Source and variability of formaldehyde (HCHO) at northern high latitude: an integrated satellite, ground/aircraft, and model study

Here we use satellite observations of HCHO vertical column densities (VCD) from the TROPOspheric Monitoring Instrument (TROPOMI), ground-based and aircraft measurements, 20 combined with a nested regional chemical transport model (GEOS-Chem at 0.5°×0.625° resolution), to understand the variability and sources of summertime HCHO better in Alaska. We first evaluate GEOSChem with in-situ airborne measurements during Atmospheric Tomography Mission 1 (ATom-1) aircraft campaign and ground-based measurements from Multi-AXis Differential Optical Absorption Spectroscopy (MAX-DOAS). We show reasonable agreement between observed and modeled HCHO, 25 isoprene and monoterpenes. In particular, HCHO profiles show spatial homogeneity in Alaska, suggesting a minor contribution of biogenic emissions to HCHO VCD. We further examine the TROPOMI HCHO product in Alaska during boreal summer, which is in good agreement with GEOSChem model results. We find that HCHO VCDs are dominated by free-tropospheric background in wildfire-free regions. During the summer of 2018, the model suggests that the background HCHO 30 column, resulting from methane oxidation, contributes to 66 to 80% of the HCHO VCD, while wildfires contribute to 14% and biogenic VOC contributes to 5 to 9% respectively. For the summer of 2019, which had intense wildfires, the model suggests that wildfires contribute to 40 to 65%, and the background column accounts for 30 to 50% of HCHO VCD in June and July. In particular, the model indicates a major contribution of wildfires from direct emissions of HCHO, instead of secondary 35 production of HCHO from oxidation of larger VOCs. We find that the column contributed by biogenic https://doi.org/10.5194/acp-2021-820 Preprint. Discussion started: 11 October 2021 c © Author(s) 2021. CC BY 4.0 License.

Here SCD1,SAT is the measured slant column density, SCDRef,SAT is the background slant column correction in reference sector. AMFSAT is the air mass factor provided by the TROPOMI HCHO product. AMF0,SAT is the air mass factor for the background column in the reference sector. VCD0,CTM is the vertical column in reference sector calculated by a CTM model (TM5-MP CTM), in the TROPOMI 190 HCHO product.
Following S5P TROPOMI HCHO L2 user manual (Veefkind et al., 2012), we applied several criteria to ensure the data quality. This includes: (1) quality assurance values (QA) greater than 0.5; (2) cloud fraction at 340 nm less than 0.5; (3) Solar Zenith Angle (SZA) less than 60º;(4) surface albedo less than 195 0.1, and (5) derived AMF greater than 0.1. In particular, northern Alaska can be covered by snow and ice even in summer with the criteria of surface albedo. We do not use the data over snow/ice surface as the retrieval algorithm may not work well on these surfaces (De Smedt et al., 2018). We use the overpass data in the local time window 12:00-15:00 AKDT (20:00-23:00 UTC).

200
To compare the HCHO column density from TROPOMI with our model, we recalculate the AMF based on vertical shapes from GEOS-Chem simulations and scattering weight from TROPOMI HCHO product. This method has been applied in a number of previous studies (Palmer et al., 2001;Boersma et al., 2004;González Abad et al., 2015;Zhu et al., 2016).
Here ( ) is the column density of the air parcel at vertical air pressure , for a specific air column.
, is the total column of the specific air column. ( ) is scattering weight of TROPOMI HCHO product at each altitude, calculated by the product of TROPOMI averaging kernel ASAT(p) and air mass factor AMFSAT. is surface layer pressure.

210
The standard TROPOMI HCHO VCD relies on the TM5-MP CTM for background HCHO fields, but in this work, we are using a different CTM, GEOS-Chem. Therefore, to be consistent with the GEOS-Chem CTM, we reprocessed TROPOMI HCHO vertical column that by replacing the original background (VCD0,SAT in Equation (1)) with VCD0,GC from GEOS-Chem background simulation (González Abad et al., 2015;Kaiser et al., 2018).We assume that background is approximately equal to 215 GEOS-Chem 0, and neglect the variability of the AMF0,GC/AMFGC (AMF0,GC is AMFGC in reference sector averaged in 5 latitude bins) (De Smedt et al., 2018). The GEOS-Chem background simulation is performed over reference sector and excludes biogenic and biomass burning emissions (Table 1). Finally, the reprocessed TROPOMI HCHO VCD is expressed as: Our reprocessed TROPOMI HCHO VCD might have several advantages over the TROPOMI HCHO operational product that is based on TM5-MP model. First, our reprocessed HCHO VCD are based on GEOS-Chem nested simulation with finer resolution than TM5-MP model. Second, our GEOS-Chem simulation includes year-specific wildfire emissions that was not available for TM5-MP model when 225 TROPOMI operational product was produced. As we show below, this reprocessed HCHO VCD shows higher values in central Alaska than the original product, leading to a better agreement with model results.

ATom-1 aircraft campaign
The NASA Atmospheric Tomography (ATom) studied atmospheric composition in remote regions 230 (Wofsy et al., 2018). ATom had four phases over a 4-year period, with each phase sampling the global atmosphere in one of four seasons. ATom deployed a comprehensive gas and aerosol particle measurement payload on the NASA DC-8 aircraft. During ATom-1, two flights performed vertical profiling over Alaska during August 1-3 in 2016. We make use of HCHO measurement by Laser Induced Fluorescence technique (Cazorla et al., 2015) and VOC measurement by whole air sampling 235 (WAS) followed by laboratory Gas Chromatography (GC) analysis (Simpson et al., 2020) during these flights to evaluate model performance on HCHO, isoprene and monoterpenes ( -pinene and -pinene).
We use 1 minutes averaged data for HCHO and 3-5 minutes average data for isoprene and monoterpenes. The reported measurement uncertainties are 10% for HCHO and 10% for isoprene and monoterpenes. 240

MAX-DOAS
The Multi-AXis Differential Optical Absorption Spectroscopy (MAX-DOAS) measurement technique is employed to measure atmospheric trace gases such as HCHO at urban and remote sites (Honninger, 2004). Previous HCHO measurements by MAX-DOAS spectroscopy provided ground validation for satellite HCHO retrievals and model results (Pinardi et al., 2013). A comprehensive description of 245 MAX-DOAS retrieval algorithm theoretical basis can be found in Honninger et al (2004).
In this study, we use HCHO VCD timeseries from two MAX-DOAS instruments deployed in Fairbanks Fairbanks is in the central Alaska boreal forest region, and TFS is located on the North Slope and is covered by tundra. The two instruments sample profiles every 12 minutes, which are averaged to 2-hour intervals for this work and are selected to be in the 12:00-15:00 local time window. In this study, the VCD is calculated using the formula VCD = dSCD20° / dAMF20° where the dSCD20° is the measured differential slant columns, the difference in HCHO absorption between a 20° elevation angle and the 255 zenith view. The dAMF20° = 1.93 is calculated geometrically (Ma et al., 2013), and 20° was chosen as the highest elevation in all measurement sequences. Wildfire polluted records are removed by selecting UV visibility > 5 km; foggy records are removed by selecting dewpoint depression < 1.5 C (at Fairbanks) or < 80% RH (at TFS). The footprint of MAX-DOAS is about 20 km for clear-sky conditions, shorter with clouds or high particulate loading. The 2 detection limit of VCD = 1.010 15 260 molecules cm -2 by this geometric method.

Nested GEOS-Chem simulation
Here we use GEOS-Chem v12.5.0 (doi: 10.5281/zenodo.3403111). GEOS-Chem is a 3-D global  (Guenther et al., 2006(Guenther et al., , 2012. In this work, BVOC emission activity factors are calculated online, expressed as: Here is a standard environment coefficient normalizing to 1 under standard environmental condition. LAI is the leaf area index (m 2 m -2 ), and are emission activity factors accounting for light and temperature effects, respectively. is calculated based on the photosynthetic photon flux density (PPFD) (µmol of photons in 400-700 nm range m -2 s -1 ). Terrestrial vegetation for BVOC emissions is based on the plant functional type (PFT) distribution derived from Community Land Model 290 (CLM4) (Lawrence et al., 2011;Oleson et al., 2013). CLM4 output suggests two dominating PFTs in the continent of Alaska: needle leaf evergreen boreal tree (mainly in the interior boreal forest region) and broadleaf deciduous boreal shrub (mainly over north slope and southwest Alaska), both with high emission factors in isoprene (3000 µg m -2 h -1 and 4000 µg m -2 h -1 respectively) and low EFs in monoterpenes ( -pinine + -pinine, 800 µg m -2 h -1 and 300 µg m -2 h -1 respectively). Thus, we expect a 295 major contribution from isoprene to BVOC emissions in Alaska in model results. Despite that shrub has a higher emission factor of isoprene, we expect a larger isoprene emission flux from central Alaska boreal forest region due to warmer "continental" temperatures and higher LAI.
In this work we use the detailed O3-NOx-HOx-VOC chemistry ("tropchem" mechanism) (Park et al., 300 2004;Mao et al., 2010Mao et al., , 2013, with updates on isoprene chemistry (Fisher et al., 2016). This version of isoprene chemistry in GEOS-Chem have been extensively evaluated by recent field campaign data and satellite observations (Fisher et al., 2016;Travis et al., 2016), including HCHO production from isoprene oxidation (Zhu et al., 2016, Kaiser et al. 2018. In general, under high-NOx condition (1 ppbv), the HCHO production is prompt, reaching 70-80% of its maximum yield within a few hours. 305 While under low-NOx condition (0.1 ppbv or lower), it takes several days to reach the maximum yield and the cumulative yield is still lower than the high-NOx condition by a factor of 2-3 (Marais et al., 2012). As we show below, this slow production of HCHO under low-NOx conditions leads to weak but widespread HCHO enhancement in regional scale.
To examine the influence of different sources on HCHO columns in Alaska, we conducted a series of nested GEOS-Chem simulations, as described in Table 1. The background HCHO column (VCD0,GC) is calculated from a GEOS-Chem simulation where biogenic emission and biomass burning emission are turned off. The HCHO differential column induced by wildfire or biogenic emission is derived from the difference between the control run and the run with wildfire or biogenic emission turned off. 315  Figure 1 (ac) shows measured vertical profiles of formaldehyde, isoprene and monoterpenes across the Alaska domain during ATom-1. In Figure 1(a), the measured HCHO mixing ratio decreases exponentially from surface (around 320 pptv) to the upper troposphere (around 100 pptv). The HCHO surface mixing ratio in Alaska is an order of magnitude lower than other high-BVOC regions such as 325

Model evaluation by ATom-1
Southeast US (Li et al., 2016). (c) show that observed isoprene and monoterpenes have much higher mixing ratios in the lowest 2 km layer than above. The mean observed isoprene mixing ratio is about 120 pptv in the boundary layer, a factor of three higher than that of monoterpenes. As isoprene has a shorter lifetime 330

Figures 1(b) and
(1.1 hours) than monoterpene (2.1 hours), this indicates a stronger isoprene emission flux than monoterpene emission flux in Alaskan boreal forest. The predominance of isoprene emission in Alaskan boreal forest is different from some European boreal forests, where monoterpenes are often the predominant BVOC species (Juráň et al., 2017;Bäck et al., 2012).

335
To evaluate model performance with ATom-1 measurements, a nested GEOS-Chem simulation is conducted during ATom-1 mission period over Alaska. We sampled the model along the flight track at the flight time with 1-hour model time resolution for comparisons between model and observations. As shown in Figure 1(d) -(f), our nested GEOS-Chem model well reproduce the ATom-1 vertical and spatial variability of HCHO, isoprene and monoterpenes mixing ratios. Modeled isoprene and 340 monoterpenes mixing ratios concentrate in the surface layer (0 -2 km) and show a median value of around 100 pptv and 10 pptv respectively. Modeled isoprene mixing ratio is comparable with ATom-1 observations, while monoterpenes mixing ratio is lower than ATom-1 averaged value in lower than 2 km (around 40 pptv).

345
One remarkable feature in Figure 1 is the spatial homogeneity in HCHO vertical profiles, as shown in both observations and model. We find that all sampled HCHO vertical profiles in Alaska show similar magnitude and vertical distribution, despite different land types and locations of these sampled profiles.
The homogeneity is not observed in isoprene and monoterpene mixing ratios, which show maximums in central and south Alaska, where boreal forests are located ( Figure S1). Such spatial discrepancies 350 between HCHO and isoprene/monoterpenes suggest a minor contribution of biogenic VOC emissions to HCHO column density.
We further examine the abundance of isoprene and monoterpenes in Alaska with available surface VOC measurements from field campaigns at TFS. Angot et al. (2020)

370
The MAX-DOAS technique is most sensitive to boundary layer species and loses sensitivity significantly above the first kilometer or two. The ATom vertical profiles shown in Figure 1 indicate that "background" HCHO has a large fraction of its column well above the lowest kilometers, which would be detected with a lower sensitivity. This effect appears to make the simple geometric approximation used here to retrieve HCHO have a sensitivity more like the differential VCD calculated 375 in GEOS-Chem (dVCDGC), which also mainly concentrates in the < 2 km layer ( Figure S4). For this reason, here we compare MAX-DOAS HCHO total column retrieval VCDMD with GEOS-Chem dVCDGC. A detailed comparison between GEOS-Chem and MAX-DOAS retrievals using an optimal estimation method that appropriately deals with the reduced sensitivity aloft will be described in a follow-up study. 380

Evaluating TROPOMI HCHO product
In this section we evaluate the TROPOMI HCHO product over Alaska during the summer of 2018 and 2019. As noted above, these two years differ substantially on local wildfire emissions, providing useful information on satellite capability of detecting biogenic and wildfire HCHO in remote regions. 410

Predominance of background chemistry in mild wildfire summer
In Figure 3 To understand the drivers for HCHO variability, we first examine the background HCHO VCD provided by GEOS-Chem (VCD0,GC). Figure 3(b) shows that from 2018 May to August, VCD0,GC in 420 central Alaska increases from 2.0×10 15 molecules cm -2 to 3.5×10 15 molecules cm -2 , then decreases to 2.6×10 15 molecules cm -2 , accounting for 66%-80% of VCDSAT,GC. This indicates that VCDSAT,GC is largely dominated by background signals VCD0,GC in 2018. The spatial pattern of VCD0,GC, most noticeable in July, is largely driven by the geography in Alaska. As the majority of HCHO VCD stems from lowest atmospheric layers (Figure 1), the high elevation in the Alaska Range in southern Alaska 425 (63N, 151W, peaks at Denali, elevation 6190 m) and the Brooks Range in northern Alaska (68N, 152W, peaks at Mount Isto, elevation 2736 m) are responsible for the significantly lower HCHO VCD in these regions. We also find high VCD0,GC (2.0-3.210 15 molecules cm -2 ) over northern Pacific in July and August, due to enhanced methane oxidation via CH3O2 + NO reactions near surface and CH3O2 + CH3O2 at higher altitudes. This enhanced methane oxidation also leads to temperature dependence of 430 VCD0,GC ( Figure S2).

445
To quantify the sources of HCHO dVCD, we derive two variables: dVCD induced by wildfire emission (dVCDGC,Fire) and biogenic emission (dVCDGC,Bio), computed by the differences between model control run and sensitivity runs with wildfire or biogenic emissions turned off (Table 1). Figure 4(c) that dVCDGC,Bio presents a similar spatial pattern and monthly cycle as modeled 450 isoprene emission ( Figure S6), with high values over central boreal forest region (4.610 14 molecules cm -2 ) and low values in other parts (5.0-8.010 13 molecules cm -2 ). The widespread biogenic HCHO enhancement can be in part explained by the slow photooxidation in Alaska under low NOx conditions (25~35 pptv near surface in GEOS-Chem). Indeed, the HCHO production from isoprene and monoterpene emissions are lower under low NOx conditions than high NOx conditions by a factor of 10 455 after 24-h oxidation, and it only reaches 20% of its 5-day cumulative yield (Marais et al., 2012). As a result, dVCDGC,Bio in Alaska is lower than that in mid-latitude by more than a factor of 10 for the same amount of isoprene emissions.

We show in
Despite the relatively weak Alaskan fire in 2018 summer, we find a higher fraction of dVCDGC,Fire than 460 dVCDGC,Bio in total dVCDGC. Figure 4(b) shows several regions with high dVCDGC,Fire (1.010 15 molecules cm -2 ), often co-located with fire hot spots. The GFED4s burning area measured by MODIS is shown in Figure S5. A model sensitivity test in 2018 suggests that over 90% of dVCDGC,Fire is from wildfire direct emission, instead of secondary production of HCHO from oxidation of other VOCs. It is partly due to the missing of wildfire VOC emissions (Akagi et al., 2011)

Wildfire emission impacts HCHO column in Alaska
We further examine the summer of 2019. Figure 5 . We emphasize that modeled direct emissions of HCHO from wildfires contributes to 56% of dVCDGC,Fire. Consequently, dVCDGC,Fire is higher than dVCDGC,Bio by a factor of 10 in 2019 Alaska 510 summer, despite that NMVOC from wildfires (498 GgC) are only higher than biogenic emissions (374 GgC) by a factor of 1 to 2.

Uncertainty and capability of TROPOMI in capturing biogenic emission HCHO signals
The total uncertainty of TROPOMI HCHO in Alaska is composed of random and systematic 520 uncertainties and contributed by errors in dSCDSAT, AMFSAT and VCD0,SAT. According to TROPOMI HCHO ATBD, systematic uncertainties from AMFSAT accounts for 30-50% of total columns, while the total contribution of AMFSAT uncertainties is around 75% of total column uncertainty. We expect the uncertainty in AMFSAT to be larger for pixels containing fire smoke due to errors in a priori profile (Zhu et al., 2020). We find that for regions with heavy smoke, our calculated GEOS-Chem AMFGC is 50% 525 lower than TROPOMI AMFSAT provided, due to the difference in HCHO a priori profiles ( Figure S3), suggesting large uncertainty in the HCHO a priori vertical profiles used for retrieval. In 2019 July, VCDSAT,GC in central Alaska is enhanced by 3.0-5.010 15 molecules cm -2 than VCDSAT after applying the AMFGC that based on GEOS-Chem HCHO profile a priori ( Figure S10). Scattering and absorbing aerosols can also introduce large uncertainties to HCHO AMF by changing the observed radiance 530 (Gonzi et al., 2011;Jung et al., 2019), especially over strong biomass burning scenes when AMF can be very sensitive to the vertical profiles of aerosols (Barkley et al., 2012;Fu et al., 2007). errors can also be due to errors of radiation transfer model and other external parameters like cloud fraction, surface albedo etc. The aerosol and cloud-related error can be as high as 30% of total columns, due to the relative low aerosol layer height (around 1 km) of the wildfire smoke in Alaska (Jung et al., 2019).
contributes to the second most majority of random and systematic errors in VCDSAT. The systematic error in dSCDSAT contributed by reference sector correction leads to the bias pattern of dSCDSAT in Figure 4(a) and Figure 6(a), within the error range of reference sector correction (0-4.010 15 molecules cm -2 ) in TROPOMI HCHO Algorithm Theoretical Basis Document (ATBD). The 540 negative over southwest Alaska and Gulf of Alaska can be partly contributed by overcorrection in processing dSCDSAT. During the correction, values higher than 5.010 16 molecules cm -2 are removed to remove wildfire signals, but the criteria is much higher than wildfire related HCHO enhancement in Alaska. According to the bias found in , The systematic slant columns uncertainty in Alaska can contribute higher than 25% of . 545 Based on ATBD, the uncertainty in VCD0,SAT is estimated to be 0. found a bias in OMI HCHO when comparing to the whole ATom 1-2 dataset. Since Alaska lies in the reference sector defined by most retrieval algorithms (González Abad et al., 2015;De Smedt et al., 2018), errors in background correction can lead to bias in corrected slant column dSCDSAT, not only in Alaska but also in all northern high latitudes region.

Conclusions and discussions
The Arctic / boreal terrestrial ecosystem is undergoing rapid changes in recent decades, but VOC emissions from Arctic and boreal vegetation and wildfires remains poorly quantified, limiting our capability for understanding biosphere-atmosphere exchange in this region and its feedback on Arctic climate and air quality. HCHO serves as an important indicator for biogenic and wildfire VOC 570 emissions. In this work, we use satellite-based observations of HCHO VCD from the TROPOMI instrument on-board S5P satellite, and ground-based measurements of HCHO VCD from MAX-DOAS, combined with a nested grid regional chemical transport model (GEOS-Chem at (0.50.625), to examine source and variability of HCHO VCD in Alaska for the summers of 2018 and 2019.

575
We first evaluate the GEOS-Chem nested simulation with in-situ airborne measurements from the August 2016 ATom-1 mission and ground-based MAX-DOAS observations in summer 2018 and 2019.
Our model well reproduces magnitude and vertical distribution of HCHO, isoprene and monoterpenes abundance when wildfire is weak. Both measurements and model highlight the spatial homogeneity in HCHO vertical profiles, suggesting a minor contribution of biogenic VOCs to HCHO VCD. With a 580 high sensitivity to near surface signals, MAX-DOAS measurements provides evaluation for modeled biogenic HCHO variability. With a simple geometric approximation to retrieve HCHO from MAX-DOAS measurements, we show that MAX-DOAS HCHO retrievals agree well with model results on the seasonal trend of HCHO signal in both Fairbanks and Toolik Field Station. We also find good correlation between MAX-DOAS HCHO and modeled isoprene emissions. Future work is warranted to 585 investigate MAX-DOAS retrievals with optimal estimation method and its comparison with model results.
We further compared the model results to TROPOMI HCHO L2 product, reprocessed with background HCHO VCD and AMF using GEOS-Chem model output. GEOS-Chem provides HCHO vertical 590 profiles a priori and background columns in a higher horizontal and vertical resolution than TM5-MP https://doi.org/10.5194/acp-2021-820 Preprint. Discussion started: 11 October 2021 c Author(s) 2021. CC BY 4.0 License.
CTM, the default model used in TROPOMI HCHO product. GEOS-Chem profiles includes the wildfire emission of the corresponding year, which TM5-MP did not, could be another advantage of GEOS-Chem. These advantages of GEOS-Chem model may improve the reliability of reprocessed TROPOMI HCHO column (VCDSAT,GC). We find that TROPOMI HCHO VCDSAT,GC in a mild wildfire summer is 595 dominated by background HCHO VCD0,GC from methane oxidation. We find that wildfires have a larger contribution to HCHO total column than biogenic emissions, even in a year with mild wildfires.
This result is in part due to the direct emission of HCHO from wildfires, and in part due to the slow and small production of HCHO from isoprene and monoterpenes oxidation under low NOx conditions. We find that HCHO VCD from biogenic VOC is too small for TROPOMI to be able to detect. 600 For the year with large wildfires (2019), we find that TROPOMI and model show good agreement on magnitude and spatial pattern of HCHO VCD, and wildfire becomes the largest contributor to HCHO VCD inside fire-related enhancements. To a large extent this is driven by the direct emission of HCHO from wildfires. We consider this good agreement to be partly fortuitous, due to uncertainties associated 605 with satellite retrieval in smoke conditions, emissions strength and speciation, and detailed chemical mechanism for HCHO production. However, we show that wildfire emission signals are detectable in TROPOMI HCHO product, making TROPOMI a semi-quantitative tool to constrain wildfire emissions in Alaska. As the Arctic and boreal region continue to warm, we expect HCHO VCD in Alaska will continue to be driven by wildfires and background methane oxidation. atmospheric composition for climate, air quality and ozone layer applications, Remote Sensing of Environment, 120, 70-83,