospheric CO2 modeling. Finally, in a densely populated region like the Amsterdam city center, our modeled fluxes underestimate the magnitude of measured eddy covariance fluxes but capture their substantial diurnal variations in summertime and wintertime well. We conclude that our product is a promising tool for modeling the European carbon budget at a high resolution in near real time. The fluxes are freely available from the ICOS Carbon Portal (CC-BY-4.0) to be used for near-real-time monitoring and modeling, for example, as an a priori flux product in a CO2 data assimilation system. The data are available at https://doi.org/10.18160/20Z1-AYJ2 (van der Woude, 2022a).

Effects of point source emission heights in WRF–STILT: a step towards exploiting nocturnal observations in models

Abstract. An appropriate representation of point source emissions in atmospheric transport models is very challenging. In the Stochastic Time-Inverted Lagrangian Transport model (STILT), all point source emissions are typically released from the surface, meaning that the actual emission stack height plus subsequent plume rise is not considered. This can lead to erroneous predictions of trace gas concentrations, especially during nighttime when vertical atmospheric mixing is minimal. In this study we use two Weather Research and Forecasting (WRF)–STILT model approaches to simulate fossil fuel CO2 (ffCO2) concentrations: (1) the standard “surface source influence (SSI)” approach and (2) an alternative “volume source influence (VSI)” approach where nearby point sources release CO2 according to their effective emission height profiles. The comparison with 14C-based measured ffCO2 data from 2-week integrated afternoon and nighttime samples collected at Heidelberg, 30 m above ground level shows that the root-mean-square deviation (RMSD) between modelled and measured ffCO2 is indeed almost twice as high during the night (RMSD =6.3 ppm) compared to the afternoon (RMSD =3.7 ppm) when using the standard SSI approach. In contrast, the VSI approach leads to a much better performance at nighttime (RMSD =3.4 ppm), which is similar to its performance during afternoon (RMSD =3.7 ppm). Representing nearby point source emissions with the VSI approach could thus be a first step towards exploiting nocturnal observations in STILT. The ability to use nighttime observations in atmospheric inversions would dramatically increase the observational data and allow for the investigation of different source mixtures or diurnal cycles. To further investigate the differences between these two approaches, we conducted a model experiment in which we simulated the ffCO2 contributions from 12 artificial power plants with typical annual emissions of 1 million tonnes of CO2 and with distances between 5 and 200 km from the Heidelberg observation site. We find that such a power plant must be more than 50 km away from the observation site in order for the mean modelled ffCO2 concentration difference between the SSI and VSI approach to fall below 0.1 ppm during situations with low mixing heights smaller than 500 m.

Short-term reduction of regional enhancement of atmospheric CO2 in China during the first COVID-19 pandemic period

Recent studies have reported a 9% decrease in global carbon emissions during the COVID-19 lockdown period; however, its impact on the variation of atmospheric CO2 level remains under question. Using atmospheric CO2 observed at Anmyeondo station (AMY) in South Korea, downstream of China, this study examines whether the decrease in China’s emissions due to COVID-19 can be detected from the enhancement of CO2 mole fraction (ΔCO2) relative to the background value. The Weather Research and Forecasting–Stochastic Time-Inverted Lagrangian Transport model was applied to determine when the observed mole fractions at AMY were affected by air parcels from China. Atmospheric observations at AMY showed up to a −20% (−1.92 ppm) decrease in ΔCO2 between February and March 2020 compared to the same period in 2018 and 2019, particularly with a −34% (−3.61 ppm) decrease in March. ΔCO, which was analyzed to explore the short-term effect of emission reductions, had a decrease of −43% (−80.66 ppb) during the lockdown in China. Particularly in East China, where emissions are more concentrated than in Northeast China, ΔCO2 and ΔCO decreased by −44% and −65%, respectively. The ΔCO/ΔCO2 ratio (24.8 ppb ppm−1), which is the indicator of emission characteristics, did not show a significant difference before and after the COVID-19 lockdown period (α = 0.05), suggesting that this decrease in ΔCO2 and ΔCO was associated with emission reductions rather than changes in emission sources or combustion efficiency in China. Reduced carbon emissions due to limited human activity resulted in a decrease in the short-term regional enhancement to the observed atmospheric CO2.

Incorporating features from the Stochastic Time-Inverted Lagrangian Transport (STILT) model into the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model: a unified dispersion model for time-forward and time-reversed applications

AbstractThe Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model is a state-of-the-science atmospheric dispersion model that is developed and maintained at the National Oceanic Atmospheric Administration’s (NOAA) Air Resources Laboratory (ARL). In the early 2000s, HYSPLIT served as the starting point for development of the Stochastic Time-Inverted Lagrangian Transport (STILT) model that emphasizes backward-in-time dispersion simulations to determine source regions of receptors. STILT continued its separate development and gained a wide user base. Since STILT was built on a now outdated version of HYSPLIT and lacks long-term institutional support to maintain the model, incorporating STILT features into HYSPLIT allows these features to stay up to date. This paper describes the STILT features incorporated into HYSPLIT, which include: a new vertical interpolation algorithm for WRF derived meteorological input files, a detailed algorithm for estimating boundary layer height, a new turbulence parameterization, a vertical Lagrangian timescale that varies in time and space, a complex dispersion algorithm, and two new convection schemes. An evaluation of these new features was performed using tracer release data from the Cross Appalachian Tracer Experiment and the Across North America Tracer Experiment. Results show the dispersion module from STILT, which takes up to double the amount of time to run, is less dispersive in the vertical and in better agreement with observations than the existing HYSPLIT option. The other new modeling features from STILT were not consistently statistically different than existing HYSPLIT options. Forward-time simulations from the new model were also compared against backward-time equivalents and found to be statistically comparable to one another.

A dedicated flask sampling strategy developed for Integrated Carbon Observation System (ICOS) stations based on CO<sub>2</sub> and CO measurements and Stochastic Time-Inverted Lagrangian Transport (STILT) footprint modelling

Abstract. In situ CO2 and CO measurements from five Integrated Carbon Observation System (ICOS) atmosphere stations have been analysed together with footprint model runs from the regional Stochastic Time-Inverted Lagrangian Transport (STILT) model to develop a dedicated strategy for flask sampling with an automated sampler. Flask sampling in ICOS has three different purposes, namely (1) to provide an independent quality control for in situ observations, (2) to provide representative information on atmospheric components currently not monitored in situ at the stations, and (3) to collect samples for 14CO2 analysis that are significantly influenced by fossil fuel CO2 (ffCO2) emission areas. Based on the existing data and experimental results obtained at the Heidelberg pilot station with a prototype flask sampler, we suggest that single flask samples are collected regularly every third day around noon or in the afternoon from the highest level of a tower station. Air samples shall be collected over 1 h, with equal temporal weighting, to obtain a true hourly mean. At all stations studied, more than 50 % of flasks collected around midday will likely be sampled during low ambient variability (<0.5 parts per million (ppm) standard deviation of 1 min values). Based on a first application at the Hohenpeißenberg ICOS site, such flask data are principally suitable for detecting CO2 concentration biases larger than 0.1 ppm with a 1σ confidence level between flask and in situ observations from only five flask comparisons. In order to have a maximum chance to also sample ffCO2 emission areas, additional flasks are collected on all other days in the afternoon. To check if the ffCO2 component will indeed be large in these samples, we use the continuous in situ CO observations. The CO deviation from an estimated background value is determined the day after each flask sampling, and depending on this offset, an automated decision is made as to whether a flask shall be retained for 14CO2 analysis. It turned out that, based on existing data, ffCO2 events of more than 4–5 ppm that would allow ffCO2 estimates with an uncertainty below 30 % were very rare at all stations studied, particularly in summer (only zero to five events per month from May to August). During the other seasons, events could be collected more frequently. The strategy developed in this project is currently being implemented at the ICOS stations.

On deriving influences of upwind agricultural and anthropogenic emissions on greenhouse gas concentrations and air quality over Delhi in India: A stochastic Lagrangian footprint approach

Delhi, the capital city of India witnesses severe degradation of air quality and rapid enhancement of trace gases during winter. Still it is unclear about the relative role of the meteorological conditions and the post-monsoon agricultural stubble burning on the occurrence of these events. To overcome this, we examine the use of applying high-resolution transport model to establish the link between atmospheric concentrations and upstream surface fluxes. This study reports the implementation of a Lagrangian approach and demonstrates its capability in deriving the upwind influences over Delhi. We simulate stochastic back trajectories over Delhi by implementing stochastic time-inverted Lagrangian transport (STILT) model, driven by the meteorological fields from the European Centre for Medium Range Weather Forecasts (ECMWF) model. During the post-monsoon, when mixing layer height is shallow, we find high near-field influence. The variations in footprint simulations with receptor heights show the effect of mixing layer dynamics on the air-parcels. By using atemporal emission fields, we find a considerable impact of meteorological conditions during November that contributes to the enhancements of trace gases. Together with strong emissions (anthropogenic and biomass burning), these enhancements can be several orders higher compared to other seasons. Through the receptor-oriented STILT implementation over India, we envision a wide range of applications spanning from air quality to climate change. An advantage of this implementation is that it allows the use of pre-calculated footprints in simulating any trace gas species and particulate matter, making it computationally less demanding than running an ensemble of full atmospheric transport model.

Atmospheric inverse estimates of CO emissions from Zhengzhou, China

Carbon monoxide (CO) is an important gas that affects human health and causes air pollution. However, the estimates of CO emissions in China are still subject to large uncertainties. Based on the CO mass concentration and the coupled Weather Research and Forecast (WRF) and Stochastic Time-Inverted Lagrangian Transport (STILT) model (WRF-STILT), this study estimates the CO emissions over Zhengzhou, China. The results show that the mean CO mass concentration was 1.17 mg m−3 from November 2017 to February 2018, with a clear diurnal variation. There were two periods of rapidly increasing CO concentration in the diurnal variation, which are 06:00–09:00 and 16:00–20:00 local time. The footprint analysis shows that the observation site is highly influenced by local emissions. The most influential regions to the site observations are northeast and northwest Zhengzhou, which are associated with the geographical barrier of the Taihang Mountains in the north and narrow Fenwei Plain in the west. The inversion result shows that the actual emissions are lower than the inventory estimates. Using the optimal scaling factors, the WRF-STILT simulations of CO concentration agree closely with the CO measurements with the linear fitting regression equation y = 0.87x + 0.15. The slopes of the linear fitting regressions between the WRF-STILT-simulated CO concentrations determined using the optimal emissions and the observations range from 0.72 to 0.89 for four months, and all the fitting results passed the significance test (P < 0.001). These results indicate that the new optimal emissions derived with the scaling factors could better represent the real emission conditions than the a priori emissions if the WRF-STILT model is assumed to be reliable.

Using Space-Based Observations and Lagrangian Modeling to Evaluate Urban Carbon Dioxide Emissions in the Middle East.

Improved observational understanding of urban CO2 emissions, a large and dynamic global source of fossil CO2, can provide essential insights for both carbon cycle science and mitigation decision making. Here we compare three distinct global CO2 emissions inventory representations of urban CO2 emissions for five Middle Eastern cities (Riyadh, Mecca, Tabuk, Jeddah, and Baghdad) and use independent satellite observations from the Orbiting Carbon Observatory‐2 (OCO‐2) satellite to evaluate the inventory representations of afternoon emissions. We use the column version of the Stochastic Time‐Inverted Lagrangian Transport (X‐STILT) model to account for atmospheric transport and link emissions to observations. We compare XCO2 simulations with observations to determine optimum inventory scaling factors. Applying these factors, we find that the average summed emissions for all five cities are 100 MtC year−1 (50–151, 90% CI), which is 2.0 (1.0, 3.0) times the average prior inventory magnitudes. The total adjustment of the emissions of these cities comes out to ~7% (0%, 14%) of total Middle Eastern emissions (~700 MtC year−1). We find our results to be insensitive to the prior spatial distributions in inventories of the cities' emissions, facilitating robust quantitative assessments of urban emission magnitudes without accurate high‐resolution gridded inventories.

Wintertime spatial distribution of ammonia and its emission sources in the Great Salt Lake region

Abstract. Ammonium-containing aerosols are a major component of wintertime air pollution in many densely populated regions around the world. Especially in mountain basins, the formation of persistent cold-air pools (PCAPs) can enhance particulate matter with diameters less than 2.5 µm (PM2.5) to levels above air quality standards. Under these conditions, PM2.5 in the Great Salt Lake region of northern Utah has been shown to be primarily composed of ammonium nitrate; however, its formation processes and sources of its precursors are not fully understood. Hence, it is key to understanding the emission sources of its gas phase precursor, ammonia (NH3). To investigate the formation of ammonium nitrate, a suite of trace gases and aerosol composition were sampled from the NOAA Twin Otter aircraft during the Utah Winter Fine Particulate Study (UWFPS) in January and February 2017. NH3 was measured using a quantum cascade tunable infrared laser differential absorption spectrometer (QC-TILDAS), while aerosol composition, including particulate ammonium (pNH4), was measured with an aerosol mass spectrometer (AMS). The origin of the sampled air masses was investigated using the Stochastic Time-Inverted Lagrangian Transport (STILT) model and combined with an NH3 emission inventory to obtain model-predicted NHx (=NH3+pNH4) enhancements. Enhancements represent the increase in NH3 mixing ratios within the last 24 h due to emissions within the model footprint. Comparison of these NHx enhancements with measured NHx from the Twin Otter shows that modelled values are a factor of 1.6 to 4.4 lower for the three major valleys in the region. Among these, the underestimation is largest for Cache Valley, an area with intensive agricultural activities. We find that one explanation for the underestimation of wintertime emissions may be the seasonality factors applied to NH3 emissions from livestock. An investigation of inter-valley exchange revealed that transport of NH3 between major valleys was limited and PM2.5 in Salt Lake Valley (the most densely populated area in Utah) was not significantly impacted by NH3 from the agricultural areas in Cache Valley. We found that in Salt Lake Valley around two thirds of NHx originated within the valley, while about 30 % originated from mobile sources and 60 % from area source emissions in the region. For Cache Valley, a large fraction of NOx potentially leading to PM2.5 formation may not be locally emitted but mixed in from other counties.

Novel approaches to improve estimates of short-lived halocarbon emissions during summer from the Southern Ocean using airborne observations

Abstract. Fluxes of halogenated volatile organic compounds (VOCs) over the Southern Ocean remain poorly understood, and few atmospheric measurements exist to constrain modeled emissions of these compounds. We present observations of CHBr3, CH2Br2, CH3I, CHClBr2, CHBrCl2, and CH3Br during the O2∕N2 Ratio and CO2 Airborne Southern Ocean (ORCAS) study and the second Atmospheric Tomography mission (ATom-2) in January and February of 2016 and 2017. Good model–measurement correlations were obtained between these observations and simulations from the Community Earth System Model (CESM) atmospheric component with chemistry (CAM-Chem) for CHBr3, CH2Br2, CH3I, and CHClBr2 but all showed significant differences in model : measurement ratios. The model : measurement comparison for CH3Br was satisfactory and for CHBrCl2 the low levels present precluded us from making a complete assessment. Thereafter, we demonstrate two novel approaches to estimate halogenated VOC fluxes; the first approach takes advantage of the robust relationships that were found between airborne observations of O2 and CHBr3, CH2Br2, and CHClBr2. We use these linear regressions with O2 and modeled O2 distributions to infer a biological flux of halogenated VOCs. The second approach uses the Stochastic Time-Inverted Lagrangian Transport (STILT) particle dispersion model to explore the relationships between observed mixing ratios and the product of the upstream surface influence of sea ice, chl a, absorption due to detritus, and downward shortwave radiation at the surface, which in turn relate to various regional hypothesized sources of halogenated VOCs such as marine phytoplankton, phytoplankton in sea-ice brines, and decomposing organic matter in surface seawater. These relationships can help evaluate the likelihood of particular halogenated VOC sources and in the case of statistically significant correlations, such as was found for CH3I, may be used to derive an estimated flux field. Our results are consistent with a biogenic regional source of CHBr3 and both nonbiological and biological sources of CH3I over these regions.

Influence of the Underlying Surface on Greenhouse Gas Concentrations in the Atmosphere Over Central Siberia

A crucial issue in atmospheric studies on greenhouse gas content involves assessing the representativeness (footprint) having influence on their concentrations measured by tall towers. In this study, the Stochastic Time-Inverted Lagrangian Transport (STILT) model was used to estimate seasonal cumulative footprint climatology for greenhouse gases measurements obtained on the 301-meter-high Zotino Tall Tower Observation Facility (ZOTTO) for the growing seasons (May-September) from 2008 to 2012 (with the exception of 2011). Results showed that the ZOTTO seasonal concentration cumulative footprint climatology for four years reached 6.9×106 km2 and the 75% cumulative footprints varied from 1.9 to 2.3×106 km2. For the same period, the Russian Land Cover map based on MODIS data for 2014 was used to estimate the impact of land cover surrounding the ZOTTO tower on concentration measurements. The analysis showed that in the 75% seasonal cumulative footprint the largest area is occupied by bogs, followed (in decreasing order) by larch, mixed, light-coniferous evergreen forests, grassland, and by other classes. Furthermore, analysis of the contributions from individual cells making up a footprint showed that the largest influence on formation of greenhouse gas concentrations as recorded by ZOTTO comes from the types of vegetation growing in the immediate vicinity of the tall tower, namely bogs, mixed forests, and light and dark coniferous forest stands.